thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-14 19:45:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'master' into sql-for-workload-management

Browse Source
This commit is contained in:
serxa 2024-10-29 09:09:05 +00:00
commit 78c8b43479
1006 changed files with 26734 additions and 24082 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
.gitmodules vendored
View File

@ -227,12 +227,6 @@
[submodule "contrib/minizip-ng"]
path = contrib/minizip-ng
url = https://github.com/zlib-ng/minizip-ng
[submodule "contrib/qpl"]
path = contrib/qpl
url = https://github.com/intel/qpl
[submodule "contrib/idxd-config"]
path = contrib/idxd-config
url = https://github.com/intel/idxd-config
[submodule "contrib/QAT-ZSTD-Plugin"]
path = contrib/QAT-ZSTD-Plugin
url = https://github.com/intel/QAT-ZSTD-Plugin

2
CMakeLists.txt
View File

@ -545,7 +545,7 @@ endif()
if (CMAKE_BUILD_TYPE_UC STREQUAL "RELWITHDEBINFO"
AND NOT SANITIZE AND NOT SANITIZE_COVERAGE AND NOT ENABLE_FUZZING
AND OS_LINUX AND (ARCH_AMD64 OR ARCH_AARCH64))
AND OMIT_HEAVY_DEBUG_SYMBOLS AND OS_LINUX AND (ARCH_AMD64 OR ARCH_AARCH64))
set(CHECK_LARGE_OBJECT_SIZES_DEFAULT ON)
else ()
set(CHECK_LARGE_OBJECT_SIZES_DEFAULT OFF)

8
base/base/StringRef.h
View File

@ -369,11 +369,15 @@ namespace PackedZeroTraits
{
template <typename Second, template <typename, typename> class PackedPairNoInit>
inline bool check(const PackedPairNoInit<StringRef, Second> p)
{ return 0 == p.key.size; }
{
return 0 == p.key.size;
}
template <typename Second, template <typename, typename> class PackedPairNoInit>
inline void set(PackedPairNoInit<StringRef, Second> & p)
{ p.key.size = 0; }
{
p.key.size = 0;
}
}

5
base/glibc-compatibility/CMakeLists.txt
View File

@ -11,6 +11,9 @@ if (GLIBC_COMPATIBILITY)
if (ARCH_AARCH64)
list (APPEND glibc_compatibility_sources musl/aarch64/syscall.s musl/aarch64/longjmp.s)
set (musl_arch_include_dir musl/aarch64)
# Disable getauxval in aarch64. ARM glibc minimum requirement for the project is 2.18 and getauxval is present
# in 2.16. Having a custom one introduces issues with sanitizers
list (REMOVE_ITEM glibc_compatibility_sources musl/getauxval.c)
elseif (ARCH_AMD64)
list (APPEND glibc_compatibility_sources musl/x86_64/syscall.s musl/x86_64/longjmp.s)
set (musl_arch_include_dir musl/x86_64)
@ -18,7 +21,7 @@ if (GLIBC_COMPATIBILITY)
message (FATAL_ERROR "glibc_compatibility can only be used on x86_64 or aarch64.")
endif ()
if (SANITIZE STREQUAL thread)
if (SANITIZE STREQUAL thread AND ARCH_AMD64)
# Disable TSAN instrumentation that conflicts with re-exec due to high ASLR entropy using getauxval
# See longer comment in __auxv_init_procfs
# In the case of tsan we need to make sure getauxval is not instrumented as that would introduce tsan

7
base/glibc-compatibility/musl/getauxval.c
View File

@ -25,9 +25,10 @@
// We don't have libc struct available here.
// Compute aux vector manually (from /proc/self/auxv).
//
// Right now there is only 51 AT_* constants,
// so 64 should be enough until this implementation will be replaced with musl.
static unsigned long __auxv_procfs[64];
// Right now there are 51 AT_* constants. Custom kernels have been encountered
// making use of up to 71. 128 should be enough until this implementation is
// replaced with musl.
static unsigned long __auxv_procfs[128];
static unsigned long __auxv_secure = 0;
// Common
static unsigned long * __auxv_environ = NULL;

2
base/poco/Foundation/include/Poco/Logger.h
View File

@ -952,6 +952,8 @@ private:
static std::pair<LoggerMapIterator, bool> add(Logger * pLogger);
static std::optional<LoggerMapIterator> find(const std::string & name);
static Logger * findRawPtr(const std::string & name);
void unsafeSetChannel(Channel * pChannel);
Channel* unsafeGetChannel() const;
Logger();
Logger(const Logger &);

21
base/poco/Foundation/src/Logger.cpp
View File

@ -61,6 +61,13 @@ Logger::~Logger()
void Logger::setChannel(Channel* pChannel)
{
std::lock_guard<std::mutex> lock(getLoggerMutex());
unsafeSetChannel(pChannel);
}
void Logger::unsafeSetChannel(Channel* pChannel)
{
if (_pChannel) _pChannel->release();
_pChannel = pChannel;
@ -69,6 +76,14 @@ void Logger::setChannel(Channel* pChannel)
Channel* Logger::getChannel() const
{
std::lock_guard<std::mutex> lock(getLoggerMutex());
return unsafeGetChannel();
}
Channel* Logger::unsafeGetChannel() const
{
return _pChannel;
}
@ -89,7 +104,7 @@ void Logger::setLevel(const std::string& level)
void Logger::setProperty(const std::string& name, const std::string& value)
{
if (name == "channel")
setChannel(LoggingRegistry::defaultRegistry().channelForName(value));
unsafeSetChannel(LoggingRegistry::defaultRegistry().channelForName(value));
else if (name == "level")
setLevel(value);
else
@ -160,7 +175,7 @@ void Logger::setChannel(const std::string& name, Channel* pChannel)
if (len == 0 ||
(it.first.compare(0, len, name) == 0 && (it.first.length() == len || it.first[len] == '.')))
{
it.second.logger->setChannel(pChannel);
it.second.logger->unsafeSetChannel(pChannel);
}
}
}
@ -393,7 +408,7 @@ std::pair<Logger::LoggerMapIterator, bool> Logger::unsafeGet(const std::string&
else
{
Logger& par = parent(name);
logger = new Logger(name, par.getChannel(), par.getLevel());
logger = new Logger(name, par.unsafeGetChannel(), par.getLevel());
}
return add(logger);

1
ci/README.md Normal file
View File

@ -0,0 +1 @@
Note: This directory is under active development for CI improvements and is not currently in use within the scope of the existing CI pipeline.

109
ci/docker/fasttest/Dockerfile Normal file
View File

@ -0,0 +1,109 @@
# docker build -t clickhouse/fasttest .
FROM ubuntu:22.04
# ARG for quick switch to a given ubuntu mirror
ARG apt_archive="http://archive.ubuntu.com"
RUN sed -i "s|http://archive.ubuntu.com|$apt_archive|g" /etc/apt/sources.list
ENV DEBIAN_FRONTEND=noninteractive LLVM_VERSION=18
RUN apt-get update \
&& apt-get install \
apt-transport-https \
apt-utils \
ca-certificates \
curl \
gnupg \
lsb-release \
wget \
git \
--yes --no-install-recommends --verbose-versions \
&& export LLVM_PUBKEY_HASH="bda960a8da687a275a2078d43c111d66b1c6a893a3275271beedf266c1ff4a0cdecb429c7a5cccf9f486ea7aa43fd27f" \
&& wget -nv -O /tmp/llvm-snapshot.gpg.key https://apt.llvm.org/llvm-snapshot.gpg.key \
&& echo "${LLVM_PUBKEY_HASH} /tmp/llvm-snapshot.gpg.key" | sha384sum -c \
&& apt-key add /tmp/llvm-snapshot.gpg.key \
&& export CODENAME="$(lsb_release --codename --short | tr 'A-Z' 'a-z')" \
&& echo "deb https://apt.llvm.org/${CODENAME}/ llvm-toolchain-${CODENAME}-${LLVM_VERSION} main" >> \
/etc/apt/sources.list \
&& apt-get update \
&& apt-get install --yes --no-install-recommends --verbose-versions llvm-${LLVM_VERSION} \
&& apt-get clean \
&& rm -rf /var/lib/apt/lists/* /var/cache/debconf /tmp/*
# moreutils - provides ts fo FT
# expect, bzip2 - requried by FT
# bsdmainutils - provides hexdump for FT
# nasm - nasm copiler for one of submodules, required from normal build
# yasm - asssembler for libhdfs3, required from normal build
RUN apt-get update \
&& apt-get install \
clang-${LLVM_VERSION} \
cmake \
libclang-${LLVM_VERSION}-dev \
libclang-rt-${LLVM_VERSION}-dev \
lld-${LLVM_VERSION} \
llvm-${LLVM_VERSION}-dev \
lsof \
ninja-build \
python3 \
python3-pip \
zstd \
moreutils \
expect \
bsdmainutils \
pv \
jq \
bzip2 \
nasm \
yasm \
--yes --no-install-recommends \
&& apt-get clean \
&& rm -rf /var/lib/apt/lists/* /var/cache/debconf /tmp/*
COPY --from=clickhouse/cctools:0d6b90a7a490 /opt/gdb /opt/gdb
# Give suid to gdb to grant it attach permissions
RUN chmod u+s /opt/gdb/bin/gdb
ENV PATH="/opt/gdb/bin:${PATH}"
# This symlink is required by gcc to find the lld linker
RUN ln -s /usr/bin/lld-${LLVM_VERSION} /usr/bin/ld.lld
# FIXME: workaround for "The imported target "merge-fdata" references the file" error
# https://salsa.debian.org/pkg-llvm-team/llvm-toolchain/-/commit/992e52c0b156a5ba9c6a8a54f8c4857ddd3d371d
RUN sed -i '/_IMPORT_CHECK_FILES_FOR_\(mlir-\|llvm-bolt\|merge-fdata\|MLIR\)/ {s|^|#|}' /usr/lib/llvm-${LLVM_VERSION}/lib/cmake/llvm/LLVMExports-*.cmake
# LLVM changes paths for compiler-rt libraries. For some reason clang-18.1.8 cannot catch up libraries from default install path.
# It's very dirty workaround, better to build compiler and LLVM ourself and use it. Details: https://github.com/llvm/llvm-project/issues/95792
RUN test ! -d /usr/lib/llvm-18/lib/clang/18/lib/x86_64-pc-linux-gnu || ln -s /usr/lib/llvm-18/lib/clang/18/lib/x86_64-pc-linux-gnu /usr/lib/llvm-18/lib/clang/18/lib/x86_64-unknown-linux-gnu
ARG TARGETARCH
ARG SCCACHE_VERSION=v0.7.7
ENV SCCACHE_IGNORE_SERVER_IO_ERROR=1
# sccache requires a value for the region. So by default we use The Default Region
ENV SCCACHE_REGION=us-east-1
RUN arch=${TARGETARCH} \
&& case $arch in \
amd64) rarch=x86_64 ;; \
arm64) rarch=aarch64 ;; \
esac \
&& curl -Ls "https://github.com/mozilla/sccache/releases/download/$SCCACHE_VERSION/sccache-$SCCACHE_VERSION-$rarch-unknown-linux-musl.tar.gz" | \
tar xz -C /tmp \
&& mv "/tmp/sccache-$SCCACHE_VERSION-$rarch-unknown-linux-musl/sccache" /usr/bin \
&& rm "/tmp/sccache-$SCCACHE_VERSION-$rarch-unknown-linux-musl" -r
COPY requirements.txt /
RUN pip3 install --no-cache-dir -r /requirements.txt
# chmod 777 to make the container user independent
RUN mkdir -p /var/lib/clickhouse \
&& chmod 777 /var/lib/clickhouse
ENV TZ=Europe/Amsterdam
RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
RUN groupadd --system --gid 1000 clickhouse \
&& useradd --system --gid 1000 --uid 1000 -m clickhouse \
&& mkdir -p /.cache/sccache && chmod 777 /.cache/sccache
ENV PYTHONPATH="/wd"
ENV PYTHONUNBUFFERED=1

6
ci/docker/fasttest/requirements.txt Normal file
View File

@ -0,0 +1,6 @@
Jinja2==3.1.3
numpy==1.26.4
requests==2.32.3
pandas==1.5.3
scipy==1.12.0
#https://clickhouse-builds.s3.amazonaws.com/packages/praktika-0.1-py3-none-any.whl

0
ci_v2/docker/style-test/Dockerfile → ci/docker/style-test/Dockerfile
View File

5
ci/docker/style-test/requirements.txt Normal file
View File

@ -0,0 +1,5 @@
requests==2.32.3
yamllint==1.26.3
codespell==2.2.1
#use praktika from CH repo
#https://clickhouse-builds.s3.amazonaws.com/packages/praktika-0.1-py3-none-any.whl

102
ci/jobs/build_clickhouse.py Normal file
View File

@ -0,0 +1,102 @@
import argparse
from praktika.result import Result
from praktika.settings import Settings
from praktika.utils import MetaClasses, Shell, Utils
class JobStages(metaclass=MetaClasses.WithIter):
CHECKOUT_SUBMODULES = "checkout"
CMAKE = "cmake"
BUILD = "build"
def parse_args():
parser = argparse.ArgumentParser(description="ClickHouse Build Job")
parser.add_argument("BUILD_TYPE", help="Type: <amd|arm_debug|release_sanitizer>")
parser.add_argument("--param", help="Optional custom job start stage", default=None)
return parser.parse_args()
def main():
args = parse_args()
stop_watch = Utils.Stopwatch()
stages = list(JobStages)
stage = args.param or JobStages.CHECKOUT_SUBMODULES
if stage:
assert stage in JobStages, f"--param must be one of [{list(JobStages)}]"
print(f"Job will start from stage [{stage}]")
while stage in stages:
stages.pop(0)
stages.insert(0, stage)
cmake_build_type = "Release"
sanitizer = ""
if "debug" in args.BUILD_TYPE.lower():
print("Build type set: debug")
cmake_build_type = "Debug"
if "asan" in args.BUILD_TYPE.lower():
print("Sanitizer set: address")
sanitizer = "address"
# if Environment.is_local_run():
# build_cache_type = "disabled"
# else:
build_cache_type = "sccache"
current_directory = Utils.cwd()
build_dir = f"{Settings.TEMP_DIR}/build"
res = True
results = []
if res and JobStages.CHECKOUT_SUBMODULES in stages:
Shell.check(f"rm -rf {build_dir} && mkdir -p {build_dir}")
results.append(
Result.create_from_command_execution(
name="Checkout Submodules",
command=f"git submodule sync --recursive && git submodule init && git submodule update --depth 1 --recursive --jobs {min([Utils.cpu_count(), 20])}",
)
)
res = results[-1].is_ok()
if res and JobStages.CMAKE in stages:
results.append(
Result.create_from_command_execution(
name="Cmake configuration",
command=f"cmake --debug-trycompile -DCMAKE_VERBOSE_MAKEFILE=1 -LA -DCMAKE_BUILD_TYPE={cmake_build_type} \
-DSANITIZE={sanitizer} -DENABLE_CHECK_HEAVY_BUILDS=1 -DENABLE_CLICKHOUSE_SELF_EXTRACTING=1 -DENABLE_TESTS=0 \
-DENABLE_UTILS=0 -DCMAKE_FIND_PACKAGE_NO_PACKAGE_REGISTRY=ON -DCMAKE_INSTALL_PREFIX=/usr \
-DCMAKE_INSTALL_SYSCONFDIR=/etc -DCMAKE_INSTALL_LOCALSTATEDIR=/var -DCMAKE_SKIP_INSTALL_ALL_DEPENDENCY=ON \
-DCMAKE_C_COMPILER=clang-18 -DCMAKE_CXX_COMPILER=clang++-18 -DCOMPILER_CACHE={build_cache_type} -DENABLE_TESTS=1 \
-DENABLE_BUILD_PROFILING=1 {current_directory}",
workdir=build_dir,
with_log=True,
)
)
res = results[-1].is_ok()
if res and JobStages.BUILD in stages:
Shell.check("sccache --show-stats")
results.append(
Result.create_from_command_execution(
name="Build ClickHouse",
command="ninja clickhouse-bundle clickhouse-odbc-bridge clickhouse-library-bridge",
workdir=build_dir,
with_log=True,
)
)
Shell.check("sccache --show-stats")
Shell.check(f"ls -l {build_dir}/programs/")
res = results[-1].is_ok()
Result.create_from(results=results, stopwatch=stop_watch).finish_job_accordingly()
if __name__ == "__main__":
main()

108
ci_v2/jobs/check_style.py → ci/jobs/check_style.py
View File

@ -2,7 +2,6 @@ import math
import multiprocessing
import os
import re
import sys
from concurrent.futures import ProcessPoolExecutor
from pathlib import Path
@ -51,25 +50,6 @@ def run_check_concurrent(check_name, check_function, files, nproc=NPROC):
return result
def run_simple_check(check_name, check_function, **kwargs):
stop_watch = Utils.Stopwatch()
error = check_function(**kwargs)
result = Result(
name=check_name,
status=Result.Status.SUCCESS if not error else Result.Status.FAILED,
start_time=stop_watch.start_time,
duration=stop_watch.duration,
info=error,
)
return result
def run_check(check_name, check_function, files):
return run_check_concurrent(check_name, check_function, files, nproc=1)
def check_duplicate_includes(file_path):
includes = []
with open(file_path, "r", encoding="utf-8", errors="ignore") as f:
@ -88,7 +68,7 @@ def check_duplicate_includes(file_path):
def check_whitespaces(file_paths):
for file in file_paths:
exit_code, out, err = Shell.get_res_stdout_stderr(
f'./ci_v2/jobs/scripts/check_style/double_whitespaces.pl "{file}"',
f'./ci/jobs/scripts/check_style/double_whitespaces.pl "{file}"',
verbose=False,
)
if out or err:
@ -117,7 +97,7 @@ def check_xmllint(file_paths):
def check_functional_test_cases(files):
"""
Queries with event_date should have yesterday() not today()
NOTE: it is not that accuate, but at least something.
NOTE: it is not that accurate, but at least something.
"""
patterns = [
@ -194,7 +174,7 @@ def check_broken_links(path, exclude_paths):
def check_cpp_code():
res, out, err = Shell.get_res_stdout_stderr(
"./ci_v2/jobs/scripts/check_style/check_cpp.sh"
"./ci/jobs/scripts/check_style/check_cpp.sh"
)
if err:
out += err
@ -203,7 +183,7 @@ def check_cpp_code():
def check_repo_submodules():
res, out, err = Shell.get_res_stdout_stderr(
"./ci_v2/jobs/scripts/check_style/check_submodules.sh"
"./ci/jobs/scripts/check_style/check_submodules.sh"
)
if err:
out += err
@ -212,7 +192,7 @@ def check_repo_submodules():
def check_other():
res, out, err = Shell.get_res_stdout_stderr(
"./ci_v2/jobs/scripts/check_style/checks_to_refactor.sh"
"./ci/jobs/scripts/check_style/checks_to_refactor.sh"
)
if err:
out += err
@ -221,7 +201,7 @@ def check_other():
def check_codespell():
res, out, err = Shell.get_res_stdout_stderr(
"./ci_v2/jobs/scripts/check_style/check_typos.sh"
"./ci/jobs/scripts/check_style/check_typos.sh"
)
if err:
out += err
@ -230,7 +210,7 @@ def check_codespell():
def check_aspell():
res, out, err = Shell.get_res_stdout_stderr(
"./ci_v2/jobs/scripts/check_style/check_aspell.sh"
"./ci/jobs/scripts/check_style/check_aspell.sh"
)
if err:
out += err
@ -239,7 +219,7 @@ def check_aspell():
def check_mypy():
res, out, err = Shell.get_res_stdout_stderr(
"./ci_v2/jobs/scripts/check_style/check-mypy"
"./ci/jobs/scripts/check_style/check-mypy"
)
if err:
out += err
@ -248,7 +228,7 @@ def check_mypy():
def check_pylint():
res, out, err = Shell.get_res_stdout_stderr(
"./ci_v2/jobs/scripts/check_style/check-pylint"
"./ci/jobs/scripts/check_style/check-pylint"
)
if err:
out += err
@ -345,66 +325,58 @@ if __name__ == "__main__":
)
)
results.append(
run_check(
check_name="Check Tests Numbers",
check_function=check_gaps_in_tests_numbers,
files=functional_test_files,
Result.create_from_command_execution(
name="Check Tests Numbers",
command=check_gaps_in_tests_numbers,
command_args=[functional_test_files],
)
)
results.append(
run_simple_check(
check_name="Check Broken Symlinks",
check_function=check_broken_links,
path="./",
exclude_paths=["contrib/", "metadata/", "programs/server/data"],
Result.create_from_command_execution(
name="Check Broken Symlinks",
command=check_broken_links,
command_kwargs={
"path": "./",
"exclude_paths": ["contrib/", "metadata/", "programs/server/data"],
},
)
)
results.append(
run_simple_check(
check_name="Check CPP code",
check_function=check_cpp_code,
Result.create_from_command_execution(
name="Check CPP code",
command=check_cpp_code,
)
)
results.append(
run_simple_check(
check_name="Check Submodules",
check_function=check_repo_submodules,
Result.create_from_command_execution(
name="Check Submodules",
command=check_repo_submodules,
)
)
results.append(
run_check(
check_name="Check File Names",
check_function=check_file_names,
files=all_files,
Result.create_from_command_execution(
name="Check File Names",
command=check_file_names,
command_args=[all_files],
)
)
results.append(
run_simple_check(
check_name="Check Many Different Things",
check_function=check_other,
Result.create_from_command_execution(
name="Check Many Different Things",
command=check_other,
)
)
results.append(
run_simple_check(
check_name="Check Codespell",
check_function=check_codespell,
Result.create_from_command_execution(
name="Check Codespell",
command=check_codespell,
)
)
results.append(
run_simple_check(
check_name="Check Aspell",
check_function=check_aspell,
Result.create_from_command_execution(
name="Check Aspell",
command=check_aspell,
)
)
res = Result.create_from(results=results, stopwatch=stop_watch).dump()
if not res.is_ok():
print("Style check: failed")
for result in results:
if not result.is_ok():
print("Failed check:")
print(" | ", result)
sys.exit(1)
else:
print("Style check: ok")
Result.create_from(results=results, stopwatch=stop_watch).finish_job_accordingly()

337
ci/jobs/fast_test.py Normal file
View File

@ -0,0 +1,337 @@
import argparse
import threading
from pathlib import Path
from praktika.result import Result
from praktika.settings import Settings
from praktika.utils import MetaClasses, Shell, Utils
from ci.jobs.scripts.functional_tests_results import FTResultsProcessor
class ClickHouseProc:
def __init__(self):
self.ch_config_dir = f"{Settings.TEMP_DIR}/etc/clickhouse-server"
self.pid_file = f"{self.ch_config_dir}/clickhouse-server.pid"
self.config_file = f"{self.ch_config_dir}/config.xml"
self.user_files_path = f"{self.ch_config_dir}/user_files"
self.test_output_file = f"{Settings.OUTPUT_DIR}/test_result.txt"
self.command = f"clickhouse-server --config-file {self.config_file} --pid-file {self.pid_file} -- --path {self.ch_config_dir} --user_files_path {self.user_files_path} --top_level_domains_path {self.ch_config_dir}/top_level_domains --keeper_server.storage_path {self.ch_config_dir}/coordination"
self.proc = None
self.pid = 0
nproc = int(Utils.cpu_count() / 2)
self.fast_test_command = f"clickhouse-test --hung-check --fast-tests-only --no-random-settings --no-random-merge-tree-settings --no-long --testname --shard --zookeeper --check-zookeeper-session --order random --print-time --report-logs-stats --jobs {nproc} -- '' | ts '%Y-%m-%d %H:%M:%S' \
| tee -a \"{self.test_output_file}\""
# TODO: store info in case of failure
self.info = ""
self.info_file = ""
Utils.set_env("CLICKHOUSE_CONFIG_DIR", self.ch_config_dir)
Utils.set_env("CLICKHOUSE_CONFIG", self.config_file)
Utils.set_env("CLICKHOUSE_USER_FILES", self.user_files_path)
Utils.set_env("CLICKHOUSE_SCHEMA_FILES", f"{self.ch_config_dir}/format_schemas")
def start(self):
print("Starting ClickHouse server")
Shell.check(f"rm {self.pid_file}")
def run_clickhouse():
self.proc = Shell.run_async(
self.command, verbose=True, suppress_output=True
)
thread = threading.Thread(target=run_clickhouse)
thread.daemon = True # Allow program to exit even if thread is still running
thread.start()
# self.proc = Shell.run_async(self.command, verbose=True)
started = False
try:
for _ in range(5):
pid = Shell.get_output(f"cat {self.pid_file}").strip()
if not pid:
Utils.sleep(1)
continue
started = True
print(f"Got pid from fs [{pid}]")
_ = int(pid)
break
except Exception:
pass
if not started:
stdout = self.proc.stdout.read().strip() if self.proc.stdout else ""
stderr = self.proc.stderr.read().strip() if self.proc.stderr else ""
Utils.print_formatted_error("Failed to start ClickHouse", stdout, stderr)
return False
print(f"ClickHouse server started successfully, pid [{pid}]")
return True
def wait_ready(self):
res, out, err = 0, "", ""
attempts = 30
delay = 2
for attempt in range(attempts):
res, out, err = Shell.get_res_stdout_stderr(
'clickhouse-client --query "select 1"', verbose=True
)
if out.strip() == "1":
print("Server ready")
break
else:
print(f"Server not ready, wait")
Utils.sleep(delay)
else:
Utils.print_formatted_error(
f"Server not ready after [{attempts*delay}s]", out, err
)
return False
return True
def run_fast_test(self):
if Path(self.test_output_file).exists():
Path(self.test_output_file).unlink()
exit_code = Shell.run(self.fast_test_command)
return exit_code == 0
def terminate(self):
print("Terminate ClickHouse process")
timeout = 10
if self.proc:
Utils.terminate_process_group(self.proc.pid)
self.proc.terminate()
try:
self.proc.wait(timeout=10)
print(f"Process {self.proc.pid} terminated gracefully.")
except Exception:
print(
f"Process {self.proc.pid} did not terminate in {timeout} seconds, killing it..."
)
Utils.terminate_process_group(self.proc.pid, force=True)
self.proc.wait() # Wait for the process to be fully killed
print(f"Process {self.proc} was killed.")
def clone_submodules():
submodules_to_update = [
"contrib/sysroot",
"contrib/magic_enum",
"contrib/abseil-cpp",
"contrib/boost",
"contrib/zlib-ng",
"contrib/libxml2",
"contrib/libunwind",
"contrib/fmtlib",
"contrib/aklomp-base64",
"contrib/cctz",
"contrib/libcpuid",
"contrib/libdivide",
"contrib/double-conversion",
"contrib/llvm-project",
"contrib/lz4",
"contrib/zstd",
"contrib/fastops",
"contrib/rapidjson",
"contrib/re2",
"contrib/sparsehash-c11",
"contrib/croaring",
"contrib/miniselect",
"contrib/xz",
"contrib/dragonbox",
"contrib/fast_float",
"contrib/NuRaft",
"contrib/jemalloc",
"contrib/replxx",
"contrib/wyhash",
"contrib/c-ares",
"contrib/morton-nd",
"contrib/xxHash",
"contrib/expected",
"contrib/simdjson",
"contrib/liburing",
"contrib/libfiu",
"contrib/incbin",
"contrib/yaml-cpp",
]
res = Shell.check("git submodule sync", verbose=True, strict=True)
res = res and Shell.check("git submodule init", verbose=True, strict=True)
res = res and Shell.check(
command=f"xargs --max-procs={min([Utils.cpu_count(), 20])} --null --no-run-if-empty --max-args=1 git submodule update --depth 1 --single-branch",
stdin_str="\0".join(submodules_to_update) + "\0",
timeout=120,
retries=3,
verbose=True,
)
res = res and Shell.check("git submodule foreach git reset --hard", verbose=True)
res = res and Shell.check("git submodule foreach git checkout @ -f", verbose=True)
res = res and Shell.check("git submodule foreach git clean -xfd", verbose=True)
return res
def update_path_ch_config(config_file_path=""):
print("Updating path in clickhouse config")
config_file_path = (
config_file_path or f"{Settings.TEMP_DIR}/etc/clickhouse-server/config.xml"
)
ssl_config_file_path = (
f"{Settings.TEMP_DIR}/etc/clickhouse-server/config.d/ssl_certs.xml"
)
try:
with open(config_file_path, "r", encoding="utf-8") as file:
content = file.read()
with open(ssl_config_file_path, "r", encoding="utf-8") as file:
ssl_config_content = file.read()
content = content.replace(">/var/", f">{Settings.TEMP_DIR}/var/")
content = content.replace(">/etc/", f">{Settings.TEMP_DIR}/etc/")
ssl_config_content = ssl_config_content.replace(
">/etc/", f">{Settings.TEMP_DIR}/etc/"
)
with open(config_file_path, "w", encoding="utf-8") as file:
file.write(content)
with open(ssl_config_file_path, "w", encoding="utf-8") as file:
file.write(ssl_config_content)
except Exception as e:
print(f"ERROR: failed to update config, exception: {e}")
return False
return True
class JobStages(metaclass=MetaClasses.WithIter):
CHECKOUT_SUBMODULES = "checkout"
CMAKE = "cmake"
BUILD = "build"
CONFIG = "config"
TEST = "test"
def parse_args():
parser = argparse.ArgumentParser(description="ClickHouse Fast Test Job")
parser.add_argument("--param", help="Optional custom job start stage", default=None)
return parser.parse_args()
def main():
args = parse_args()
stop_watch = Utils.Stopwatch()
stages = list(JobStages)
stage = args.param or JobStages.CHECKOUT_SUBMODULES
if stage:
assert stage in JobStages, f"--param must be one of [{list(JobStages)}]"
print(f"Job will start from stage [{stage}]")
while stage in stages:
stages.pop(0)
stages.insert(0, stage)
current_directory = Utils.cwd()
build_dir = f"{Settings.TEMP_DIR}/build"
Utils.add_to_PATH(f"{build_dir}/programs:{current_directory}/tests")
res = True
results = []
if res and JobStages.CHECKOUT_SUBMODULES in stages:
Shell.check(f"rm -rf {build_dir} && mkdir -p {build_dir}")
results.append(
Result.create_from_command_execution(
name="Checkout Submodules for Minimal Build",
command=clone_submodules,
)
)
res = results[-1].is_ok()
if res and JobStages.CMAKE in stages:
results.append(
Result.create_from_command_execution(
name="Cmake configuration",
command=f"cmake {current_directory} -DCMAKE_CXX_COMPILER=clang++-18 -DCMAKE_C_COMPILER=clang-18 \
-DCMAKE_TOOLCHAIN_FILE={current_directory}/cmake/linux/toolchain-x86_64-musl.cmake -DENABLE_LIBRARIES=0 \
-DENABLE_TESTS=0 -DENABLE_UTILS=0 -DENABLE_THINLTO=0 -DENABLE_NURAFT=1 -DENABLE_SIMDJSON=1 \
-DENABLE_JEMALLOC=1 -DENABLE_LIBURING=1 -DENABLE_YAML_CPP=1 -DCOMPILER_CACHE=sccache",
workdir=build_dir,
with_log=True,
)
)
res = results[-1].is_ok()
if res and JobStages.BUILD in stages:
Shell.check("sccache --show-stats")
results.append(
Result.create_from_command_execution(
name="Build ClickHouse",
command="ninja clickhouse-bundle clickhouse-stripped",
workdir=build_dir,
with_log=True,
)
)
Shell.check("sccache --show-stats")
res = results[-1].is_ok()
if res and JobStages.BUILD in stages:
commands = [
f"mkdir -p {Settings.OUTPUT_DIR}/binaries",
f"cp ./programs/clickhouse {Settings.OUTPUT_DIR}/binaries/clickhouse",
f"zstd --threads=0 --force programs/clickhouse-stripped -o {Settings.OUTPUT_DIR}/binaries/clickhouse-stripped.zst",
"sccache --show-stats",
"clickhouse-client --version",
"clickhouse-test --help",
]
results.append(
Result.create_from_command_execution(
name="Check and Compress binary",
command=commands,
workdir=build_dir,
with_log=True,
)
)
res = results[-1].is_ok()
if res and JobStages.CONFIG in stages:
commands = [
f"rm -rf {Settings.TEMP_DIR}/etc/ && mkdir -p {Settings.TEMP_DIR}/etc/clickhouse-client {Settings.TEMP_DIR}/etc/clickhouse-server",
f"cp {current_directory}/programs/server/config.xml {current_directory}/programs/server/users.xml {Settings.TEMP_DIR}/etc/clickhouse-server/",
f"{current_directory}/tests/config/install.sh {Settings.TEMP_DIR}/etc/clickhouse-server {Settings.TEMP_DIR}/etc/clickhouse-client",
# f"cp -a {current_directory}/programs/server/config.d/log_to_console.xml {Settings.TEMP_DIR}/etc/clickhouse-server/config.d/",
f"rm -f {Settings.TEMP_DIR}/etc/clickhouse-server/config.d/secure_ports.xml",
update_path_ch_config,
]
results.append(
Result.create_from_command_execution(
name="Install ClickHouse Config",
command=commands,
with_log=True,
)
)
res = results[-1].is_ok()
CH = ClickHouseProc()
if res and JobStages.TEST in stages:
stop_watch_ = Utils.Stopwatch()
step_name = "Start ClickHouse Server"
print(step_name)
res = CH.start()
res = res and CH.wait_ready()
results.append(
Result.create_from(name=step_name, status=res, stopwatch=stop_watch_)
)
if res and JobStages.TEST in stages:
step_name = "Tests"
print(step_name)
res = res and CH.run_fast_test()
if res:
results.append(FTResultsProcessor(wd=Settings.OUTPUT_DIR).run())
CH.terminate()
Result.create_from(results=results, stopwatch=stop_watch).finish_job_accordingly()
if __name__ == "__main__":
main()

0
ci_v2/jobs/scripts/check_style/aspell-ignore/en/aspell-dict.txt → ci/jobs/scripts/check_style/aspell-ignore/en/aspell-dict.txt
View File

0
ci_v2/jobs/scripts/check_style/check_aspell.sh → ci/jobs/scripts/check_style/check_aspell.sh
View File

74
ci_v2/jobs/scripts/check_style/check_cpp.sh → ci/jobs/scripts/check_style/check_cpp.sh
View File

@ -14,7 +14,8 @@
LC_ALL="en_US.UTF-8"
ROOT_PATH="."
EXCLUDE_DIRS='build/|integration/|widechar_width/|glibc-compatibility/|poco/|memcpy/|consistent-hashing|benchmark|tests/.*.cpp|utils/keeper-bench/example.yaml'
EXCLUDE='build/|integration/|widechar_width/|glibc-compatibility/|poco/|memcpy/|consistent-hashing|benchmark|tests/.*.cpp|utils/keeper-bench/example.yaml'
EXCLUDE_DOCS='Settings\.cpp|FormatFactorySettingsDeclaration\.h'
# From [1]:
# But since array_to_string_internal() in array.c still loops over array
@ -31,7 +32,8 @@ function in_array()
}
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' 2>/dev/null |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
grep -vP $EXCLUDE_DOCS |
xargs grep $@ -P '((class|struct|namespace|enum|if|for|while|else|throw|switch).*|\)(\s*const)?(\s*override)?\s*)\{$|\s$|^ {1,3}[^\* ]\S|\t|^\s*(if|else if|if constexpr|else if constexpr|for|while|catch|switch)\(|\( [^\s\\]|\S \)' |
# a curly brace not in a new line, but not for the case of C++11 init or agg. initialization | trailing whitespace | number of ws not a multiple of 4, but not in the case of comment continuation | missing whitespace after for/if/while... before opening brace | whitespaces inside braces
grep -v -P '(//|:\s+\*|\$\(\()| \)"'
@ -39,39 +41,19 @@ find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' 2>/dev/n
# Tabs
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' 2>/dev/null |
grep -vP $EXCLUDE_DIRS |
xargs grep $@ -F $'\t'
grep -vP $EXCLUDE |
xargs grep $@ -F $'\t' && echo '^ tabs are not allowed'
# // namespace comments are unneeded
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' 2>/dev/null |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep $@ -P '}\s*//+\s*namespace\s*'
# Broken symlinks
find -L $ROOT_PATH -type l 2>/dev/null | grep -v contrib && echo "^ Broken symlinks found"
# Duplicated or incorrect setting declarations
SETTINGS_FILE=$(mktemp)
cat $ROOT_PATH/src/Core/Settings.cpp $ROOT_PATH/src/Core/FormatFactorySettingsDeclaration.h | grep "M(" | awk '{print substr($2, 0, length($2) - 1) " " substr($1, 3, length($1) - 3) " SettingsDeclaration" }' > ${SETTINGS_FILE}
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' | xargs grep "extern const Settings" -T | awk '{print substr($5, 0, length($5) -1) " " substr($4, 9) " " substr($1, 0, length($1) - 1)}' >> ${SETTINGS_FILE}
# Duplicate extern declarations for settings
awk '{if (seen[$0]++) print $3 " -> " $1 ;}' ${SETTINGS_FILE} | while read line;
do
echo "Found duplicated setting declaration in: $line"
done
# Incorrect declarations for settings
for setting in $(awk '{print $1 " " $2}' ${SETTINGS_FILE} | sort | uniq | awk '{ print $1 }' | sort | uniq -d);
do
expected=$(grep "^$setting " ${SETTINGS_FILE} | grep SettingsDeclaration | awk '{ print $2 }')
grep "^$setting " ${SETTINGS_FILE} | grep -v " $expected" | awk '{ print $3 " found setting " $1 " with type " $2 }' | while read line;
do
echo "In $line but it should be $expected"
done
done
rm ${SETTINGS_FILE}
bash $ROOT_PATH/utils/check-style/check-settings-style
# Unused/Undefined/Duplicates ErrorCodes/ProfileEvents/CurrentMetrics
declare -A EXTERN_TYPES
@ -91,12 +73,14 @@ EXTERN_TYPES_EXCLUDES=(
ProfileEvents::Timer
ProfileEvents::Type
ProfileEvents::TypeEnum
ProfileEvents::ValueType
ProfileEvents::dumpToMapColumn
ProfileEvents::getProfileEvents
ProfileEvents::ThreadIdToCountersSnapshot
ProfileEvents::LOCAL_NAME
ProfileEvents::keeper_profile_events
ProfileEvents::CountersIncrement
ProfileEvents::size
CurrentMetrics::add
CurrentMetrics::sub
@ -108,6 +92,7 @@ EXTERN_TYPES_EXCLUDES=(
CurrentMetrics::values
CurrentMetrics::Value
CurrentMetrics::keeper_metrics
CurrentMetrics::size
ErrorCodes::ErrorCode
ErrorCodes::getName
@ -130,7 +115,7 @@ for extern_type in ${!EXTERN_TYPES[@]}; do
# and this matches with zkutil::CreateMode
grep -v -e 'src/Common/ZooKeeper/Types.h' -e 'src/Coordination/KeeperConstants.cpp'
} | {
grep -vP $EXCLUDE_DIRS | xargs grep -l -P "extern const $type_of_extern $allowed_chars"
grep -vP $EXCLUDE | xargs grep -l -P "extern const $type_of_extern $allowed_chars"
} | while read file; do
grep -P "extern const $type_of_extern $allowed_chars;" $file | sed -r -e "s/^.*?extern const $type_of_extern ($allowed_chars);.*?$/\1/" | while read val; do
if ! grep -q "$extern_type::$val" $file; then
@ -148,7 +133,7 @@ for extern_type in ${!EXTERN_TYPES[@]}; do
# sed -i -r "0,/(\s*)extern const $type_of_extern [$allowed_chars]+/s//\1extern const $type_of_extern $val;\n&/" $file || \
# awk '{ print; if (ns == 1) { ns = 2 }; if (ns == 2) { ns = 0; print "namespace $extern_type\n{\n extern const $type_of_extern '$val';\n}" } }; /namespace DB/ { ns = 1; };' < $file > ${file}.tmp && mv ${file}.tmp $file )
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' | {
grep -vP $EXCLUDE_DIRS | xargs grep -l -P "$extern_type::$allowed_chars"
grep -vP $EXCLUDE | xargs grep -l -P "$extern_type::$allowed_chars"
} | while read file; do
grep -P "$extern_type::$allowed_chars" $file | grep -P -v '^\s*//' | sed -r -e "s/^.*?$extern_type::($allowed_chars).*?$/\1/" | while read val; do
if ! grep -q "extern const $type_of_extern $val" $file; then
@ -161,7 +146,7 @@ for extern_type in ${!EXTERN_TYPES[@]}; do
# Duplicates
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' | {
grep -vP $EXCLUDE_DIRS | xargs grep -l -P "$extern_type::$allowed_chars"
grep -vP $EXCLUDE | xargs grep -l -P "$extern_type::$allowed_chars"
} | while read file; do
grep -P "extern const $type_of_extern $allowed_chars;" $file | sort | uniq -c | grep -v -P ' +1 ' && echo "Duplicate $extern_type in file $file"
done
@ -169,32 +154,32 @@ done
# Three or more consecutive empty lines
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
while read file; do awk '/^$/ { ++i; if (i > 2) { print "More than two consecutive empty lines in file '$file'" } } /./ { i = 0 }' $file; done
# Check that every header file has #pragma once in first line
find $ROOT_PATH/{src,programs,utils} -name '*.h' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
while read file; do [[ $(head -n1 $file) != '#pragma once' ]] && echo "File $file must have '#pragma once' in first line"; done
# Too many exclamation marks
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -F '!!!' | grep -P '.' && echo "Too many exclamation marks (looks dirty, unconfident)."
# Exclamation mark in a message
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -F '!",' | grep -P '.' && echo "No need for an exclamation mark (looks dirty, unconfident)."
# Trailing whitespaces
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -n -P ' $' | grep -n -P '.' && echo "^ Trailing whitespaces."
# Forbid stringstream because it's easy to use them incorrectly and hard to debug possible issues
find $ROOT_PATH/{src,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -P 'std::[io]?stringstream' | grep -v "STYLE_CHECK_ALLOW_STD_STRING_STREAM" && echo "Use WriteBufferFromOwnString or ReadBufferFromString instead of std::stringstream"
# Forbid std::cerr/std::cout in src (fine in programs/utils)
@ -204,6 +189,7 @@ std_cerr_cout_excludes=(
_fuzzer
# OK
src/Common/ProgressIndication.cpp
src/Common/ProgressTable.cpp
# only under #ifdef DBMS_HASH_MAP_DEBUG_RESIZES, that is used only in tests
src/Common/HashTable/HashTable.h
# SensitiveDataMasker::printStats()
@ -230,11 +216,10 @@ std_cerr_cout_excludes=(
)
sources_with_std_cerr_cout=( $(
find $ROOT_PATH/{src,base} -name '*.h' -or -name '*.cpp' | \
grep -vP $EXCLUDE_DIRS | \
grep -vP $EXCLUDE | \
grep -F -v $(printf -- "-e %s " "${std_cerr_cout_excludes[@]}") | \
xargs grep -F --with-filename -e std::cerr -e std::cout | cut -d: -f1 | sort -u
) )
# Exclude comments
for src in "${sources_with_std_cerr_cout[@]}"; do
# suppress stderr, since it may contain warning for #pargma once in headers
@ -279,23 +264,23 @@ fi
# Forbid std::filesystem::is_symlink and std::filesystem::read_symlink, because it's easy to use them incorrectly
find $ROOT_PATH/{src,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -P '::(is|read)_symlink' | grep -v "STYLE_CHECK_ALLOW_STD_FS_SYMLINK" && echo "Use DB::FS::isSymlink and DB::FS::readSymlink instead"
# Forbid __builtin_unreachable(), because it's hard to debug when it becomes reachable
find $ROOT_PATH/{src,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -P '__builtin_unreachable' && echo "Use UNREACHABLE() from defines.h instead"
# Forbid mt19937() and random_device() which are outdated and slow
find $ROOT_PATH/{src,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -P '(std::mt19937|std::mersenne_twister_engine|std::random_device)' && echo "Use pcg64_fast (from pcg_random.h) and randomSeed (from Common/randomSeed.h) instead"
# Require checking return value of close(),
# since it can hide fd misuse and break other places.
find $ROOT_PATH/{src,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -e ' close(.*fd' -e ' ::close(' | grep -v = && echo "Return value of close() should be checked"
# A small typo can lead to debug code in release builds, see https://github.com/ClickHouse/ClickHouse/pull/47647
@ -322,18 +307,15 @@ ls -1d $ROOT_PATH/contrib/*-cmake | xargs -I@ find @ -name 'CMakeLists.txt' -or
# Wrong spelling of abbreviations, e.g. SQL is right, Sql is wrong. XMLHttpRequest is very wrong.
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -P 'Sql|Html|Xml|Cpu|Tcp|Udp|Http|Db|Json|Yaml' | grep -v -P 'RabbitMQ|Azure|Aws|aws|Avro|IO/S3' &&
echo "Abbreviations such as SQL, XML, HTTP, should be in all caps. For example, SQL is right, Sql is wrong. XMLHttpRequest is very wrong."
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' |
grep -vP $EXCLUDE_DIRS |
grep -vP $EXCLUDE |
xargs grep -F -i 'ErrorCodes::LOGICAL_ERROR, "Logical error:' &&
echo "If an exception has LOGICAL_ERROR code, there is no need to include the text 'Logical error' in the exception message, because then the phrase 'Logical error' will be printed twice."
# There shouldn't be any code snippets under GPL or LGPL
find $ROOT_PATH/{src,base,programs} -name '*.h' -or -name '*.cpp' 2>/dev/null | xargs grep -i -F 'General Public License' && echo "There shouldn't be any code snippets under GPL or LGPL"
PATTERN="allow_";
DIFF=$(comm -3 <(grep -o "\b$PATTERN\w*\b" $ROOT_PATH/src/Core/Settings.cpp | sort -u) <(grep -o -h "\b$PATTERN\w*\b" $ROOT_PATH/src/Databases/enableAllExperimentalSettings.cpp $ROOT_PATH/utils/check-style/experimental_settings_ignore.txt | sort -u));
[ -n "$DIFF" ] && echo "$DIFF" && echo "^^ Detected 'allow_*' settings that might need to be included in src/Databases/enableAllExperimentalSettings.cpp" && echo "Alternatively, consider adding an exception to utils/check-style/experimental_settings_ignore.txt"

0
ci_v2/jobs/scripts/check_style/check_submodules.sh → ci/jobs/scripts/check_style/check_submodules.sh
View File

0
ci_v2/jobs/scripts/check_style/check_typos.sh → ci/jobs/scripts/check_style/check_typos.sh
View File

0
ci_v2/jobs/scripts/check_style/checks_to_refactor.sh → ci/jobs/scripts/check_style/checks_to_refactor.sh
View File

0
ci_v2/jobs/scripts/check_style/double_whitespaces.pl → ci/jobs/scripts/check_style/double_whitespaces.pl
View File

284
ci/jobs/scripts/functional_tests_results.py Executable file
View File

@ -0,0 +1,284 @@
import dataclasses
from typing import List
from praktika.environment import Environment
from praktika.result import Result
OK_SIGN = "[ OK "
FAIL_SIGN = "[ FAIL "
TIMEOUT_SIGN = "[ Timeout! "
UNKNOWN_SIGN = "[ UNKNOWN "
SKIPPED_SIGN = "[ SKIPPED "
HUNG_SIGN = "Found hung queries in processlist"
SERVER_DIED_SIGN = "Server died, terminating all processes"
SERVER_DIED_SIGN2 = "Server does not respond to health check"
DATABASE_SIGN = "Database: "
SUCCESS_FINISH_SIGNS = ["All tests have finished", "No tests were run"]
RETRIES_SIGN = "Some tests were restarted"
# def write_results(results_file, status_file, results, status):
# with open(results_file, "w", encoding="utf-8") as f:
# out = csv.writer(f, delimiter="\t")
# out.writerows(results)
# with open(status_file, "w", encoding="utf-8") as f:
# out = csv.writer(f, delimiter="\t")
# out.writerow(status)
BROKEN_TESTS_ANALYZER_TECH_DEBT = [
"01624_soft_constraints",
# Check after ConstantNode refactoring
"02944_variant_as_common_type",
]
class FTResultsProcessor:
@dataclasses.dataclass
class Summary:
total: int
skipped: int
unknown: int
failed: int
success: int
test_results: List[Result]
hung: bool = False
server_died: bool = False
retries: bool = False
success_finish: bool = False
test_end: bool = True
def __init__(self, wd):
self.tests_output_file = f"{wd}/test_result.txt"
# self.test_results_parsed_file = f"{wd}/test_result.tsv"
# self.status_file = f"{wd}/check_status.tsv"
self.broken_tests = BROKEN_TESTS_ANALYZER_TECH_DEBT
def _process_test_output(self):
total = 0
skipped = 0
unknown = 0
failed = 0
success = 0
hung = False
server_died = False
retries = False
success_finish = False
test_results = []
test_end = True
with open(self.tests_output_file, "r", encoding="utf-8") as test_file:
for line in test_file:
original_line = line
line = line.strip()
if any(s in line for s in SUCCESS_FINISH_SIGNS):
success_finish = True
# Ignore hung check report, since it may be quite large.
# (and may break python parser which has limit of 128KiB for each row).
if HUNG_SIGN in line:
hung = True
break
if SERVER_DIED_SIGN in line or SERVER_DIED_SIGN2 in line:
server_died = True
if RETRIES_SIGN in line:
retries = True
if any(
sign in line
for sign in (OK_SIGN, FAIL_SIGN, UNKNOWN_SIGN, SKIPPED_SIGN)
):
test_name = line.split(" ")[2].split(":")[0]
test_time = ""
try:
time_token = line.split("]")[1].strip().split()[0]
float(time_token)
test_time = time_token
except:
pass
total += 1
if TIMEOUT_SIGN in line:
if test_name in self.broken_tests:
success += 1
test_results.append((test_name, "BROKEN", test_time, []))
else:
failed += 1
test_results.append((test_name, "Timeout", test_time, []))
elif FAIL_SIGN in line:
if test_name in self.broken_tests:
success += 1
test_results.append((test_name, "BROKEN", test_time, []))
else:
failed += 1
test_results.append((test_name, "FAIL", test_time, []))
elif UNKNOWN_SIGN in line:
unknown += 1
test_results.append((test_name, "FAIL", test_time, []))
elif SKIPPED_SIGN in line:
skipped += 1
test_results.append((test_name, "SKIPPED", test_time, []))
else:
if OK_SIGN in line and test_name in self.broken_tests:
skipped += 1
test_results.append(
(
test_name,
"NOT_FAILED",
test_time,
[
"This test passed. Update analyzer_tech_debt.txt.\n"
],
)
)
else:
success += int(OK_SIGN in line)
test_results.append((test_name, "OK", test_time, []))
test_end = False
elif (
len(test_results) > 0
and test_results[-1][1] == "FAIL"
and not test_end
):
test_results[-1][3].append(original_line)
# Database printed after everything else in case of failures,
# so this is a stop marker for capturing test output.
#
# And it is handled after everything else to include line with database into the report.
if DATABASE_SIGN in line:
test_end = True
test_results = [
Result(
name=test[0],
status=test[1],
start_time=None,
duration=float(test[2]),
info="".join(test[3])[:8192],
)
for test in test_results
]
s = self.Summary(
total=total,
skipped=skipped,
unknown=unknown,
failed=failed,
success=success,
test_results=test_results,
hung=hung,
server_died=server_died,
success_finish=success_finish,
retries=retries,
)
return s
def run(self):
state = Result.Status.SUCCESS
s = self._process_test_output()
test_results = s.test_results
# # Check test_results.tsv for sanitizer asserts, crashes and other critical errors.
# # If the file is present, it's expected to be generated by stress_test.lib check for critical errors
# # In the end this file will be fully regenerated, including both results from critical errors check and
# # functional test results.
# if test_results_path and os.path.exists(test_results_path):
# with open(test_results_path, "r", encoding="utf-8") as test_results_file:
# existing_test_results = list(
# csv.reader(test_results_file, delimiter="\t")
# )
# for test in existing_test_results:
# if len(test) < 2:
# unknown += 1
# else:
# test_results.append(test)
#
# if test[1] != "OK":
# failed += 1
# else:
# success += 1
# is_flaky_check = 1 < int(os.environ.get("NUM_TRIES", 1))
# logging.info("Is flaky check: %s", is_flaky_check)
# # If no tests were run (success == 0) it indicates an error (e.g. server did not start or crashed immediately)
# # But it's Ok for "flaky checks" - they can contain just one test for check which is marked as skipped.
# if failed != 0 or unknown != 0 or (success == 0 and (not is_flaky_check)):
if s.failed != 0 or s.unknown != 0:
state = Result.Status.FAILED
if s.hung:
state = Result.Status.FAILED
test_results.append(
Result("Some queries hung", "FAIL", info="Some queries hung")
)
elif s.server_died:
state = Result.Status.FAILED
# When ClickHouse server crashes, some tests are still running
# and fail because they cannot connect to server
for result in test_results:
if result.status == "FAIL":
result.status = "SERVER_DIED"
test_results.append(Result("Server died", "FAIL", info="Server died"))
elif not s.success_finish:
state = Result.Status.FAILED
test_results.append(
Result("Tests are not finished", "FAIL", info="Tests are not finished")
)
elif s.retries:
test_results.append(
Result("Some tests restarted", "SKIPPED", info="Some tests restarted")
)
else:
pass
# TODO: !!!
# def test_result_comparator(item):
# # sort by status then by check name
# order = {
# "FAIL": 0,
# "SERVER_DIED": 1,
# "Timeout": 2,
# "NOT_FAILED": 3,
# "BROKEN": 4,
# "OK": 5,
# "SKIPPED": 6,
# }
# return order.get(item[1], 10), str(item[0]), item[1]
#
# test_results.sort(key=test_result_comparator)
return Result.create_from(
name=Environment.JOB_NAME,
results=test_results,
status=state,
files=[self.tests_output_file],
with_info_from_results=False,
)
# if __name__ == "__main__":
# logging.basicConfig(level=logging.INFO, format="%(asctime)s %(message)s")
# parser = argparse.ArgumentParser(
# description="ClickHouse script for parsing results of functional tests"
# )
#
# parser.add_argument("--out-results-file", default="/test_output/test_results.tsv")
# parser.add_argument("--out-status-file", default="/test_output/check_status.tsv")
# args = parser.parse_args()
#
# broken_tests = []
# state, description, test_results = process_result(
# args.in_results_dir,
# broken_tests,
# args.in_test_result_file,
# args.in_results_file,
# )
# logging.info("Result parsed")
# status = (state, description)
#
#
#
# write_results(args.out_results_file, args.out_status_file, test_results, status)
# logging.info("Result written")

5
ci/praktika/__init__.py Normal file
View File

@ -0,0 +1,5 @@
from .artifact import Artifact
from .docker import Docker
from .job import Job
from .secret import Secret
from .workflow import Workflow

94
ci/praktika/__main__.py Normal file
View File

@ -0,0 +1,94 @@
import argparse
import sys
from praktika.html_prepare import Html
from praktika.utils import Utils
from praktika.validator import Validator
from praktika.yaml_generator import YamlGenerator
def create_parser():
parser = argparse.ArgumentParser(prog="python3 -m praktika")
subparsers = parser.add_subparsers(dest="command", help="Available subcommands")
run_parser = subparsers.add_parser("run", help="Job Runner")
run_parser.add_argument("--job", help="Job Name", type=str, required=True)
run_parser.add_argument(
"--workflow",
help="Workflow Name (required if job name is not uniq per config)",
type=str,
default="",
)
run_parser.add_argument(
"--no-docker",
help="Do not run job in docker even if job config says so, for local test",
action="store_true",
)
run_parser.add_argument(
"--docker",
help="Custom docker image for job run, for local test",
type=str,
default="",
)
run_parser.add_argument(
"--param",
help="Custom parameter to pass into a job script, it's up to job script how to use it, for local test",
type=str,
default=None,
)
run_parser.add_argument(
"--ci",
help="When not set - dummy env will be generated, for local test",
action="store_true",
default="",
)
_yaml_parser = subparsers.add_parser("yaml", help="Generates Yaml Workflows")
_html_parser = subparsers.add_parser("html", help="Uploads HTML page for reports")
return parser
if __name__ == "__main__":
parser = create_parser()
args = parser.parse_args()
if args.command == "yaml":
Validator().validate()
YamlGenerator().generate()
elif args.command == "html":
Html.prepare()
elif args.command == "run":
from praktika.mangle import _get_workflows
from praktika.runner import Runner
workflows = _get_workflows(name=args.workflow or None)
job_workflow_pairs = []
for workflow in workflows:
job = workflow.find_job(args.job, lazy=True)
if job:
job_workflow_pairs.append((job, workflow))
if not job_workflow_pairs:
Utils.raise_with_error(
f"Failed to find job [{args.job}] workflow [{args.workflow}]"
)
elif len(job_workflow_pairs) > 1:
Utils.raise_with_error(
f"More than one job [{args.job}] found - try specifying workflow name with --workflow"
)
else:
job, workflow = job_workflow_pairs[0][0], job_workflow_pairs[0][1]
print(f"Going to run job [{job.name}], workflow [{workflow.name}]")
Runner().run(
workflow=workflow,
job=job,
docker=args.docker,
dummy_env=not args.ci,
no_docker=args.no_docker,
param=args.param,
)
else:
parser.print_help()
sys.exit(1)

198
ci/praktika/_environment.py Normal file
View File

@ -0,0 +1,198 @@
import dataclasses
import json
import os
from pathlib import Path
from types import SimpleNamespace
from typing import Any, Dict, List, Type
from praktika import Workflow
from praktika._settings import _Settings
from praktika.utils import MetaClasses, T
@dataclasses.dataclass
class _Environment(MetaClasses.Serializable):
WORKFLOW_NAME: str
JOB_NAME: str
REPOSITORY: str
BRANCH: str
SHA: str
PR_NUMBER: int
EVENT_TYPE: str
JOB_OUTPUT_STREAM: str
EVENT_FILE_PATH: str
CHANGE_URL: str
COMMIT_URL: str
BASE_BRANCH: str
RUN_ID: str
RUN_URL: str
INSTANCE_TYPE: str
INSTANCE_ID: str
INSTANCE_LIFE_CYCLE: str
LOCAL_RUN: bool = False
PARAMETER: Any = None
REPORT_INFO: List[str] = dataclasses.field(default_factory=list)
name = "environment"
@classmethod
def file_name_static(cls, _name=""):
return f"{_Settings.TEMP_DIR}/{cls.name}.json"
@classmethod
def from_dict(cls: Type[T], obj: Dict[str, Any]) -> T:
JOB_OUTPUT_STREAM = os.getenv("GITHUB_OUTPUT", "")
obj["JOB_OUTPUT_STREAM"] = JOB_OUTPUT_STREAM
if "PARAMETER" in obj:
obj["PARAMETER"] = _to_object(obj["PARAMETER"])
return cls(**obj)
def add_info(self, info):
self.REPORT_INFO.append(info)
self.dump()
@classmethod
def get(cls):
if Path(cls.file_name_static()).is_file():
return cls.from_fs("environment")
else:
print("WARNING: Environment: get from env")
env = cls.from_env()
env.dump()
return env
def set_job_name(self, job_name):
self.JOB_NAME = job_name
self.dump()
return self
@staticmethod
def get_needs_statuses():
if Path(_Settings.WORKFLOW_STATUS_FILE).is_file():
with open(_Settings.WORKFLOW_STATUS_FILE, "r", encoding="utf8") as f:
return json.load(f)
else:
print(
f"ERROR: Status file [{_Settings.WORKFLOW_STATUS_FILE}] does not exist"
)
raise RuntimeError()
@classmethod
def from_env(cls) -> "_Environment":
WORKFLOW_NAME = os.getenv("GITHUB_WORKFLOW", "")
JOB_NAME = os.getenv("JOB_NAME", "")
REPOSITORY = os.getenv("GITHUB_REPOSITORY", "")
BRANCH = os.getenv("GITHUB_HEAD_REF", "")
EVENT_FILE_PATH = os.getenv("GITHUB_EVENT_PATH", "")
JOB_OUTPUT_STREAM = os.getenv("GITHUB_OUTPUT", "")
RUN_ID = os.getenv("GITHUB_RUN_ID", "0")
RUN_URL = f"https://github.com/{REPOSITORY}/actions/runs/{RUN_ID}"
BASE_BRANCH = os.getenv("GITHUB_BASE_REF", "")
if EVENT_FILE_PATH:
with open(EVENT_FILE_PATH, "r", encoding="utf-8") as f:
github_event = json.load(f)
if "pull_request" in github_event:
EVENT_TYPE = Workflow.Event.PULL_REQUEST
PR_NUMBER = github_event["pull_request"]["number"]
SHA = github_event["pull_request"]["head"]["sha"]
CHANGE_URL = github_event["pull_request"]["html_url"]
COMMIT_URL = CHANGE_URL + f"/commits/{SHA}"
elif "commits" in github_event:
EVENT_TYPE = Workflow.Event.PUSH
SHA = github_event["after"]
CHANGE_URL = github_event["head_commit"]["url"] # commit url
PR_NUMBER = 0
COMMIT_URL = CHANGE_URL
else:
assert False, "TODO: not supported"
else:
print("WARNING: Local execution - dummy Environment will be generated")
SHA = "TEST"
PR_NUMBER = -1
EVENT_TYPE = Workflow.Event.PUSH
CHANGE_URL = ""
COMMIT_URL = ""
INSTANCE_TYPE = (
os.getenv("INSTANCE_TYPE", None)
# or Shell.get_output("ec2metadata --instance-type")
or ""
)
INSTANCE_ID = (
os.getenv("INSTANCE_ID", None)
# or Shell.get_output("ec2metadata --instance-id")
or ""
)
INSTANCE_LIFE_CYCLE = (
os.getenv("INSTANCE_LIFE_CYCLE", None)
# or Shell.get_output(
# "curl -s --fail http://169.254.169.254/latest/meta-data/instance-life-cycle"
# )
or ""
)
return _Environment(
WORKFLOW_NAME=WORKFLOW_NAME,
JOB_NAME=JOB_NAME,
REPOSITORY=REPOSITORY,
BRANCH=BRANCH,
EVENT_FILE_PATH=EVENT_FILE_PATH,
JOB_OUTPUT_STREAM=JOB_OUTPUT_STREAM,
SHA=SHA,
EVENT_TYPE=EVENT_TYPE,
PR_NUMBER=PR_NUMBER,
RUN_ID=RUN_ID,
CHANGE_URL=CHANGE_URL,
COMMIT_URL=COMMIT_URL,
RUN_URL=RUN_URL,
BASE_BRANCH=BASE_BRANCH,
INSTANCE_TYPE=INSTANCE_TYPE,
INSTANCE_ID=INSTANCE_ID,
INSTANCE_LIFE_CYCLE=INSTANCE_LIFE_CYCLE,
REPORT_INFO=[],
)
def get_s3_prefix(self, latest=False):
return self.get_s3_prefix_static(self.PR_NUMBER, self.BRANCH, self.SHA, latest)
@classmethod
def get_s3_prefix_static(cls, pr_number, branch, sha, latest=False):
prefix = ""
if pr_number > 0:
prefix += f"{pr_number}"
else:
prefix += f"{branch}"
if latest:
prefix += f"/latest"
elif sha:
prefix += f"/{sha}"
return prefix
# TODO: find a better place for the function. This file should not import praktika.settings
# as it's requires reading users config, that's why imports nested inside the function
def get_report_url(self):
import urllib
from praktika.settings import Settings
from praktika.utils import Utils
path = Settings.HTML_S3_PATH
for bucket, endpoint in Settings.S3_BUCKET_TO_HTTP_ENDPOINT.items():
if bucket in path:
path = path.replace(bucket, endpoint)
break
REPORT_URL = f"https://{path}/{Path(Settings.HTML_PAGE_FILE).name}?PR={self.PR_NUMBER}&sha={self.SHA}&name_0={urllib.parse.quote(self.WORKFLOW_NAME, safe='')}&name_1={urllib.parse.quote(self.JOB_NAME, safe='')}"
return REPORT_URL
def is_local_run(self):
return self.LOCAL_RUN
def _to_object(data):
if isinstance(data, dict):
return SimpleNamespace(**{k: _to_object(v) for k, v in data.items()})
elif isinstance(data, list):
return [_to_object(i) for i in data]
else:
return data

124
ci/praktika/_settings.py Normal file
View File

@ -0,0 +1,124 @@
import dataclasses
from pathlib import Path
from typing import Dict, Iterable, List, Optional
@dataclasses.dataclass
class _Settings:
######################################
# Pipeline generation settings #
######################################
CI_PATH = "./ci"
WORKFLOW_PATH_PREFIX: str = "./.github/workflows"
WORKFLOWS_DIRECTORY: str = f"{CI_PATH}/workflows"
SETTINGS_DIRECTORY: str = f"{CI_PATH}/settings"
CI_CONFIG_JOB_NAME = "Config Workflow"
DOCKER_BUILD_JOB_NAME = "Docker Builds"
FINISH_WORKFLOW_JOB_NAME = "Finish Workflow"
READY_FOR_MERGE_STATUS_NAME = "Ready for Merge"
CI_CONFIG_RUNS_ON: Optional[List[str]] = None
DOCKER_BUILD_RUNS_ON: Optional[List[str]] = None
VALIDATE_FILE_PATHS: bool = True
######################################
# Runtime Settings #
######################################
MAX_RETRIES_S3 = 3
MAX_RETRIES_GH = 3
######################################
# S3 (artifact storage) settings #
######################################
S3_ARTIFACT_PATH: str = ""
######################################
# CI workspace settings #
######################################
TEMP_DIR: str = "/tmp/praktika"
OUTPUT_DIR: str = f"{TEMP_DIR}/output"
INPUT_DIR: str = f"{TEMP_DIR}/input"
PYTHON_INTERPRETER: str = "python3"
PYTHON_PACKET_MANAGER: str = "pip3"
PYTHON_VERSION: str = "3.9"
INSTALL_PYTHON_FOR_NATIVE_JOBS: bool = False
INSTALL_PYTHON_REQS_FOR_NATIVE_JOBS: str = "./ci/requirements.txt"
ENVIRONMENT_VAR_FILE: str = f"{TEMP_DIR}/environment.json"
RUN_LOG: str = f"{TEMP_DIR}/praktika_run.log"
SECRET_GH_APP_ID: str = "GH_APP_ID"
SECRET_GH_APP_PEM_KEY: str = "GH_APP_PEM_KEY"
ENV_SETUP_SCRIPT: str = "/tmp/praktika_setup_env.sh"
WORKFLOW_STATUS_FILE: str = f"{TEMP_DIR}/workflow_status.json"
######################################
# CI Cache settings #
######################################
CACHE_VERSION: int = 1
CACHE_DIGEST_LEN: int = 20
CACHE_S3_PATH: str = ""
CACHE_LOCAL_PATH: str = f"{TEMP_DIR}/ci_cache"
######################################
# Report settings #
######################################
HTML_S3_PATH: str = ""
HTML_PAGE_FILE: str = "./praktika/json.html"
TEXT_CONTENT_EXTENSIONS: Iterable[str] = frozenset([".txt", ".log"])
S3_BUCKET_TO_HTTP_ENDPOINT: Optional[Dict[str, str]] = None
DOCKERHUB_USERNAME: str = ""
DOCKERHUB_SECRET: str = ""
DOCKER_WD: str = "/wd"
######################################
# CI DB Settings #
######################################
SECRET_CI_DB_URL: str = "CI_DB_URL"
SECRET_CI_DB_PASSWORD: str = "CI_DB_PASSWORD"
CI_DB_DB_NAME = ""
CI_DB_TABLE_NAME = ""
CI_DB_INSERT_TIMEOUT_SEC = 5
_USER_DEFINED_SETTINGS = [
"S3_ARTIFACT_PATH",
"CACHE_S3_PATH",
"HTML_S3_PATH",
"S3_BUCKET_TO_HTTP_ENDPOINT",
"TEXT_CONTENT_EXTENSIONS",
"TEMP_DIR",
"OUTPUT_DIR",
"INPUT_DIR",
"CI_CONFIG_RUNS_ON",
"DOCKER_BUILD_RUNS_ON",
"CI_CONFIG_JOB_NAME",
"PYTHON_INTERPRETER",
"PYTHON_VERSION",
"PYTHON_PACKET_MANAGER",
"INSTALL_PYTHON_FOR_NATIVE_JOBS",
"INSTALL_PYTHON_REQS_FOR_NATIVE_JOBS",
"MAX_RETRIES_S3",
"MAX_RETRIES_GH",
"VALIDATE_FILE_PATHS",
"DOCKERHUB_USERNAME",
"DOCKERHUB_SECRET",
"READY_FOR_MERGE_STATUS_NAME",
"SECRET_CI_DB_URL",
"SECRET_CI_DB_PASSWORD",
"CI_DB_DB_NAME",
"CI_DB_TABLE_NAME",
"CI_DB_INSERT_TIMEOUT_SEC",
"SECRET_GH_APP_PEM_KEY",
"SECRET_GH_APP_ID",
]
class GHRunners:
ubuntu = "ubuntu-latest"
if __name__ == "__main__":
for setting in _USER_DEFINED_SETTINGS:
print(_Settings().__getattribute__(setting))
# print(dataclasses.asdict(_Settings()))

33
ci/praktika/artifact.py Normal file
View File

@ -0,0 +1,33 @@
from dataclasses import dataclass
class Artifact:
class Type:
GH = "github"
S3 = "s3"
PHONY = "phony"
@dataclass
class Config:
"""
name - artifact name
type - artifact type, see Artifact.Type
path - file path or glob, e.g. "path/**/[abc]rtifac?/*"
"""
name: str
type: str
path: str
_provided_by: str = ""
_s3_path: str = ""
def is_s3_artifact(self):
return self.type == Artifact.Type.S3
@classmethod
def define_artifact(cls, name, type, path):
return cls.Config(name=name, type=type, path=path)
@classmethod
def define_gh_artifact(cls, name, path):
return cls.define_artifact(name=name, type=cls.Type.GH, path=path)

127
ci/praktika/cache.py Normal file
View File

@ -0,0 +1,127 @@
import dataclasses
import json
from pathlib import Path
from praktika import Artifact, Job, Workflow
from praktika._environment import _Environment
from praktika.digest import Digest
from praktika.s3 import S3
from praktika.settings import Settings
from praktika.utils import Utils
class Cache:
@dataclasses.dataclass
class CacheRecord:
class Type:
SUCCESS = "success"
type: str
sha: str
pr_number: int
branch: str
def dump(self, path):
with open(path, "w", encoding="utf8") as f:
json.dump(dataclasses.asdict(self), f)
@classmethod
def from_fs(cls, path):
with open(path, "r", encoding="utf8") as f:
return Cache.CacheRecord(**json.load(f))
@classmethod
def from_dict(cls, obj):
return Cache.CacheRecord(**obj)
def __init__(self):
self.digest = Digest()
self.success = {} # type Dict[str, Any]
@classmethod
def push_success_record(cls, job_name, job_digest, sha):
type_ = Cache.CacheRecord.Type.SUCCESS
record = Cache.CacheRecord(
type=type_,
sha=sha,
pr_number=_Environment.get().PR_NUMBER,
branch=_Environment.get().BRANCH,
)
assert (
Settings.CACHE_S3_PATH
), f"Setting CACHE_S3_PATH must be defined with enabled CI Cache"
record_path = f"{Settings.CACHE_S3_PATH}/v{Settings.CACHE_VERSION}/{Utils.normalize_string(job_name)}/{job_digest}"
record_file = Path(Settings.TEMP_DIR) / type_
record.dump(record_file)
S3.copy_file_to_s3(s3_path=record_path, local_path=record_file)
record_file.unlink()
def fetch_success(self, job_name, job_digest):
type_ = Cache.CacheRecord.Type.SUCCESS
assert (
Settings.CACHE_S3_PATH
), f"Setting CACHE_S3_PATH must be defined with enabled CI Cache"
record_path = f"{Settings.CACHE_S3_PATH}/v{Settings.CACHE_VERSION}/{Utils.normalize_string(job_name)}/{job_digest}/{type_}"
record_file_local_dir = (
f"{Settings.CACHE_LOCAL_PATH}/{Utils.normalize_string(job_name)}/"
)
Path(record_file_local_dir).mkdir(parents=True, exist_ok=True)
if S3.head_object(record_path):
res = S3.copy_file_from_s3(
s3_path=record_path, local_path=record_file_local_dir
)
else:
res = None
if res:
print(f"Cache record found, job [{job_name}], digest [{job_digest}]")
self.success[job_name] = True
return Cache.CacheRecord.from_fs(Path(record_file_local_dir) / type_)
return None
if __name__ == "__main__":
# test
c = Cache()
workflow = Workflow.Config(
name="TEST",
event=Workflow.Event.PULL_REQUEST,
jobs=[
Job.Config(
name="JobA",
runs_on=["some"],
command="python -m unittest ./ci/tests/example_1/test_example_produce_artifact.py",
provides=["greet"],
job_requirements=Job.Requirements(
python_requirements_txt="./ci/requirements.txt"
),
digest_config=Job.CacheDigestConfig(
# example: use glob to include files
include_paths=["./ci/tests/example_1/test_example_consume*.py"],
),
),
Job.Config(
name="JobB",
runs_on=["some"],
command="python -m unittest ./ci/tests/example_1/test_example_consume_artifact.py",
requires=["greet"],
job_requirements=Job.Requirements(
python_requirements_txt="./ci/requirements.txt"
),
digest_config=Job.CacheDigestConfig(
# example: use dir to include files recursively
include_paths=["./ci/tests/example_1"],
# example: use glob to exclude files from digest
exclude_paths=[
"./ci/tests/example_1/test_example_consume*",
"./**/*.pyc",
],
),
),
],
artifacts=[Artifact.Config(type="s3", name="greet", path="hello")],
enable_cache=True,
)
for job in workflow.jobs:
print(c.digest.calc_job_digest(job))

136
ci/praktika/cidb.py Normal file
View File

@ -0,0 +1,136 @@
import copy
import dataclasses
import json
from typing import Optional
import requests
from praktika._environment import _Environment
from praktika.result import Result
from praktika.settings import Settings
from praktika.utils import Utils
class CIDB:
@dataclasses.dataclass
class TableRecord:
pull_request_number: int
commit_sha: str
commit_url: str
check_name: str
check_status: str
check_duration_ms: int
check_start_time: int
report_url: str
pull_request_url: str
base_ref: str
base_repo: str
head_ref: str
head_repo: str
task_url: str
instance_type: str
instance_id: str
test_name: str
test_status: str
test_duration_ms: Optional[int]
test_context_raw: str
def __init__(self, url, passwd):
self.url = url
self.auth = {
"X-ClickHouse-User": "default",
"X-ClickHouse-Key": passwd,
}
@classmethod
def json_data_generator(cls, result: Result):
env = _Environment.get()
base_record = cls.TableRecord(
pull_request_number=env.PR_NUMBER,
commit_sha=env.SHA,
commit_url=env.COMMIT_URL,
check_name=result.name,
check_status=result.status,
check_duration_ms=int(result.duration * 1000),
check_start_time=Utils.timestamp_to_str(result.start_time),
report_url=env.get_report_url(),
pull_request_url=env.CHANGE_URL,
base_ref=env.BASE_BRANCH,
base_repo=env.REPOSITORY,
head_ref=env.BRANCH,
# TODO: remove from table?
head_repo=env.REPOSITORY,
# TODO: remove from table?
task_url="",
instance_type=",".join([env.INSTANCE_TYPE, env.INSTANCE_LIFE_CYCLE]),
instance_id=env.INSTANCE_ID,
test_name="",
test_status="",
test_duration_ms=None,
test_context_raw=result.info,
)
yield json.dumps(dataclasses.asdict(base_record))
for result_ in result.results:
record = copy.deepcopy(base_record)
record.test_name = result_.name
if result_.start_time:
record.check_start_time = (Utils.timestamp_to_str(result.start_time),)
record.test_status = result_.status
record.test_duration_ms = int(result_.duration * 1000)
record.test_context_raw = result_.info
yield json.dumps(dataclasses.asdict(record))
def insert(self, result: Result):
# Create a session object
params = {
"database": Settings.CI_DB_DB_NAME,
"query": f"INSERT INTO {Settings.CI_DB_TABLE_NAME} FORMAT JSONEachRow",
"date_time_input_format": "best_effort",
"send_logs_level": "warning",
}
session = requests.Session()
for json_str in self.json_data_generator(result):
try:
response1 = session.post(
url=self.url,
params=params,
data=json_str,
headers=self.auth,
timeout=Settings.CI_DB_INSERT_TIMEOUT_SEC,
)
except Exception as ex:
raise ex
session.close()
def check(self):
# Create a session object
params = {
"database": Settings.CI_DB_DB_NAME,
"query": f"SELECT 1",
}
try:
response = requests.post(
url=self.url,
params=params,
data="",
headers=self.auth,
timeout=Settings.CI_DB_INSERT_TIMEOUT_SEC,
)
if not response.ok:
print("ERROR: No connection to CI DB")
return (
False,
f"ERROR: No connection to CI DB [{response.status_code}/{response.reason}]",
)
if not response.json() == 1:
print("ERROR: CI DB smoke test failed select 1 == 1")
return (
False,
f"ERROR: CI DB smoke test failed [select 1 ==> {response.json()}]",
)
except Exception as ex:
print(f"ERROR: Exception [{ex}]")
return False, "CIDB: ERROR: Exception [{ex}]"
return True, ""

112
ci/praktika/digest.py Normal file
View File

@ -0,0 +1,112 @@
import dataclasses
import hashlib
import os
from hashlib import md5
from pathlib import Path
from typing import List
from praktika import Job
from praktika.docker import Docker
from praktika.settings import Settings
from praktika.utils import Utils
class Digest:
def __init__(self):
self.digest_cache = {}
@staticmethod
def _hash_digest_config(digest_config: Job.CacheDigestConfig) -> str:
data_dict = dataclasses.asdict(digest_config)
hash_obj = md5()
hash_obj.update(str(data_dict).encode())
hash_string = hash_obj.hexdigest()
return hash_string
def calc_job_digest(self, job_config: Job.Config):
config = job_config.digest_config
if not config:
return "f" * Settings.CACHE_DIGEST_LEN
cache_key = self._hash_digest_config(config)
if cache_key in self.digest_cache:
return self.digest_cache[cache_key]
included_files = Utils.traverse_paths(
job_config.digest_config.include_paths,
job_config.digest_config.exclude_paths,
sorted=True,
)
print(
f"calc digest for job [{job_config.name}]: hash_key [{cache_key}], include [{len(included_files)}] files"
)
# Sort files to ensure consistent hash calculation
included_files.sort()
# Calculate MD5 hash
res = ""
if not included_files:
res = "f" * Settings.CACHE_DIGEST_LEN
print(f"NOTE: empty digest config [{config}] - return dummy digest")
else:
hash_md5 = hashlib.md5()
for file_path in included_files:
res = self._calc_file_digest(file_path, hash_md5)
assert res
self.digest_cache[cache_key] = res
return res
def calc_docker_digest(
self,
docker_config: Docker.Config,
dependency_configs: List[Docker.Config],
hash_md5=None,
):
"""
:param hash_md5:
:param dependency_configs: list of Docker.Config(s) that :param docker_config: depends on
:param docker_config: Docker.Config to calculate digest for
:return:
"""
print(f"Calculate digest for docker [{docker_config.name}]")
paths = Utils.traverse_path(docker_config.path, sorted=True)
if not hash_md5:
hash_md5 = hashlib.md5()
dependencies = []
for dependency_name in docker_config.depends_on:
for dependency_config in dependency_configs:
if dependency_config.name == dependency_name:
print(
f"Add docker [{dependency_config.name}] as dependency for docker [{docker_config.name}] digest calculation"
)
dependencies.append(dependency_config)
for dependency in dependencies:
_ = self.calc_docker_digest(dependency, dependency_configs, hash_md5)
for path in paths:
_ = self._calc_file_digest(path, hash_md5=hash_md5)
return hash_md5.hexdigest()[: Settings.CACHE_DIGEST_LEN]
@staticmethod
def _calc_file_digest(file_path, hash_md5):
# Resolve file path if it's a symbolic link
resolved_path = file_path
if Path(file_path).is_symlink():
resolved_path = os.path.realpath(file_path)
if not Path(resolved_path).is_file():
print(
f"WARNING: No valid file resolved by link {file_path} -> {resolved_path} - skipping digest calculation"
)
return hash_md5.hexdigest()[: Settings.CACHE_DIGEST_LEN]
with open(resolved_path, "rb") as f:
for chunk in iter(lambda: f.read(4096), b""):
hash_md5.update(chunk)
return hash_md5.hexdigest()[: Settings.CACHE_DIGEST_LEN]

60
ci/praktika/docker.py Normal file
View File

@ -0,0 +1,60 @@
import dataclasses
from typing import List
from praktika.utils import Shell
class Docker:
class Platforms:
ARM = "linux/arm64"
AMD = "linux/amd64"
arm_amd = [ARM, AMD]
@dataclasses.dataclass
class Config:
name: str
path: str
depends_on: List[str]
platforms: List[str]
@classmethod
def build(cls, config: "Docker.Config", log_file, digests, add_latest):
tags_substr = f" -t {config.name}:{digests[config.name]}"
if add_latest:
tags_substr = f" -t {config.name}:latest"
from_tag = ""
if config.depends_on:
assert (
len(config.depends_on) == 1
), f"Only one dependency in depends_on is currently supported, docker [{config}]"
from_tag = f" --build-arg FROM_TAG={digests[config.depends_on[0]]}"
command = f"docker buildx build --platform {','.join(config.platforms)} {tags_substr} {from_tag} --cache-to type=inline --cache-from type=registry,ref={config.name} --push {config.path}"
return Shell.run(command, log_file=log_file, verbose=True)
@classmethod
def sort_in_build_order(cls, dockers: List["Docker.Config"]):
ready_names = []
i = 0
while i < len(dockers):
docker = dockers[i]
if not docker.depends_on or all(
dep in ready_names for dep in docker.depends_on
):
ready_names.append(docker.name)
i += 1
else:
dockers.append(dockers.pop(i))
return dockers
@classmethod
def login(cls, user_name, user_password):
print("Docker: log in to dockerhub")
return Shell.check(
f"docker login --username '{user_name}' --password-stdin",
strict=True,
stdin_str=user_password,
encoding="utf-8",
verbose=True,
)

3
ci/praktika/environment.py Normal file
View File

@ -0,0 +1,3 @@
from praktika._environment import _Environment
Environment = _Environment.get()

0
tests/integration/test_system_replicated_fetches/__init__.py → ci/praktika/execution/__init__.py
View File

4
ci/praktika/execution/__main__.py Normal file
View File

@ -0,0 +1,4 @@
from praktika.execution.machine_init import run
if __name__ == "__main__":
run()

31
ci/praktika/execution/execution_settings.py Normal file
View File

@ -0,0 +1,31 @@
import os
from praktika.utils import MetaClasses
class ScalingType(metaclass=MetaClasses.WithIter):
DISABLED = "disabled"
AUTOMATIC_SCALE_DOWN = "scale_down"
AUTOMATIC_SCALE_UP_DOWN = "scale"
class DefaultExecutionSettings:
GH_ACTIONS_DIRECTORY: str = "/home/ubuntu/gh_actions"
RUNNER_SCALING_TYPE: str = ScalingType.AUTOMATIC_SCALE_UP_DOWN
MAX_WAIT_TIME_BEFORE_SCALE_DOWN_SEC: int = 30
class ExecutionSettings:
GH_ACTIONS_DIRECTORY = os.getenv(
"GH_ACTIONS_DIRECTORY", DefaultExecutionSettings.GH_ACTIONS_DIRECTORY
)
RUNNER_SCALING_TYPE = os.getenv(
"RUNNER_SCALING_TYPE", DefaultExecutionSettings.RUNNER_SCALING_TYPE
)
MAX_WAIT_TIME_BEFORE_SCALE_DOWN_SEC = int(
os.getenv(
"MAX_WAIT_TIME_BEFORE_SCALE_DOWN_SEC",
DefaultExecutionSettings.MAX_WAIT_TIME_BEFORE_SCALE_DOWN_SEC,
)
)
LOCAL_EXECUTION = bool(os.getenv("CLOUD", "0") == "0")

338
ci/praktika/execution/machine_init.py Normal file
View File

@ -0,0 +1,338 @@
import os
import platform
import signal
import time
import traceback
import requests
from praktika.execution.execution_settings import ExecutionSettings, ScalingType
from praktika.utils import ContextManager, Shell
class StateMachine:
class StateNames:
INIT = "init"
WAIT = "wait"
RUN = "run"
def __init__(self):
self.state = self.StateNames.INIT
self.scale_type = ExecutionSettings.RUNNER_SCALING_TYPE
self.machine = Machine(scaling_type=self.scale_type).update_instance_info()
self.state_updated_at = int(time.time())
self.forked = False
def kick(self):
if self.state == self.StateNames.INIT:
self.machine.config_actions().run_actions_async()
print("State Machine: INIT -> WAIT")
self.state = self.StateNames.WAIT
self.state_updated_at = int(time.time())
# TODO: add monitoring
if not self.machine.is_actions_process_healthy():
print(f"ERROR: GH runner process unexpectedly died")
self.machine.self_terminate(decrease_capacity=False)
elif self.state == self.StateNames.WAIT:
res = self.machine.check_job_assigned()
if res:
print("State Machine: WAIT -> RUN")
self.state = self.StateNames.RUN
self.state_updated_at = int(time.time())
self.check_scale_up()
else:
self.check_scale_down()
elif self.state == self.StateNames.RUN:
res = self.machine.check_job_running()
if res:
pass
else:
print("State Machine: RUN -> INIT")
self.state = self.StateNames.INIT
self.state_updated_at = int(time.time())
def check_scale_down(self):
if self.scale_type not in (
ScalingType.AUTOMATIC_SCALE_DOWN,
ScalingType.AUTOMATIC_SCALE_UP_DOWN,
):
return
if ScalingType.AUTOMATIC_SCALE_UP_DOWN and not self.forked:
print(
f"Scaling type is AUTOMATIC_SCALE_UP_DOWN and machine has not run a job - do not scale down"
)
return
if (
int(time.time()) - self.state_updated_at
> ExecutionSettings.MAX_WAIT_TIME_BEFORE_SCALE_DOWN_SEC
):
print(
f"No job assigned for more than MAX_WAIT_TIME_BEFORE_SCALE_DOWN_SEC [{ExecutionSettings.MAX_WAIT_TIME_BEFORE_SCALE_DOWN_SEC}] - scale down the instance"
)
if not ExecutionSettings.LOCAL_EXECUTION:
self.machine.self_terminate(decrease_capacity=True)
else:
print("Local execution - skip scaling operation")
def check_scale_up(self):
if self.scale_type not in (ScalingType.AUTOMATIC_SCALE_UP_DOWN,):
return
if self.forked:
print("This instance already forked once - do not scale up")
return
self.machine.self_fork()
self.forked = True
def run(self):
self.machine.unconfig_actions()
while True:
self.kick()
time.sleep(5)
def terminate(self):
try:
self.machine.unconfig_actions()
except:
print("WARNING: failed to unconfig runner")
if not ExecutionSettings.LOCAL_EXECUTION:
if self.machine is not None:
self.machine.self_terminate(decrease_capacity=False)
time.sleep(10)
# wait termination
print("ERROR: failed to terminate instance via aws cli - try os call")
os.system("sudo shutdown now")
else:
print("NOTE: Local execution - machine won't be terminated")
class Machine:
@staticmethod
def get_latest_gh_actions_release():
url = f"https://api.github.com/repos/actions/runner/releases/latest"
response = requests.get(url, timeout=5)
if response.status_code == 200:
latest_release = response.json()
return latest_release["tag_name"].removeprefix("v")
else:
print(f"Failed to get the latest release: {response.status_code}")
return None
def __init__(self, scaling_type):
self.os_name = platform.system().lower()
assert self.os_name == "linux", f"Unsupported OS [{self.os_name}]"
if platform.machine() == "x86_64":
self.arch = "x64"
elif "aarch64" in platform.machine().lower():
self.arch = "arm64"
else:
assert False, f"Unsupported arch [{platform.machine()}]"
self.instance_id = None
self.asg_name = None
self.runner_api_endpoint = None
self.runner_type = None
self.labels = []
self.proc = None
assert scaling_type in ScalingType
self.scaling_type = scaling_type
def install_gh_actions_runner(self):
gh_actions_version = self.get_latest_gh_actions_release()
assert self.os_name and gh_actions_version and self.arch
Shell.check(
f"rm -rf {ExecutionSettings.GH_ACTIONS_DIRECTORY}",
strict=True,
verbose=True,
)
Shell.check(
f"mkdir {ExecutionSettings.GH_ACTIONS_DIRECTORY}", strict=True, verbose=True
)
with ContextManager.cd(ExecutionSettings.GH_ACTIONS_DIRECTORY):
Shell.check(
f"curl -O -L https://github.com/actions/runner/releases/download/v{gh_actions_version}/actions-runner-{self.os_name}-{self.arch}-{gh_actions_version}.tar.gz",
strict=True,
verbose=True,
)
Shell.check(f"tar xzf *tar.gz", strict=True, verbose=True)
Shell.check(f"rm -f *tar.gz", strict=True, verbose=True)
Shell.check(f"sudo ./bin/installdependencies.sh", strict=True, verbose=True)
Shell.check(
f"chown -R ubuntu:ubuntu {ExecutionSettings.GH_ACTIONS_DIRECTORY}",
strict=True,
verbose=True,
)
def _get_gh_token_from_ssm(self):
gh_token = Shell.get_output_or_raise(
"/usr/local/bin/aws ssm get-parameter --name github_runner_registration_token --with-decryption --output text --query Parameter.Value"
)
return gh_token
def update_instance_info(self):
self.instance_id = Shell.get_output_or_raise("ec2metadata --instance-id")
assert self.instance_id
self.asg_name = Shell.get_output(
f"aws ec2 describe-instances --instance-id {self.instance_id} --query \"Reservations[].Instances[].Tags[?Key=='aws:autoscaling:groupName'].Value\" --output text"
)
# self.runner_type = Shell.get_output_or_raise(
# f'/usr/local/bin/aws ec2 describe-tags --filters "Name=resource-id,Values={self.instance_id}" --query "Tags[?Key==\'github:runner-type\'].Value" --output text'
# )
self.runner_type = self.asg_name
if (
self.scaling_type != ScalingType.DISABLED
and not ExecutionSettings.LOCAL_EXECUTION
):
assert (
self.asg_name and self.runner_type
), f"Failed to retrieve ASG name, which is required for scaling_type [{self.scaling_type}]"
org = os.getenv("MY_ORG", "")
assert (
org
), "MY_ORG env variable myst be set to use init script for runner machine"
self.runner_api_endpoint = f"https://github.com/{org}"
self.labels = ["self-hosted", self.runner_type]
return self
@classmethod
def check_job_assigned(cls):
runner_pid = Shell.get_output_or_raise("pgrep Runner.Listener")
if not runner_pid:
print("check_job_assigned: No runner pid")
return False
log_file = Shell.get_output_or_raise(
f"lsof -p {runner_pid} | grep -o {ExecutionSettings.GH_ACTIONS_DIRECTORY}/_diag/Runner.*log"
)
if not log_file:
print("check_job_assigned: No log file")
return False
return Shell.check(f"grep -q 'Terminal] .* Running job:' {log_file}")
def check_job_running(self):
if self.proc is None:
print(f"WARNING: No job started")
return False
exit_code = self.proc.poll()
if exit_code is None:
return True
else:
print(f"Job runner finished with exit code [{exit_code}]")
self.proc = None
return False
def config_actions(self):
if not self.instance_id:
self.update_instance_info()
token = self._get_gh_token_from_ssm()
assert token and self.instance_id and self.runner_api_endpoint and self.labels
command = f"sudo -u ubuntu {ExecutionSettings.GH_ACTIONS_DIRECTORY}/config.sh --token {token} \
--url {self.runner_api_endpoint} --ephemeral --unattended --replace \
--runnergroup Default --labels {','.join(self.labels)} --work wd --name {self.instance_id}"
res = 1
i = 0
while i < 10 and res != 0:
res = Shell.run(command)
i += 1
if res != 0:
print(
f"ERROR: failed to configure GH actions runner after [{i}] attempts, exit code [{res}], retry after 10s"
)
time.sleep(10)
self._get_gh_token_from_ssm()
if res == 0:
print("GH action runner has been configured")
else:
assert False, "GH actions runner configuration failed"
return self
def unconfig_actions(self):
token = self._get_gh_token_from_ssm()
command = f"sudo -u ubuntu {ExecutionSettings.GH_ACTIONS_DIRECTORY}/config.sh remove --token {token}"
Shell.check(command, strict=True)
return self
def run_actions_async(self):
command = f"sudo -u ubuntu {ExecutionSettings.GH_ACTIONS_DIRECTORY}/run.sh"
self.proc = Shell.run_async(command)
assert self.proc is not None
return self
def is_actions_process_healthy(self):
try:
if self.proc.poll() is None:
return True
stdout, stderr = self.proc.communicate()
if self.proc.returncode != 0:
# Handle failure
print(
f"GH Action process failed with return code {self.proc.returncode}"
)
print(f"Error output: {stderr}")
return False
else:
print(f"GH Action process is not running")
return False
except Exception as e:
print(f"GH Action process exception: {e}")
return False
def self_terminate(self, decrease_capacity):
print(
f"WARNING: Self terminate is called, decrease_capacity [{decrease_capacity}]"
)
traceback.print_stack()
if not self.instance_id:
self.update_instance_info()
assert self.instance_id
command = f"aws autoscaling terminate-instance-in-auto-scaling-group --instance-id {self.instance_id}"
if decrease_capacity:
command += " --should-decrement-desired-capacity"
else:
command += " --no-should-decrement-desired-capacity"
Shell.check(
command=command,
verbose=True,
)
def self_fork(self):
current_capacity = Shell.get_output(
f'aws autoscaling describe-auto-scaling-groups --auto-scaling-group-name {self.asg_name} \
--query "AutoScalingGroups[0].DesiredCapacity" --output text'
)
current_capacity = int(current_capacity)
if not current_capacity:
print("ERROR: failed to get current capacity - cannot scale up")
return
desired_capacity = current_capacity + 1
command = f"aws autoscaling set-desired-capacity --auto-scaling-group-name {self.asg_name} --desired-capacity {desired_capacity}"
print(f"Increase capacity [{current_capacity} -> {desired_capacity}]")
res = Shell.check(
command=command,
verbose=True,
)
if not res:
print("ERROR: failed to increase capacity - cannot scale up")
def handle_signal(signum, _frame):
print(f"FATAL: Received signal {signum}")
raise RuntimeError(f"killed by signal {signum}")
def run():
signal.signal(signal.SIGINT, handle_signal)
signal.signal(signal.SIGTERM, handle_signal)
m = None
try:
m = StateMachine()
m.run()
except Exception as e:
print(f"FATAL: Exception [{e}] - terminate instance")
time.sleep(10)
if m:
m.terminate()
raise e
if __name__ == "__main__":
run()

102
ci/praktika/favicon/lambda_function.py Normal file
View File

@ -0,0 +1,102 @@
import base64
import random
import struct
import zlib
def create_favicon():
# Image dimensions
width = 32
height = 32
# Initialize a transparent background image (RGBA: 4 bytes per pixel)
image_data = bytearray(
[0, 0, 0, 0] * width * height
) # Set alpha to 0 for transparency
# Draw 4 vertical lines with color #FAFF68 (RGB: 250, 255, 104)
line_color = [250, 255, 104, 255] # RGBA for #FAFF68 with full opacity
line_width = 4
space_width = 3
x_start = space_width
line_number = 4
line_height = height - space_width
for i in range(line_number):
# Randomly pick a starting y position for each line
y_start = random.randint(0, height - 1)
# Draw the line with random shift along Y-axis
for y in range(line_height):
y_pos = (y + y_start) % height
for x in range(line_width):
pixel_index = (y_pos * width + x_start + x) * 4
image_data[pixel_index : pixel_index + 4] = line_color
x_start += line_width + space_width
# Convert the RGBA image to PNG format
png_data = create_png(width, height, image_data)
# Convert PNG to ICO format
ico_data = create_ico(png_data)
return ico_data
def create_png(width, height, image_data):
def write_chunk(chunk_type, data):
chunk_len = struct.pack(">I", len(data))
chunk_crc = struct.pack(">I", zlib.crc32(chunk_type + data) & 0xFFFFFFFF)
return chunk_len + chunk_type + data + chunk_crc
png_signature = b"\x89PNG\r\n\x1a\n"
ihdr_chunk = struct.pack(">IIBBBBB", width, height, 8, 6, 0, 0, 0)
idat_data = zlib.compress(
b"".join(
b"\x00" + image_data[y * width * 4 : (y + 1) * width * 4]
for y in range(height)
),
9,
)
idat_chunk = write_chunk(b"IDAT", idat_data)
iend_chunk = write_chunk(b"IEND", b"")
return png_signature + write_chunk(b"IHDR", ihdr_chunk) + idat_chunk + iend_chunk
def create_ico(png_data):
# ICO header: reserved (2 bytes), type (2 bytes), image count (2 bytes)
ico_header = struct.pack("<HHH", 0, 1, 1)
# ICO entry: width, height, color count, reserved, color planes, bits per pixel, size, offset
ico_entry = struct.pack("<BBBBHHII", 32, 32, 0, 0, 1, 32, len(png_data), 22)
return ico_header + ico_entry + png_data
def save_favicon_to_disk(ico_data, file_path="favicon.ico"):
with open(file_path, "wb") as f:
f.write(ico_data)
print(f"Favicon saved to {file_path}")
def lambda_handler(event, context):
# Generate the favicon
favicon_data = create_favicon()
# Return the favicon as a binary response
return {
"statusCode": 200,
"headers": {
"Content-Type": "image/x-icon",
"Content-Disposition": 'inline; filename="favicon.ico"',
},
"body": base64.b64encode(favicon_data).decode("utf-8"),
"isBase64Encoded": True,
}
# Optional: Call the function directly to generate and save favicon locally (if running outside Lambda)
if __name__ == "__main__":
favicon_data = create_favicon()
save_favicon_to_disk(favicon_data)

105
ci/praktika/gh.py Normal file
View File

@ -0,0 +1,105 @@
import json
import time
from praktika._environment import _Environment
from praktika.result import Result
from praktika.settings import Settings
from praktika.utils import Shell
class GH:
@classmethod
def do_command_with_retries(cls, command):
res = False
retry_count = 0
out, err = "", ""
while retry_count < Settings.MAX_RETRIES_GH and not res:
ret_code, out, err = Shell.get_res_stdout_stderr(command, verbose=True)
res = ret_code == 0
if not res and "Validation Failed" in err:
print("ERROR: GH command validation error")
break
if not res and "Bad credentials" in err:
print("ERROR: GH credentials/auth failure")
break
if not res:
retry_count += 1
time.sleep(5)
if not res:
print(
f"ERROR: Failed to execute gh command [{command}] out:[{out}] err:[{err}] after [{retry_count}] attempts"
)
return res
@classmethod
def post_pr_comment(
cls, comment_body, or_update_comment_with_substring, repo=None, pr=None
):
if not repo:
repo = _Environment.get().REPOSITORY
if not pr:
pr = _Environment.get().PR_NUMBER
if or_update_comment_with_substring:
print(f"check comment [{comment_body}] created")
cmd_check_created = f'gh api -H "Accept: application/vnd.github.v3+json" \
"/repos/{repo}/issues/{pr}/comments" \
--jq \'.[] | {{id: .id, body: .body}}\' | grep -F "{or_update_comment_with_substring}"'
output = Shell.get_output(cmd_check_created)
if output:
comment_ids = []
try:
comment_ids = [
json.loads(item.strip())["id"] for item in output.split("\n")
]
except Exception as ex:
print(f"Failed to retrieve PR comments with [{ex}]")
for id in comment_ids:
cmd = f'gh api \
-X PATCH \
-H "Accept: application/vnd.github.v3+json" \
"/repos/{repo}/issues/comments/{id}" \
-f body=\'{comment_body}\''
print(f"Update existing comments [{id}]")
return cls.do_command_with_retries(cmd)
cmd = f'gh pr comment {pr} --body "{comment_body}"'
return cls.do_command_with_retries(cmd)
@classmethod
def post_commit_status(cls, name, status, description, url):
status = cls.convert_to_gh_status(status)
command = (
f"gh api -X POST -H 'Accept: application/vnd.github.v3+json' "
f"/repos/{_Environment.get().REPOSITORY}/statuses/{_Environment.get().SHA} "
f"-f state='{status}' -f target_url='{url}' "
f"-f description='{description}' -f context='{name}'"
)
return cls.do_command_with_retries(command)
@classmethod
def convert_to_gh_status(cls, status):
if status in (
Result.Status.PENDING,
Result.Status.SUCCESS,
Result.Status.FAILED,
Result.Status.ERROR,
):
return status
if status in Result.Status.RUNNING:
return Result.Status.PENDING
else:
assert (
False
), f"Invalid status [{status}] to be set as GH commit status.state"
if __name__ == "__main__":
# test
GH.post_pr_comment(
comment_body="foobar",
or_update_comment_with_substring="CI",
repo="ClickHouse/praktika",
pr=15,
)

71
ci/praktika/gh_auth.py Normal file
View File

@ -0,0 +1,71 @@
import sys
import time
from typing import List
import requests
from jwt import JWT, jwk_from_pem
from praktika import Workflow
from praktika.mangle import _get_workflows
from praktika.settings import Settings
from praktika.utils import Shell
class GHAuth:
@staticmethod
def _generate_jwt(client_id, pem):
pem = str.encode(pem)
signing_key = jwk_from_pem(pem)
payload = {
"iat": int(time.time()),
"exp": int(time.time()) + 600,
"iss": client_id,
}
# Create JWT
jwt_instance = JWT()
encoded_jwt = jwt_instance.encode(payload, signing_key, alg="RS256")
return encoded_jwt
@staticmethod
def _get_installation_id(jwt_token):
headers = {
"Authorization": f"Bearer {jwt_token}",
"Accept": "application/vnd.github.v3+json",
}
response = requests.get(
"https://api.github.com/app/installations", headers=headers, timeout=10
)
response.raise_for_status()
installations = response.json()
assert installations, "No installations found for the GitHub App"
return installations[0]["id"]
@staticmethod
def _get_access_token(jwt_token, installation_id):
headers = {
"Authorization": f"Bearer {jwt_token}",
"Accept": "application/vnd.github.v3+json",
}
url = (
f"https://api.github.com/app/installations/{installation_id}/access_tokens"
)
response = requests.post(url, headers=headers, timeout=10)
response.raise_for_status()
return response.json()["token"]
@classmethod
def auth(cls, workflow_name) -> None:
wf = _get_workflows(workflow_name) # type: List[Workflow.Config]
pem = wf[0].get_secret(Settings.SECRET_GH_APP_PEM_KEY).get_value()
assert pem
app_id = wf[0].get_secret(Settings.SECRET_GH_APP_ID).get_value()
# Generate JWT
jwt_token = cls._generate_jwt(app_id, pem)
# Get Installation ID
installation_id = cls._get_installation_id(jwt_token)
# Get Installation Access Token
access_token = cls._get_access_token(jwt_token, installation_id)
Shell.check(f"echo {access_token} | gh auth login --with-token", strict=True)
if __name__ == "__main__":
GHAuth.auth(sys.argv[1])

124
ci/praktika/hook_cache.py Normal file
View File

@ -0,0 +1,124 @@
from praktika._environment import _Environment
from praktika.cache import Cache
from praktika.mangle import _get_workflows
from praktika.runtime import RunConfig
from praktika.settings import Settings
from praktika.utils import Utils
class CacheRunnerHooks:
@classmethod
def configure(cls, _workflow):
workflow_config = RunConfig.from_fs(_workflow.name)
cache = Cache()
assert _Environment.get().WORKFLOW_NAME
workflow = _get_workflows(name=_Environment.get().WORKFLOW_NAME)[0]
print(f"Workflow Configure, workflow [{workflow.name}]")
assert (
workflow.enable_cache
), f"Outdated yaml pipelines or BUG. Configuration must be run only for workflow with enabled cache, workflow [{workflow.name}]"
artifact_digest_map = {}
job_digest_map = {}
for job in workflow.jobs:
if not job.digest_config:
print(
f"NOTE: job [{job.name}] has no Config.digest_config - skip cache check, always run"
)
digest = cache.digest.calc_job_digest(job_config=job)
job_digest_map[job.name] = digest
if job.provides:
# assign the job digest also to the artifacts it provides
for artifact in job.provides:
artifact_digest_map[artifact] = digest
for job in workflow.jobs:
digests_combined_list = []
if job.requires:
# include digest of required artifact to the job digest, so that they affect job state
for artifact_name in job.requires:
if artifact_name not in [
artifact.name for artifact in workflow.artifacts
]:
# phony artifact assumed to be not affecting jobs that depend on it
continue
digests_combined_list.append(artifact_digest_map[artifact_name])
digests_combined_list.append(job_digest_map[job.name])
final_digest = "-".join(digests_combined_list)
workflow_config.digest_jobs[job.name] = final_digest
assert (
workflow_config.digest_jobs
), f"BUG, Workflow with enabled cache must have job digests after configuration, wf [{workflow.name}]"
print("Check remote cache")
job_to_cache_record = {}
for job_name, job_digest in workflow_config.digest_jobs.items():
record = cache.fetch_success(job_name=job_name, job_digest=job_digest)
if record:
assert (
Utils.normalize_string(job_name)
not in workflow_config.cache_success
)
workflow_config.cache_success.append(job_name)
workflow_config.cache_success_base64.append(Utils.to_base64(job_name))
job_to_cache_record[job_name] = record
print("Check artifacts to reuse")
for job in workflow.jobs:
if job.name in workflow_config.cache_success:
if job.provides:
for artifact_name in job.provides:
workflow_config.cache_artifacts[artifact_name] = (
job_to_cache_record[job.name]
)
print(f"Write config to GH's job output")
with open(_Environment.get().JOB_OUTPUT_STREAM, "a", encoding="utf8") as f:
print(
f"DATA={workflow_config.to_json()}",
file=f,
)
print(f"WorkflowRuntimeConfig: [{workflow_config.to_json(pretty=True)}]")
print(
"Dump WorkflowConfig to fs, the next hooks in this job might want to see it"
)
workflow_config.dump()
return workflow_config
@classmethod
def pre_run(cls, _workflow, _job, _required_artifacts=None):
path_prefixes = []
if _job.name == Settings.CI_CONFIG_JOB_NAME:
# SPECIAL handling
return path_prefixes
env = _Environment.get()
runtime_config = RunConfig.from_fs(_workflow.name)
required_artifacts = []
if _required_artifacts:
required_artifacts = _required_artifacts
for artifact in required_artifacts:
if artifact.name in runtime_config.cache_artifacts:
record = runtime_config.cache_artifacts[artifact.name]
print(f"Reuse artifact [{artifact.name}] from [{record}]")
path_prefixes.append(
env.get_s3_prefix_static(
record.pr_number, record.branch, record.sha
)
)
else:
path_prefixes.append(env.get_s3_prefix())
return path_prefixes
@classmethod
def run(cls, workflow, job):
pass
@classmethod
def post_run(cls, workflow, job):
if job.name == Settings.CI_CONFIG_JOB_NAME:
return
if job.digest_config:
# cache is enabled, and it's a job that supposed to be cached (has defined digest config)
workflow_runtime = RunConfig.from_fs(workflow.name)
job_digest = workflow_runtime.digest_jobs[job.name]
Cache.push_success_record(job.name, job_digest, workflow_runtime.sha)

198
ci/praktika/hook_html.py Normal file
View File

@ -0,0 +1,198 @@
import dataclasses
import json
import urllib.parse
from pathlib import Path
from typing import List
from praktika._environment import _Environment
from praktika.gh import GH
from praktika.parser import WorkflowConfigParser
from praktika.result import Result, ResultInfo
from praktika.runtime import RunConfig
from praktika.s3 import S3
from praktika.settings import Settings
from praktika.utils import Shell, Utils
@dataclasses.dataclass
class GitCommit:
date: str
message: str
sha: str
@staticmethod
def from_json(json_data: str) -> List["GitCommit"]:
commits = []
try:
data = json.loads(json_data)
commits = [
GitCommit(
message=commit["messageHeadline"],
sha=commit["oid"],
date=commit["committedDate"],
)
for commit in data.get("commits", [])
]
except Exception as e:
print(
f"ERROR: Failed to deserialize commit's data: [{json_data}], ex: [{e}]"
)
return commits
class HtmlRunnerHooks:
@classmethod
def configure(cls, _workflow):
def _get_pr_commits(pr_number):
res = []
if not pr_number:
return res
output = Shell.get_output(f"gh pr view {pr_number} --json commits")
if output:
res = GitCommit.from_json(output)
return res
# generate pending Results for all jobs in the workflow
if _workflow.enable_cache:
skip_jobs = RunConfig.from_fs(_workflow.name).cache_success
else:
skip_jobs = []
env = _Environment.get()
results = []
for job in _workflow.jobs:
if job.name not in skip_jobs:
result = Result.generate_pending(job.name)
else:
result = Result.generate_skipped(job.name)
results.append(result)
summary_result = Result.generate_pending(_workflow.name, results=results)
summary_result.aux_links.append(env.CHANGE_URL)
summary_result.aux_links.append(env.RUN_URL)
summary_result.start_time = Utils.timestamp()
page_url = "/".join(
["https:/", Settings.HTML_S3_PATH, str(Path(Settings.HTML_PAGE_FILE).name)]
)
for bucket, endpoint in Settings.S3_BUCKET_TO_HTTP_ENDPOINT.items():
page_url = page_url.replace(bucket, endpoint)
# TODO: add support for non-PRs (use branch?)
page_url += f"?PR={env.PR_NUMBER}&sha=latest&name_0={urllib.parse.quote(env.WORKFLOW_NAME, safe='')}"
summary_result.html_link = page_url
# clean the previous latest results in PR if any
if env.PR_NUMBER:
S3.clean_latest_result()
S3.copy_result_to_s3(
summary_result,
unlock=False,
)
print(f"CI Status page url [{page_url}]")
res1 = GH.post_commit_status(
name=_workflow.name,
status=Result.Status.PENDING,
description="",
url=page_url,
)
res2 = GH.post_pr_comment(
comment_body=f"Workflow [[{_workflow.name}]({page_url})], commit [{_Environment.get().SHA[:8]}]",
or_update_comment_with_substring=f"Workflow [",
)
if not (res1 or res2):
Utils.raise_with_error(
"Failed to set both GH commit status and PR comment with Workflow Status, cannot proceed"
)
if env.PR_NUMBER:
commits = _get_pr_commits(env.PR_NUMBER)
# TODO: upload commits data to s3 to visualise it on a report page
print(commits)
@classmethod
def pre_run(cls, _workflow, _job):
result = Result.from_fs(_job.name)
S3.copy_result_from_s3(
Result.file_name_static(_workflow.name),
)
workflow_result = Result.from_fs(_workflow.name)
workflow_result.update_sub_result(result)
S3.copy_result_to_s3(
workflow_result,
unlock=True,
)
@classmethod
def run(cls, _workflow, _job):
pass
@classmethod
def post_run(cls, _workflow, _job, info_errors):
result = Result.from_fs(_job.name)
env = _Environment.get()
S3.copy_result_from_s3(
Result.file_name_static(_workflow.name),
lock=True,
)
workflow_result = Result.from_fs(_workflow.name)
print(f"Workflow info [{workflow_result.info}], info_errors [{info_errors}]")
env_info = env.REPORT_INFO
if env_info:
print(
f"WARNING: some info lines are set in Environment - append to report [{env_info}]"
)
info_errors += env_info
if info_errors:
info_errors = [f" | {error}" for error in info_errors]
info_str = f"{_job.name}:\n"
info_str += "\n".join(info_errors)
print("Update workflow results with new info")
workflow_result.set_info(info_str)
old_status = workflow_result.status
S3.upload_result_files_to_s3(result)
workflow_result.update_sub_result(result)
skipped_job_results = []
if not result.is_ok():
print(
"Current job failed - find dependee jobs in the workflow and set their statuses to skipped"
)
workflow_config_parsed = WorkflowConfigParser(_workflow).parse()
for dependee_job in workflow_config_parsed.workflow_yaml_config.jobs:
if _job.name in dependee_job.needs:
if _workflow.get_job(dependee_job.name).run_unless_cancelled:
continue
print(
f"NOTE: Set job [{dependee_job.name}] status to [{Result.Status.SKIPPED}] due to current failure"
)
skipped_job_results.append(
Result(
name=dependee_job.name,
status=Result.Status.SKIPPED,
info=ResultInfo.SKIPPED_DUE_TO_PREVIOUS_FAILURE
+ f" [{_job.name}]",
)
)
for skipped_job_result in skipped_job_results:
workflow_result.update_sub_result(skipped_job_result)
S3.copy_result_to_s3(
workflow_result,
unlock=True,
)
if workflow_result.status != old_status:
print(
f"Update GH commit status [{result.name}]: [{old_status} -> {workflow_result.status}], link [{workflow_result.html_link}]"
)
GH.post_commit_status(
name=workflow_result.name,
status=GH.convert_to_gh_status(workflow_result.status),
description="",
url=workflow_result.html_link,
)

43
ci/praktika/hook_interface.py Normal file
View File

@ -0,0 +1,43 @@
from abc import ABC, abstractmethod
from praktika import Workflow
class HookInterface(ABC):
@abstractmethod
def pre_run(self, _workflow, _job):
"""
runs in pre-run step
:param _workflow:
:param _job:
:return:
"""
pass
@abstractmethod
def run(self, _workflow, _job):
"""
runs in run step
:param _workflow:
:param _job:
:return:
"""
pass
@abstractmethod
def post_run(self, _workflow, _job):
"""
runs in post-run step
:param _workflow:
:param _job:
:return:
"""
pass
@abstractmethod
def configure(self, _workflow: Workflow.Config):
"""
runs in initial WorkflowConfig job in run step
:return:
"""
pass

10
ci/praktika/html_prepare.py Normal file
View File

@ -0,0 +1,10 @@
from praktika.s3 import S3
from praktika.settings import Settings
class Html:
@classmethod
def prepare(cls):
S3.copy_file_to_s3(
s3_path=Settings.HTML_S3_PATH, local_path=Settings.HTML_PAGE_FILE
)

102
ci/praktika/job.py Normal file
View File

@ -0,0 +1,102 @@
import copy
import json
from dataclasses import dataclass, field
from typing import Any, List, Optional
class Job:
@dataclass
class Requirements:
python: bool = False
python_requirements_txt: str = ""
@dataclass
class CacheDigestConfig:
include_paths: List[str] = field(default_factory=list)
exclude_paths: List[str] = field(default_factory=list)
@dataclass
class Config:
# Job Name
name: str
# Machine's label to run job on. For instance [ubuntu-latest] for free gh runner
runs_on: List[str]
# Job Run Command
command: str
# What job requires
# May be phony or physical names
requires: List[str] = field(default_factory=list)
# What job provides
# May be phony or physical names
provides: List[str] = field(default_factory=list)
job_requirements: Optional["Job.Requirements"] = None
timeout: int = 1 * 3600
digest_config: Optional["Job.CacheDigestConfig"] = None
run_in_docker: str = ""
run_unless_cancelled: bool = False
allow_merge_on_failure: bool = False
parameter: Any = None
def parametrize(
self,
parameter: Optional[List[Any]] = None,
runs_on: Optional[List[List[str]]] = None,
timeout: Optional[List[int]] = None,
):
assert (
parameter or runs_on
), "Either :parameter or :runs_on must be non empty list for parametrisation"
if not parameter:
parameter = [None] * len(runs_on)
if not runs_on:
runs_on = [None] * len(parameter)
if not timeout:
timeout = [None] * len(parameter)
assert (
len(parameter) == len(runs_on) == len(timeout)
), "Parametrization lists must be of the same size"
res = []
for parameter_, runs_on_, timeout_ in zip(parameter, runs_on, timeout):
obj = copy.deepcopy(self)
if parameter_:
obj.parameter = parameter_
if runs_on_:
obj.runs_on = runs_on_
if timeout_:
obj.timeout = timeout_
obj.name = obj.get_job_name_with_parameter()
res.append(obj)
return res
def get_job_name_with_parameter(self):
name, parameter, runs_on = self.name, self.parameter, self.runs_on
res = name
name_params = []
if isinstance(parameter, list) or isinstance(parameter, dict):
name_params.append(json.dumps(parameter))
elif parameter is not None:
name_params.append(parameter)
if runs_on:
assert isinstance(runs_on, list)
name_params.append(json.dumps(runs_on))
if name_params:
name_params = [str(param) for param in name_params]
res += f" ({', '.join(name_params)})"
self.name = res
return res
def __repr__(self):
return self.name

745
ci/praktika/json.html Normal file
View File

@ -0,0 +1,745 @@
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>praktika report</title>
<link rel="icon" href="https://w4z3pajszlbkfcw2wcylfei5km0xmwag.lambda-url.us-east-1.on.aws/" type="image/x-icon">
<style>
/* Default (Day Theme) */
:root {
--background-color: white;
--text-color: #000;
--tile-background: #f9f9f9;
--footer-background: #f1f1f1;
--footer-text-color: #000;
--status-width: 300px;
}
body {
background-color: var(--background-color);
color: var(--text-color);
height: 100%;
margin: 0;
display: flex;
flex-direction: column;
font-family: 'IBM Plex Mono Condensed', monospace, sans-serif;
--header-background-color: #f4f4f4;
}
body.night-theme {
--background-color: #1F1F1C;
--text-color: #fff;
--tile-background: black;
--header-background-color: #1F1F1C;
}
#info-container {
margin-left: calc(var(--status-width) + 20px);
margin-bottom: 10px;
background-color: var(--tile-background);
padding: 10px;
text-align: left;
}
#status-container {
position: fixed;
top: 0;
bottom: 0;
left: 0;
width: var(--status-width);
background-color: var(--tile-background);
padding: 20px;
box-sizing: border-box;
font-size: 18px;
margin: 0;
}
#status-container a {
color: #007bff;
text-decoration: underline;
font-weight: bold;
cursor: pointer;
display: inline-block;
margin-top: 5px;
margin-left: 20px;
padding: 2px 0;
font-size: 0.8em;
}
#status-container a:hover {
color: #0056b3;
text-decoration: none;
}
.key-value-pair {
display: flex; /* Enable Flexbox for alignment */
justify-content: space-between; /* Distribute space between key and value */
margin-bottom: 20px; /* Add space between each pair */
}
.json-key {
font-weight: bold;
}
.json-value {
font-weight: normal;
font-family: 'Source Code Pro', monospace, sans-serif;
letter-spacing: -0.5px;
}
#result-container {
background-color: var(--tile-background);
margin-left: calc(var(--status-width) + 20px);
padding: 20px;
box-sizing: border-box;
text-align: center;
font-size: 18px;
font-weight: normal;
flex-grow: 1;
}
#footer {
padding: 10px;
position: fixed;
bottom: 0;
left: 0;
right: 0;
background-color: #1F1F1C;
color: white;
font-size: 14px;
display: flex;
justify-content: space-between; /* Ensure the .left expands, and .right and .settings are aligned to the right */
align-items: center;
}
#footer a {
color: white;
text-decoration: none;
}
#footer .left {
flex-grow: 1; /* Takes up all the available space */
}
/* make some space around '/' in the navigation line */
#footer .left span.separator {
margin-left: 5px;
margin-right: 5px;
}
#footer .right, #footer .settings {
display: flex;
align-items: center;
}
#footer .right a::before {
content: "#";
margin-left: 10px;
color: #e0e0e0;
}
#footer .right::before, #footer .settings::before {
content: "|"; /* Add separator before right and settings sections */
margin-left: 10px;
margin-right: 10px;
color: #e0e0e0;
}
#theme-toggle {
cursor: pointer;
font-size: 20px;
color: white;
}
#theme-toggle:hover {
color: #e0e0e0;
}
#footer a:hover {
text-decoration: underline;
}
#links {
margin-top: 10px;
padding: 15px;
border: 1px solid #ccc;
border-radius: 5px;
background-color: #f9f9f9;
}
#links a {
display: block;
margin-bottom: 5px;
padding: 5px 10px;
background-color: #D5D5D5;
color: black;
text-decoration: none;
border-radius: 5px;
}
#links a:hover {
background-color: #D5D5D5;
}
table {
width: 100%;
border-collapse: collapse;
}
th.name-column, td.name-column {
max-width: 400px; /* Set the maximum width for the column */
white-space: nowrap; /* Prevent text from wrapping */
overflow: hidden; /* Hide the overflowed text */
text-overflow: ellipsis; /* Show ellipsis (...) for overflowed text */
}
th.status-column, td.status-column {
max-width: 100px; /* Set the maximum width for the column */
white-space: nowrap; /* Prevent text from wrapping */
overflow: hidden; /* Hide the overflowed text */
text-overflow: ellipsis; /* Show ellipsis (...) for overflowed text */
}
th.time-column, td.time-column {
max-width: 120px; /* Set the maximum width for the column */
white-space: nowrap; /* Prevent text from wrapping */
text-align: right;
}
th.info-column, td.info-column {
width: 100%; /* Allow the column to take all the remaining space */
}
th, td {
padding: 8px;
border: 1px solid #ddd;
text-align: left;
}
th {
background-color: var(--header-background-color);
}
.status-success {
color: green;
font-weight: bold;
}
.status-fail {
color: red;
font-weight: bold;
}
.status-pending {
color: #d4a017;
font-weight: bold;
}
.status-broken {
color: purple;
font-weight: bold;
}
.status-run {
color: blue;
font-weight: bold;
}
.status-error {
color: darkred;
font-weight: bold;
}
.status-other {
color: grey;
font-weight: bold;
}
</style>
</head>
<body>
<div id="info-container"></div>
<div id="status-container"></div>
<div id="result-container"></div>
<footer id="footer">
<div class="left"></div>
<div class="right"></div>
<div class="settings">
<span id="theme-toggle">☀️</span>
</div>
</footer>
<script>
function toggleTheme() {
document.body.classList.toggle('night-theme');
const toggleIcon = document.getElementById('theme-toggle');
if (document.body.classList.contains('night-theme')) {
toggleIcon.textContent = '☾'; // Moon for night mode
} else {
toggleIcon.textContent = '☀️'; // Sun for day mode
}
}
// Attach the toggle function to the click event of the icon
document.getElementById('theme-toggle').addEventListener('click', toggleTheme);
function formatTimestamp(timestamp, showDate = true) {
const date = new Date(timestamp * 1000);
const day = String(date.getDate()).padStart(2, '0');
const monthNames = ["Jan", "Feb", "Mar", "Apr", "May", "Jun",
"Jul", "Aug", "Sep", "Oct", "Nov", "Dec"];
const month = monthNames[date.getMonth()];
const year = date.getFullYear();
const hours = String(date.getHours()).padStart(2, '0');
const minutes = String(date.getMinutes()).padStart(2, '0');
const seconds = String(date.getSeconds()).padStart(2, '0');
//const milliseconds = String(date.getMilliseconds()).padStart(2, '0');
return showDate
? `${day}-${month}-${year} ${hours}:${minutes}:${seconds}`
: `${hours}:${minutes}:${seconds}`;
}
function formatDuration(durationInSeconds, detailed = false) {
// Check if the duration is empty, null, or not a number
if (!durationInSeconds || isNaN(durationInSeconds)) {
return '';
}
// Ensure duration is a floating-point number
const duration = parseFloat(durationInSeconds);
if (detailed) {
// Format in the detailed format with hours, minutes, and seconds
const hours = Math.floor(duration / 3600);
const minutes = Math.floor((duration % 3600) / 60);
const seconds = Math.floor(duration % 60);
const formattedHours = hours > 0 ? `${hours}h ` : '';
const formattedMinutes = minutes > 0 ? `${minutes}m ` : '';
const formattedSeconds = `${String(seconds).padStart(2, '0')}s`;
return `${formattedHours}${formattedMinutes}${formattedSeconds}`.trim();
} else {
// Format in the default format with seconds and milliseconds
const seconds = Math.floor(duration);
const milliseconds = Math.floor((duration % 1) * 1000);
const formattedSeconds = String(seconds);
const formattedMilliseconds = String(milliseconds).padStart(3, '0');
return `${formattedSeconds}.${formattedMilliseconds}`;
}
}
// Function to determine status class based on value
function getStatusClass(status) {
const lowerStatus = status.toLowerCase();
if (lowerStatus.includes('success') || lowerStatus === 'ok') return 'status-success';
if (lowerStatus.includes('fail')) return 'status-fail';
if (lowerStatus.includes('pending')) return 'status-pending';
if (lowerStatus.includes('broken')) return 'status-broken';
if (lowerStatus.includes('run')) return 'status-run';
if (lowerStatus.includes('error')) return 'status-error';
return 'status-other';
}
function addKeyValueToStatus(key, value) {
const statusContainer = document.getElementById('status-container');
let keyValuePair = document.createElement('div');
keyValuePair.className = 'key-value-pair';
const keyElement = document.createElement('div');
keyElement.className = 'json-key';
keyElement.textContent = key + ':';
const valueElement = document.createElement('div');
valueElement.className = 'json-value';
valueElement.textContent = value;
keyValuePair.appendChild(keyElement)
keyValuePair.appendChild(valueElement)
statusContainer.appendChild(keyValuePair);
}
function addFileButtonToStatus(key, links) {
if (links == null) {
return
}
const statusContainer = document.getElementById('status-container');
const keyElement = document.createElement('div');
keyElement.className = 'json-key';
keyElement.textContent = columnSymbols[key] + ':' || key;
statusContainer.appendChild(keyElement);
if (Array.isArray(links) && links.length > 0) {
links.forEach(link => {
const textLink = document.createElement('a');
textLink.href = link;
textLink.textContent = link.split('/').pop();
textLink.target = '_blank';
statusContainer.appendChild(textLink);
statusContainer.appendChild(document.createElement('br'));
});
}
}
function addStatusToStatus(status, start_time, duration) {
const statusContainer = document.getElementById('status-container')
let keyValuePair = document.createElement('div');
keyValuePair.className = 'key-value-pair';
let keyElement = document.createElement('div');
let valueElement = document.createElement('div');
keyElement.className = 'json-key';
valueElement.className = 'json-value';
keyElement.textContent = columnSymbols['status'] + ':' || 'status:';
valueElement.classList.add('status-value');
valueElement.classList.add(getStatusClass(status));
valueElement.textContent = status;
keyValuePair.appendChild(keyElement);
keyValuePair.appendChild(valueElement);
statusContainer.appendChild(keyValuePair);
keyValuePair = document.createElement('div');
keyValuePair.className = 'key-value-pair';
keyElement = document.createElement('div');
valueElement = document.createElement('div');
keyElement.className = 'json-key';
valueElement.className = 'json-value';
keyElement.textContent = columnSymbols['start_time'] + ':' || 'start_time:';
valueElement.textContent = formatTimestamp(start_time);
keyValuePair.appendChild(keyElement);
keyValuePair.appendChild(valueElement);
statusContainer.appendChild(keyValuePair);
keyValuePair = document.createElement('div');
keyValuePair.className = 'key-value-pair';
keyElement = document.createElement('div');
valueElement = document.createElement('div');
keyElement.className = 'json-key';
valueElement.className = 'json-value';
keyElement.textContent = columnSymbols['duration'] + ':' || 'duration:';
if (duration === null) {
// Set initial value to 0 and add a unique ID or data attribute to identify the duration element
valueElement.textContent = '00:00:00';
valueElement.setAttribute('id', 'duration-value');
} else {
// Format the duration if it's a valid number
valueElement.textContent = formatDuration(duration, true);
}
keyValuePair.appendChild(keyElement);
keyValuePair.appendChild(valueElement);
statusContainer.appendChild(keyValuePair);
}
function navigatePath(jsonObj, nameArray) {
let baseParams = new URLSearchParams(window.location.search);
let keysToDelete = [];
baseParams.forEach((value, key) => {
if (key.startsWith('name_')) {
keysToDelete.push(key); // Collect the keys to delete
}
});
keysToDelete.forEach((key) => baseParams.delete(key));
let pathNames = [];
let pathLinks = [];
let currentObj = jsonObj;
// Add the first entry (root level)
baseParams.set(`name_0`, currentObj.name);
pathNames.push(currentObj.name);
pathLinks.push(`<span class="separator">/</span><a href="${window.location.pathname}?${baseParams.toString()}">${currentObj.name}</a>`);
// Iterate through the nameArray starting at index 0
for (const [index, name] of nameArray.entries()) {
if (index === 0) continue;
if (currentObj && Array.isArray(currentObj.results)) {
const nextResult = currentObj.results.find(result => result.name === name);
if (nextResult) {
baseParams.set(`name_${index}`, nextResult.name);
pathNames.push(nextResult.name); // Correctly push nextResult name, not currentObj.name
pathLinks.push(`<span class="separator">/</span><a href="${window.location.pathname}?${baseParams.toString()}">${nextResult.name}</a>`);
currentObj = nextResult; // Move to the next object in the hierarchy
} else {
console.error(`Name "${name}" not found in results array.`);
return null; // Name not found in results array
}
} else {
console.error(`Current object is not structured as expected.`);
return null; // Current object is not structured as expected
}
}
const footerLeft = document.querySelector('#footer .left');
footerLeft.innerHTML = pathLinks.join('');
return currentObj;
}
// Define the fixed columns globally, so both functions can use it
const columns = ['name', 'status', 'start_time', 'duration', 'info'];
const columnSymbols = {
name: '📂',
status: '✔️',
start_time: '🕒',
duration: '⏳',
info: 'ℹ️',
files: '📄'
};
function createResultsTable(results, nest_level) {
if (results && Array.isArray(results) && results.length > 0) {
const table = document.createElement('table');
const thead = document.createElement('thead');
const tbody = document.createElement('tbody');
// Get the current URL parameters
const currentUrl = new URL(window.location.href);
// Create table headers based on the fixed columns
const headerRow = document.createElement('tr');
columns.forEach(column => {
const th = document.createElement('th');
th.textContent = th.textContent = columnSymbols[column] || column;
th.style.cursor = 'pointer'; // Make headers clickable
th.addEventListener('click', () => sortTable(results, column, tbody, nest_level)); // Add click event to sort the table
headerRow.appendChild(th);
});
thead.appendChild(headerRow);
// Create table rows
populateTableRows(tbody, results, columns, nest_level);
table.appendChild(thead);
table.appendChild(tbody);
return table;
}
return null;
}
function populateTableRows(tbody, results, columns, nest_level) {
const currentUrl = new URL(window.location.href); // Get the current URL
// Clear existing rows if re-rendering (used in sorting)
tbody.innerHTML = '';
results.forEach((result, index) => {
const row = document.createElement('tr');
columns.forEach(column => {
const td = document.createElement('td');
const value = result[column];
if (column === 'name') {
// Create a link for the name field, using name_X
const link = document.createElement('a');
const newUrl = new URL(currentUrl); // Create a fresh copy of the URL for each row
newUrl.searchParams.set(`name_${nest_level}`, value); // Use backticks for string interpolation
link.href = newUrl.toString();
link.textContent = value;
td.classList.add('name-column');
td.appendChild(link);
} else if (column === 'status') {
// Apply status formatting
const span = document.createElement('span');
span.className = getStatusClass(value);
span.textContent = value;
td.classList.add('status-column');
td.appendChild(span);
} else if (column === 'start_time') {
td.classList.add('time-column');
td.textContent = value ? formatTimestamp(value, false) : '';
} else if (column === 'duration') {
td.classList.add('time-column');
td.textContent = value ? formatDuration(value) : '';
} else if (column === 'info') {
// For info and other columns, just display the value
td.textContent = value || '';
td.classList.add('info-column');
}
row.appendChild(td);
});
tbody.appendChild(row);
});
}
function sortTable(results, key, tbody, nest_level) {
// Find the table header element for the given key
let th = null;
const tableHeaders = document.querySelectorAll('th'); // Select all table headers
tableHeaders.forEach(header => {
if (header.textContent.trim().toLowerCase() === key.toLowerCase()) {
th = header;
}
});
if (!th) {
console.error(`No table header found for key: ${key}`);
return;
}
// Determine the current sort direction
let ascending = th.getAttribute('data-sort-direction') === 'asc' ? false : true;
// Toggle the sort direction for the next click
th.setAttribute('data-sort-direction', ascending ? 'asc' : 'desc');
// Sort the results array by the given key
results.sort((a, b) => {
if (a[key] < b[key]) return ascending ? -1 : 1;
if (a[key] > b[key]) return ascending ? 1 : -1;
return 0;
});
// Re-populate the table with sorted data
populateTableRows(tbody, results, columns, nest_level);
}
function loadJSON(PR, sha, nameParams) {
const infoElement = document.getElementById('info-container');
let lastModifiedTime = null;
const task = nameParams[0].toLowerCase();
// Construct the URL dynamically based on PR, sha, and name_X
const baseUrl = window.location.origin + window.location.pathname.replace('/json.html', '');
const path = `${baseUrl}/${encodeURIComponent(PR)}/${encodeURIComponent(sha)}/result_${task}.json`;
fetch(path, {cache: "no-cache"})
.then(response => {
if (!response.ok) {
throw new Error(`HTTP error! status: ${response.status}`);
}
lastModifiedTime = response.headers.get('Last-Modified');
return response.json();
})
.then(data => {
const linksDiv = document.getElementById('links');
const resultsDiv = document.getElementById('result-container');
const footerRight = document.querySelector('#footer .right');
let targetData = navigatePath(data, nameParams);
let nest_level = nameParams.length;
if (targetData) {
infoElement.style.display = 'none';
// Handle footer links if present
if (Array.isArray(data.aux_links) && data.aux_links.length > 0) {
data.aux_links.forEach(link => {
const a = document.createElement('a');
a.href = link;
a.textContent = link.split('/').pop();
a.target = '_blank';
footerRight.appendChild(a);
});
}
addStatusToStatus(targetData.status, targetData.start_time, targetData.duration)
// Handle links
addFileButtonToStatus('files', targetData.links)
// Handle duration update if duration is null and start_time exists
if (targetData.duration === null && targetData.start_time) {
let duration = Math.floor(Date.now() / 1000 - targetData.start_time);
const durationElement = document.getElementById('duration-value');
const intervalId = setInterval(() => {
duration++;
durationElement.textContent = formatDuration(duration, true);
}, 1000);
}
// If 'results' exists and is non-empty, create the table
const resultsData = targetData.results;
if (Array.isArray(resultsData) && resultsData.length > 0) {
const table = createResultsTable(resultsData, nest_level);
if (table) {
resultsDiv.appendChild(table);
}
}
} else {
infoElement.textContent = 'Object Not Found';
infoElement.style.display = 'block';
}
// Set up auto-reload if Last-Modified header is present
if (lastModifiedTime) {
setInterval(() => {
checkForUpdate(path, lastModifiedTime);
}, 30000); // 30000 milliseconds = 30 seconds
}
})
.catch(error => {
console.error('Error loading JSON:', error);
infoElement.textContent = 'Error loading data';
infoElement.style.display = 'block';
});
}
// Function to check if the JSON file is updated
function checkForUpdate(path, lastModifiedTime) {
fetch(path, {method: 'HEAD'})
.then(response => {
if (!response.ok) {
throw new Error(`HTTP error! status: ${response.status}`);
}
const newLastModifiedTime = response.headers.get('Last-Modified');
if (newLastModifiedTime && new Date(newLastModifiedTime) > new Date(lastModifiedTime)) {
// If the JSON file has been updated, reload the page
window.location.reload();
}
})
.catch(error => {
console.error('Error checking for update:', error);
});
}
// Initialize the page and load JSON from URL parameter
function init() {
const urlParams = new URLSearchParams(window.location.search);
const PR = urlParams.get('PR');
const sha = urlParams.get('sha');
const root_name = urlParams.get('name_0');
const nameParams = [];
urlParams.forEach((value, key) => {
if (key.startsWith('name_')) {
const index = parseInt(key.split('_')[1], 10);
nameParams[index] = value;
}
});
if (PR) {
addKeyValueToStatus("PR", PR)
} else {
console.error("TODO")
}
addKeyValueToStatus("sha", sha);
if (nameParams[1]) {
addKeyValueToStatus("job", nameParams[1]);
}
addKeyValueToStatus("workflow", nameParams[0]);
if (PR && sha && root_name) {
loadJSON(PR, sha, nameParams);
} else {
document.getElementById('title').textContent = 'Error: Missing required URL parameters: PR, sha, or name_0';
}
}
window.onload = init;
</script>
</body>
</html>

137
ci/praktika/mangle.py Normal file
View File

@ -0,0 +1,137 @@
import copy
import importlib.util
from pathlib import Path
from typing import Any, Dict
from praktika import Job
from praktika._settings import _USER_DEFINED_SETTINGS, _Settings
from praktika.utils import ContextManager, Utils
def _get_workflows(name=None, file=None):
"""
Gets user's workflow configs
"""
res = []
with ContextManager.cd():
directory = Path(_Settings.WORKFLOWS_DIRECTORY)
for py_file in directory.glob("*.py"):
if file and file not in str(py_file):
continue
module_name = py_file.name.removeprefix(".py")
spec = importlib.util.spec_from_file_location(
module_name, f"{_Settings.WORKFLOWS_DIRECTORY}/{module_name}"
)
assert spec
foo = importlib.util.module_from_spec(spec)
assert spec.loader
spec.loader.exec_module(foo)
try:
for workflow in foo.WORKFLOWS:
if name:
if name == workflow.name:
print(f"Read workflow [{name}] config from [{module_name}]")
res = [workflow]
break
else:
continue
else:
res += foo.WORKFLOWS
print(f"Read workflow configs from [{module_name}]")
except Exception as e:
print(
f"WARNING: Failed to add WORKFLOWS config from [{module_name}], exception [{e}]"
)
if not res:
Utils.raise_with_error(f"Failed to find workflow [{name or file}]")
for workflow in res:
# add native jobs
_update_workflow_with_native_jobs(workflow)
# fill in artifact properties, e.g. _provided_by
_update_workflow_artifacts(workflow)
return res
def _update_workflow_artifacts(workflow):
artifact_job = {}
for job in workflow.jobs:
for artifact_name in job.provides:
assert artifact_name not in artifact_job
artifact_job[artifact_name] = job.name
for artifact in workflow.artifacts:
artifact._provided_by = artifact_job[artifact.name]
def _update_workflow_with_native_jobs(workflow):
if workflow.dockers:
from praktika.native_jobs import _docker_build_job
print(f"Enable native job [{_docker_build_job.name}] for [{workflow.name}]")
aux_job = copy.deepcopy(_docker_build_job)
if workflow.enable_cache:
print(
f"Add automatic digest config for [{aux_job.name}] job since cache is enabled"
)
docker_digest_config = Job.CacheDigestConfig()
for docker_config in workflow.dockers:
docker_digest_config.include_paths.append(docker_config.path)
aux_job.digest_config = docker_digest_config
workflow.jobs.insert(0, aux_job)
for job in workflow.jobs[1:]:
if not job.requires:
job.requires = []
job.requires.append(aux_job.name)
if (
workflow.enable_cache
or workflow.enable_report
or workflow.enable_merge_ready_status
):
from praktika.native_jobs import _workflow_config_job
print(f"Enable native job [{_workflow_config_job.name}] for [{workflow.name}]")
aux_job = copy.deepcopy(_workflow_config_job)
workflow.jobs.insert(0, aux_job)
for job in workflow.jobs[1:]:
if not job.requires:
job.requires = []
job.requires.append(aux_job.name)
if workflow.enable_merge_ready_status:
from praktika.native_jobs import _final_job
print(f"Enable native job [{_final_job.name}] for [{workflow.name}]")
aux_job = copy.deepcopy(_final_job)
for job in workflow.jobs:
aux_job.requires.append(job.name)
workflow.jobs.append(aux_job)
def _get_user_settings() -> Dict[str, Any]:
"""
Gets user's settings
"""
res = {} # type: Dict[str, Any]
directory = Path(_Settings.SETTINGS_DIRECTORY)
for py_file in directory.glob("*.py"):
module_name = py_file.name.removeprefix(".py")
spec = importlib.util.spec_from_file_location(
module_name, f"{_Settings.SETTINGS_DIRECTORY}/{module_name}"
)
assert spec
foo = importlib.util.module_from_spec(spec)
assert spec.loader
spec.loader.exec_module(foo)
for setting in _USER_DEFINED_SETTINGS:
try:
value = getattr(foo, setting)
res[setting] = value
print(f"Apply user defined setting [{setting} = {value}]")
except Exception as e:
pass
return res

378
ci/praktika/native_jobs.py Normal file
View File

@ -0,0 +1,378 @@
import sys
from typing import Dict
from praktika import Job, Workflow
from praktika._environment import _Environment
from praktika.cidb import CIDB
from praktika.digest import Digest
from praktika.docker import Docker
from praktika.gh import GH
from praktika.hook_cache import CacheRunnerHooks
from praktika.hook_html import HtmlRunnerHooks
from praktika.mangle import _get_workflows
from praktika.result import Result, ResultInfo
from praktika.runtime import RunConfig
from praktika.s3 import S3
from praktika.settings import Settings
from praktika.utils import Shell, Utils
assert Settings.CI_CONFIG_RUNS_ON
_workflow_config_job = Job.Config(
name=Settings.CI_CONFIG_JOB_NAME,
runs_on=Settings.CI_CONFIG_RUNS_ON,
job_requirements=(
Job.Requirements(
python=Settings.INSTALL_PYTHON_FOR_NATIVE_JOBS,
python_requirements_txt=Settings.INSTALL_PYTHON_REQS_FOR_NATIVE_JOBS,
)
if Settings.INSTALL_PYTHON_REQS_FOR_NATIVE_JOBS
else None
),
command=f"{Settings.PYTHON_INTERPRETER} -m praktika.native_jobs '{Settings.CI_CONFIG_JOB_NAME}'",
)
_docker_build_job = Job.Config(
name=Settings.DOCKER_BUILD_JOB_NAME,
runs_on=Settings.DOCKER_BUILD_RUNS_ON,
job_requirements=Job.Requirements(
python=Settings.INSTALL_PYTHON_FOR_NATIVE_JOBS,
python_requirements_txt="",
),
timeout=4 * 3600,
command=f"{Settings.PYTHON_INTERPRETER} -m praktika.native_jobs '{Settings.DOCKER_BUILD_JOB_NAME}'",
)
_final_job = Job.Config(
name=Settings.FINISH_WORKFLOW_JOB_NAME,
runs_on=Settings.CI_CONFIG_RUNS_ON,
job_requirements=Job.Requirements(
python=Settings.INSTALL_PYTHON_FOR_NATIVE_JOBS,
python_requirements_txt="",
),
command=f"{Settings.PYTHON_INTERPRETER} -m praktika.native_jobs '{Settings.FINISH_WORKFLOW_JOB_NAME}'",
run_unless_cancelled=True,
)
def _build_dockers(workflow, job_name):
print(f"Start [{job_name}], workflow [{workflow.name}]")
dockers = workflow.dockers
ready = []
results = []
job_status = Result.Status.SUCCESS
job_info = ""
dockers = Docker.sort_in_build_order(dockers)
docker_digests = {} # type: Dict[str, str]
for docker in dockers:
docker_digests[docker.name] = Digest().calc_docker_digest(docker, dockers)
if not Shell.check(
"docker buildx inspect --bootstrap | grep -q docker-container", verbose=True
):
print("Install docker container driver")
if not Shell.check(
"docker buildx create --use --name mybuilder --driver docker-container",
verbose=True,
):
job_status = Result.Status.FAILED
job_info = "Failed to install docker buildx driver"
if job_status == Result.Status.SUCCESS:
if not Docker.login(
Settings.DOCKERHUB_USERNAME,
user_password=workflow.get_secret(Settings.DOCKERHUB_SECRET).get_value(),
):
job_status = Result.Status.FAILED
job_info = "Failed to login to dockerhub"
if job_status == Result.Status.SUCCESS:
for docker in dockers:
assert (
docker.name not in ready
), f"All docker names must be uniq [{dockers}]"
stopwatch = Utils.Stopwatch()
info = f"{docker.name}:{docker_digests[docker.name]}"
log_file = f"{Settings.OUTPUT_DIR}/docker_{Utils.normalize_string(docker.name)}.log"
files = []
code, out, err = Shell.get_res_stdout_stderr(
f"docker manifest inspect {docker.name}:{docker_digests[docker.name]}"
)
print(
f"Docker inspect results for {docker.name}:{docker_digests[docker.name]}: exit code [{code}], out [{out}], err [{err}]"
)
if "no such manifest" in err:
ret_code = Docker.build(
docker, log_file=log_file, digests=docker_digests, add_latest=False
)
if ret_code == 0:
status = Result.Status.SUCCESS
else:
status = Result.Status.FAILED
job_status = Result.Status.FAILED
info += f", failed with exit code: {ret_code}, see log"
files.append(log_file)
else:
print(
f"Docker image [{docker.name}:{docker_digests[docker.name]} exists - skip build"
)
status = Result.Status.SKIPPED
ready.append(docker.name)
results.append(
Result(
name=docker.name,
status=status,
info=info,
duration=stopwatch.duration,
start_time=stopwatch.start_time,
files=files,
)
)
Result.from_fs(job_name).set_status(job_status).set_results(results).set_info(
job_info
)
if job_status != Result.Status.SUCCESS:
sys.exit(1)
def _config_workflow(workflow: Workflow.Config, job_name):
def _check_yaml_up_to_date():
print("Check workflows are up to date")
stop_watch = Utils.Stopwatch()
exit_code, output, err = Shell.get_res_stdout_stderr(
f"git diff-index HEAD -- {Settings.WORKFLOW_PATH_PREFIX}"
)
info = ""
status = Result.Status.SUCCESS
if exit_code != 0:
info = f"workspace has uncommitted files unexpectedly [{output}]"
status = Result.Status.ERROR
print("ERROR: ", info)
else:
Shell.check(f"{Settings.PYTHON_INTERPRETER} -m praktika --generate")
exit_code, output, err = Shell.get_res_stdout_stderr(
f"git diff-index HEAD -- {Settings.WORKFLOW_PATH_PREFIX}"
)
if exit_code != 0:
info = f"workspace has outdated workflows [{output}] - regenerate with [python -m praktika --generate]"
status = Result.Status.ERROR
print("ERROR: ", info)
return (
Result(
name="Check Workflows updated",
status=status,
start_time=stop_watch.start_time,
duration=stop_watch.duration,
info=info,
),
info,
)
def _check_secrets(secrets):
print("Check Secrets")
stop_watch = Utils.Stopwatch()
infos = []
for secret_config in secrets:
value = secret_config.get_value()
if not value:
info = f"ERROR: Failed to read secret [{secret_config.name}]"
infos.append(info)
print(info)
info = "\n".join(infos)
return (
Result(
name="Check Secrets",
status=(Result.Status.FAILED if infos else Result.Status.SUCCESS),
start_time=stop_watch.start_time,
duration=stop_watch.duration,
info=info,
),
info,
)
def _check_db(workflow):
stop_watch = Utils.Stopwatch()
res, info = CIDB(
workflow.get_secret(Settings.SECRET_CI_DB_URL).get_value(),
workflow.get_secret(Settings.SECRET_CI_DB_PASSWORD).get_value(),
).check()
return (
Result(
name="Check CI DB",
status=(Result.Status.FAILED if not res else Result.Status.SUCCESS),
start_time=stop_watch.start_time,
duration=stop_watch.duration,
info=info,
),
info,
)
print(f"Start [{job_name}], workflow [{workflow.name}]")
results = []
files = []
info_lines = []
job_status = Result.Status.SUCCESS
workflow_config = RunConfig(
name=workflow.name,
digest_jobs={},
digest_dockers={},
sha=_Environment.get().SHA,
cache_success=[],
cache_success_base64=[],
cache_artifacts={},
).dump()
# checks:
result_, info = _check_yaml_up_to_date()
if result_.status != Result.Status.SUCCESS:
print("ERROR: yaml files are outdated - regenerate, commit and push")
job_status = Result.Status.ERROR
info_lines.append(job_name + ": " + info)
results.append(result_)
if workflow.secrets:
result_, info = _check_secrets(workflow.secrets)
if result_.status != Result.Status.SUCCESS:
print(f"ERROR: Invalid secrets in workflow [{workflow.name}]")
job_status = Result.Status.ERROR
info_lines.append(job_name + ": " + info)
results.append(result_)
if workflow.enable_cidb:
result_, info = _check_db(workflow)
if result_.status != Result.Status.SUCCESS:
job_status = Result.Status.ERROR
info_lines.append(job_name + ": " + info)
results.append(result_)
# config:
if workflow.dockers:
print("Calculate docker's digests")
dockers = workflow.dockers
dockers = Docker.sort_in_build_order(dockers)
for docker in dockers:
workflow_config.digest_dockers[docker.name] = Digest().calc_docker_digest(
docker, dockers
)
workflow_config.dump()
if workflow.enable_cache:
print("Cache Lookup")
stop_watch = Utils.Stopwatch()
workflow_config = CacheRunnerHooks.configure(workflow)
results.append(
Result(
name="Cache Lookup",
status=Result.Status.SUCCESS,
start_time=stop_watch.start_time,
duration=stop_watch.duration,
)
)
files.append(RunConfig.file_name_static(workflow.name))
workflow_config.dump()
if workflow.enable_report:
print("Init report")
stop_watch = Utils.Stopwatch()
HtmlRunnerHooks.configure(workflow)
results.append(
Result(
name="Init Report",
status=Result.Status.SUCCESS,
start_time=stop_watch.start_time,
duration=stop_watch.duration,
)
)
files.append(Result.file_name_static(workflow.name))
Result.from_fs(job_name).set_status(job_status).set_results(results).set_files(
files
).set_info("\n".join(info_lines))
if job_status != Result.Status.SUCCESS:
sys.exit(1)
def _finish_workflow(workflow, job_name):
print(f"Start [{job_name}], workflow [{workflow.name}]")
env = _Environment.get()
print("Check Actions statuses")
print(env.get_needs_statuses())
print("Check Workflow results")
S3.copy_result_from_s3(
Result.file_name_static(workflow.name),
lock=False,
)
workflow_result = Result.from_fs(workflow.name)
ready_for_merge_status = Result.Status.SUCCESS
ready_for_merge_description = ""
failed_results = []
update_final_report = False
for result in workflow_result.results:
if result.name == job_name or result.status in (
Result.Status.SUCCESS,
Result.Status.SKIPPED,
):
continue
if not result.is_completed():
print(
f"ERROR: not finished job [{result.name}] in the workflow - set status to error"
)
result.status = Result.Status.ERROR
# dump workflow result after update - to have an updated result in post
workflow_result.dump()
# add error into env - should apper in the report
env.add_info(ResultInfo.NOT_FINALIZED + f" [{result.name}]")
update_final_report = True
job = workflow.get_job(result.name)
if not job or not job.allow_merge_on_failure:
print(
f"NOTE: Result for [{result.name}] has not ok status [{result.status}]"
)
ready_for_merge_status = Result.Status.FAILED
failed_results.append(result.name.split("(", maxsplit=1)[0]) # cut name
if failed_results:
ready_for_merge_description = f"failed: {', '.join(failed_results)}"
if not GH.post_commit_status(
name=Settings.READY_FOR_MERGE_STATUS_NAME + f" [{workflow.name}]",
status=ready_for_merge_status,
description=ready_for_merge_description,
url="",
):
print(f"ERROR: failed to set status [{Settings.READY_FOR_MERGE_STATUS_NAME}]")
env.add_info(ResultInfo.GH_STATUS_ERROR)
if update_final_report:
S3.copy_result_to_s3(
workflow_result,
unlock=False,
) # no lock - no unlock
Result.from_fs(job_name).set_status(Result.Status.SUCCESS).set_info(
ready_for_merge_description
)
if __name__ == "__main__":
job_name = sys.argv[1]
assert job_name, "Job name must be provided as input argument"
workflow = _get_workflows(name=_Environment.get().WORKFLOW_NAME)[0]
if job_name == Settings.DOCKER_BUILD_JOB_NAME:
_build_dockers(workflow, job_name)
elif job_name == Settings.CI_CONFIG_JOB_NAME:
_config_workflow(workflow, job_name)
elif job_name == Settings.FINISH_WORKFLOW_JOB_NAME:
_finish_workflow(workflow, job_name)
else:
assert False, f"BUG, job name [{job_name}]"

258
ci/praktika/parser.py Normal file
View File

@ -0,0 +1,258 @@
import dataclasses
from typing import Any, Dict, List
from praktika import Artifact, Workflow
from praktika.mangle import _get_workflows
class AddonType:
PY = "py"
@dataclasses.dataclass
class WorkflowYaml:
@dataclasses.dataclass
class JobYaml:
name: str
needs: List[str]
runs_on: List[str]
artifacts_gh_requires: List["WorkflowYaml.ArtifactYaml"]
artifacts_gh_provides: List["WorkflowYaml.ArtifactYaml"]
addons: List["WorkflowYaml.JobAddonYaml"]
gh_app_auth: bool
run_unless_cancelled: bool
parameter: Any
def __repr__(self):
return self.name
@dataclasses.dataclass
class ArtifactYaml:
name: str
provided_by: str
required_by: List[str]
path: str
type: str
def __repr__(self):
return self.name
@dataclasses.dataclass
class JobAddonYaml:
install_python: bool
requirements_txt_path: str
name: str
event: str
branches: List[str]
jobs: List[JobYaml]
job_to_config: Dict[str, JobYaml]
artifact_to_config: Dict[str, ArtifactYaml]
secret_names_gh: List[str]
enable_cache: bool
class WorkflowConfigParser:
def __init__(self, config: Workflow.Config):
self.workflow_name = config.name
self.config = config
self.requires_all = [] # type: List[str]
self.provides_all = [] # type: List[str]
self.job_names_all = [] # type: List[str]
self.artifact_to_providing_job_map = {} # type: Dict[str, List[str]]
self.artifact_to_job_requires_map = {} # type: Dict[str, List[str]]
self.artifact_map = {} # type: Dict[str, List[Artifact.Config]]
self.job_to_provides_artifacts = {} # type: Dict[str, List[Artifact.Config]]
self.job_to_requires_artifacts = {} # type: Dict[str, List[Artifact.Config]]
self.workflow_yaml_config = WorkflowYaml(
name=self.workflow_name,
event=config.event,
branches=[],
jobs=[],
secret_names_gh=[],
job_to_config={},
artifact_to_config={},
enable_cache=False,
)
def parse(self):
self.workflow_yaml_config.enable_cache = self.config.enable_cache
# populate WorkflowYaml.branches
if self.config.event in (Workflow.Event.PUSH,):
assert (
self.config.branches
), f'Workflow.Config.branches (e.g. ["main"]) must be set for workflow with event [{self.config.event}], workflow [{self.workflow_name}]'
assert (
not self.config.base_branches
), f'Workflow.Config.base_branches (e.g. ["main"]) must not be set for workflow with event [{self.config.event}], workflow [{self.workflow_name}]'
assert isinstance(
self.config.branches, list
), f'Workflow.Config.branches must be of type list (e.g. ["main"]), workflow [{self.workflow_name}]'
self.workflow_yaml_config.branches = self.config.branches
elif self.config.event in (Workflow.Event.PULL_REQUEST,):
assert (
self.config.base_branches
), f'Workflow.Config.base_branches (e.g. ["main"]) must be set for workflow with event [{self.config.event}], workflow [{self.workflow_name}]'
assert (
not self.config.branches
), f'Workflow.Config.branches (e.g. ["main"]) must not be set for workflow with event [{self.config.event}], workflow [{self.workflow_name}]'
assert isinstance(
self.config.base_branches, list
), f'Workflow.Config.base_branches must be of type list (e.g. ["main"]), workflow [{self.workflow_name}]'
self.workflow_yaml_config.branches = self.config.base_branches
# populate WorkflowYaml.artifact_to_config with phony artifacts
for job in self.config.jobs:
assert (
job.name not in self.workflow_yaml_config.artifact_to_config
), f"Not uniq Job name [{job.name}], workflow [{self.workflow_name}]"
self.workflow_yaml_config.artifact_to_config[job.name] = (
WorkflowYaml.ArtifactYaml(
name=job.name,
provided_by=job.name,
required_by=[],
path="",
type=Artifact.Type.PHONY,
)
)
# populate jobs
for job in self.config.jobs:
job_yaml_config = WorkflowYaml.JobYaml(
name=job.name,
addons=[],
artifacts_gh_requires=[],
artifacts_gh_provides=[],
needs=[],
runs_on=[],
gh_app_auth=False,
run_unless_cancelled=job.run_unless_cancelled,
parameter=None,
)
self.workflow_yaml_config.jobs.append(job_yaml_config)
assert (
job.name not in self.workflow_yaml_config.job_to_config
), f"Job name [{job.name}] is not uniq, workflow [{self.workflow_name}]"
self.workflow_yaml_config.job_to_config[job.name] = job_yaml_config
# populate WorkflowYaml.artifact_to_config
if self.config.artifacts:
for artifact in self.config.artifacts:
assert (
artifact.name not in self.workflow_yaml_config.artifact_to_config
), f"Artifact name [{artifact.name}] is not uniq, workflow [{self.workflow_name}]"
artifact_yaml_config = WorkflowYaml.ArtifactYaml(
name=artifact.name,
provided_by="",
required_by=[],
path=artifact.path,
type=artifact.type,
)
self.workflow_yaml_config.artifact_to_config[artifact.name] = (
artifact_yaml_config
)
# populate ArtifactYaml.provided_by
for job in self.config.jobs:
if job.provides:
for artifact_name in job.provides:
assert (
artifact_name in self.workflow_yaml_config.artifact_to_config
), f"Artifact [{artifact_name}] has no config, job [{job.name}], workflow [{self.workflow_name}]"
assert not self.workflow_yaml_config.artifact_to_config[
artifact_name
].provided_by, f"Artifact [{artifact_name}] provided by multiple jobs [{self.workflow_yaml_config.artifact_to_config[artifact_name].provided_by}] and [{job.name}]"
self.workflow_yaml_config.artifact_to_config[
artifact_name
].provided_by = job.name
# populate ArtifactYaml.required_by
for job in self.config.jobs:
if job.requires:
for artifact_name in job.requires:
assert (
artifact_name in self.workflow_yaml_config.artifact_to_config
), f"Artifact [{artifact_name}] has no config, job [{job.name}], workflow [{self.workflow_name}]"
assert self.workflow_yaml_config.artifact_to_config[
artifact_name
].provided_by, f"Artifact [{artifact_name}] has no job providing it, required by job [{job.name}], workflow [{self.workflow_name}]"
self.workflow_yaml_config.artifact_to_config[
artifact_name
].required_by.append(job.name)
# populate JobYaml.addons
for job in self.config.jobs:
if job.job_requirements:
addon_yaml = WorkflowYaml.JobAddonYaml(
requirements_txt_path=job.job_requirements.python_requirements_txt,
install_python=job.job_requirements.python,
)
self.workflow_yaml_config.job_to_config[job.name].addons.append(
addon_yaml
)
if self.config.enable_report:
for job in self.config.jobs:
# auth required for every job with enabled HTML, so that workflow summary status can be updated
self.workflow_yaml_config.job_to_config[job.name].gh_app_auth = True
# populate JobYaml.runs_on
for job in self.config.jobs:
self.workflow_yaml_config.job_to_config[job.name].runs_on = job.runs_on
# populate JobYaml.artifacts_gh_requires, JobYaml.artifacts_gh_provides and JobYaml.needs
for (
artifact_name,
artifact,
) in self.workflow_yaml_config.artifact_to_config.items():
# assert (
# artifact.provided_by
# and artifact.provided_by in self.workflow_yaml_config.job_to_config
# ), f"Artifact [{artifact_name}] has no valid job providing it [{artifact.provided_by}]"
for job_name in artifact.required_by:
if (
artifact.provided_by
not in self.workflow_yaml_config.job_to_config[job_name].needs
):
self.workflow_yaml_config.job_to_config[job_name].needs.append(
artifact.provided_by
)
if artifact.type in (Artifact.Type.GH,):
self.workflow_yaml_config.job_to_config[
job_name
].artifacts_gh_requires.append(artifact)
elif artifact.type in (Artifact.Type.PHONY, Artifact.Type.S3):
pass
else:
assert (
False
), f"Artifact [{artifact_name}] has unsupported type [{artifact.type}]"
if not artifact.required_by and artifact.type != Artifact.Type.PHONY:
print(
f"WARNING: Artifact [{artifact_name}] provided by job [{artifact.provided_by}] not required by any job in workflow [{self.workflow_name}]"
)
if artifact.type == Artifact.Type.GH:
self.workflow_yaml_config.job_to_config[
artifact.provided_by
].artifacts_gh_provides.append(artifact)
# populate JobYaml.parametrize
for job in self.config.jobs:
self.workflow_yaml_config.job_to_config[job.name].parameter = job.parameter
# populate secrets
for secret_config in self.config.secrets:
if secret_config.is_gh():
self.workflow_yaml_config.secret_names_gh.append(secret_config.name)
return self
if __name__ == "__main__":
# test
workflows = _get_workflows()
for workflow in workflows:
WorkflowConfigParser(workflow).parse()

354
ci/praktika/result.py Normal file
View File

@ -0,0 +1,354 @@
import dataclasses
import datetime
import sys
from collections.abc import Container
from pathlib import Path
from typing import Any, Dict, List, Optional
from praktika._environment import _Environment
from praktika._settings import _Settings
from praktika.utils import ContextManager, MetaClasses, Shell, Utils
@dataclasses.dataclass
class Result(MetaClasses.Serializable):
"""
Represents the outcome of a workflow/job/task or any operation, along with associated metadata.
This class supports nesting of results to represent tasks with sub-tasks, and includes
various attributes to track status, timing, files, and links.
Attributes:
name (str): The name of the task.
status (str): The current status of the task. Should be one of the values defined in the Status class.
start_time (Optional[float]): The start time of the task in Unix timestamp format. None if not started.
duration (Optional[float]): The duration of the task in seconds. None if not completed.
results (List[Result]): A list of sub-results representing nested tasks.
files (List[str]): A list of file paths or names related to the result.
links (List[str]): A list of URLs related to the result (e.g., links to reports or resources).
info (str): Additional information about the result. Free-form text.
# TODO: rename
aux_links (List[str]): A list of auxiliary links that provide additional context for the result.
# TODO: remove
html_link (str): A direct link to an HTML representation of the result (e.g., a detailed report page).
Inner Class:
Status: Defines possible statuses for the task, such as "success", "failure", etc.
"""
class Status:
SKIPPED = "skipped"
SUCCESS = "success"
FAILED = "failure"
PENDING = "pending"
RUNNING = "running"
ERROR = "error"
name: str
status: str
start_time: Optional[float] = None
duration: Optional[float] = None
results: List["Result"] = dataclasses.field(default_factory=list)
files: List[str] = dataclasses.field(default_factory=list)
links: List[str] = dataclasses.field(default_factory=list)
info: str = ""
aux_links: List[str] = dataclasses.field(default_factory=list)
html_link: str = ""
@staticmethod
def create_from(
name="",
results: List["Result"] = None,
stopwatch: Utils.Stopwatch = None,
status="",
files=None,
info="",
with_info_from_results=True,
):
if isinstance(status, bool):
status = Result.Status.SUCCESS if status else Result.Status.FAILED
if not results and not status:
print("ERROR: Either .results or .status must be provided")
raise
if not name:
name = _Environment.get().JOB_NAME
if not name:
print("ERROR: Failed to guess the .name")
raise
result_status = status or Result.Status.SUCCESS
infos = []
if info:
if isinstance(info, Container):
infos += info
else:
infos.append(info)
if results and not status:
for result in results:
if result.status not in (Result.Status.SUCCESS, Result.Status.FAILED):
Utils.raise_with_error(
f"Unexpected result status [{result.status}] for Result.create_from call"
)
if result.status != Result.Status.SUCCESS:
result_status = Result.Status.FAILED
if results:
for result in results:
if result.info and with_info_from_results:
infos.append(f"{result.name}: {result.info}")
return Result(
name=name,
status=result_status,
start_time=stopwatch.start_time if stopwatch else None,
duration=stopwatch.duration if stopwatch else None,
info="\n".join(infos) if infos else "",
results=results or [],
files=files or [],
)
@staticmethod
def get():
return Result.from_fs(_Environment.get().JOB_NAME)
def is_completed(self):
return self.status not in (Result.Status.PENDING, Result.Status.RUNNING)
def is_running(self):
return self.status not in (Result.Status.RUNNING,)
def is_ok(self):
return self.status in (Result.Status.SKIPPED, Result.Status.SUCCESS)
def set_status(self, status) -> "Result":
self.status = status
self.dump()
return self
def set_success(self) -> "Result":
return self.set_status(Result.Status.SUCCESS)
def set_results(self, results: List["Result"]) -> "Result":
self.results = results
self.dump()
return self
def set_files(self, files) -> "Result":
for file in files:
assert Path(
file
).is_file(), f"Not valid file [{file}] from file list [{files}]"
if not self.files:
self.files = []
self.files += files
self.dump()
return self
def set_info(self, info: str) -> "Result":
if self.info:
self.info += "\n"
self.info += info
self.dump()
return self
def set_link(self, link) -> "Result":
self.links.append(link)
self.dump()
return self
@classmethod
def file_name_static(cls, name):
return f"{_Settings.TEMP_DIR}/result_{Utils.normalize_string(name)}.json"
@classmethod
def from_dict(cls, obj: Dict[str, Any]) -> "Result":
sub_results = []
for result_dict in obj["results"] or []:
sub_res = cls.from_dict(result_dict)
sub_results.append(sub_res)
obj["results"] = sub_results
return Result(**obj)
def update_duration(self):
if not self.duration and self.start_time:
self.duration = datetime.datetime.utcnow().timestamp() - self.start_time
else:
if not self.duration:
print(
f"NOTE: duration is set for job [{self.name}] Result - do not update by CI"
)
else:
print(
f"NOTE: start_time is not set for job [{self.name}] Result - do not update duration"
)
return self
def update_sub_result(self, result: "Result"):
assert self.results, "BUG?"
for i, result_ in enumerate(self.results):
if result_.name == result.name:
self.results[i] = result
self._update_status()
return self
def _update_status(self):
was_pending = False
was_running = False
if self.status == self.Status.PENDING:
was_pending = True
if self.status == self.Status.RUNNING:
was_running = True
has_pending, has_running, has_failed = False, False, False
for result_ in self.results:
if result_.status in (self.Status.RUNNING,):
has_running = True
if result_.status in (self.Status.PENDING,):
has_pending = True
if result_.status in (self.Status.ERROR, self.Status.FAILED):
has_failed = True
if has_running:
self.status = self.Status.RUNNING
elif has_pending:
self.status = self.Status.PENDING
elif has_failed:
self.status = self.Status.FAILED
else:
self.status = self.Status.SUCCESS
if (was_pending or was_running) and self.status not in (
self.Status.PENDING,
self.Status.RUNNING,
):
print("Pipeline finished")
self.update_duration()
@classmethod
def generate_pending(cls, name, results=None):
return Result(
name=name,
status=Result.Status.PENDING,
start_time=None,
duration=None,
results=results or [],
files=[],
links=[],
info="",
)
@classmethod
def generate_skipped(cls, name, results=None):
return Result(
name=name,
status=Result.Status.SKIPPED,
start_time=None,
duration=None,
results=results or [],
files=[],
links=[],
info="from cache",
)
@classmethod
def create_from_command_execution(
cls,
name,
command,
with_log=False,
fail_fast=True,
workdir=None,
command_args=None,
command_kwargs=None,
):
"""
Executes shell commands or Python callables, optionally logging output, and handles errors.
:param name: Check name
:param command: Shell command (str) or Python callable, or list of them.
:param workdir: Optional working directory.
:param with_log: Boolean flag to log output to a file.
:param fail_fast: Boolean flag to stop execution if one command fails.
:param command_args: Positional arguments for the callable command.
:param command_kwargs: Keyword arguments for the callable command.
:return: Result object with status and optional log file.
"""
# Stopwatch to track execution time
stop_watch_ = Utils.Stopwatch()
command_args = command_args or []
command_kwargs = command_kwargs or {}
# Set log file path if logging is enabled
log_file = (
f"{_Settings.TEMP_DIR}/{Utils.normalize_string(name)}.log"
if with_log
else None
)
# Ensure the command is a list for consistent iteration
if not isinstance(command, list):
fail_fast = False
command = [command]
print(f"> Starting execution for [{name}]")
res = True # Track success/failure status
error_infos = []
for command_ in command:
if callable(command_):
# If command is a Python function, call it with provided arguments
result = command_(*command_args, **command_kwargs)
if isinstance(result, bool):
res = result
elif result:
error_infos.append(str(result))
res = False
else:
# Run shell command in a specified directory with logging and verbosity
with ContextManager.cd(workdir):
exit_code = Shell.run(command_, verbose=True, log_file=log_file)
res = exit_code == 0
# If fail_fast is enabled, stop on first failure
if not res and fail_fast:
print(f"Execution stopped due to failure in [{command_}]")
break
# Create and return the result object with status and log file (if any)
return Result.create_from(
name=name,
status=res,
stopwatch=stop_watch_,
info=error_infos,
files=[log_file] if log_file else None,
)
def finish_job_accordingly(self):
self.dump()
if not self.is_ok():
print("ERROR: Job Failed")
for result in self.results:
if not result.is_ok():
print("Failed checks:")
print(" | ", result)
sys.exit(1)
else:
print("ok")
class ResultInfo:
SETUP_ENV_JOB_FAILED = (
"Failed to set up job env, it's praktika bug or misconfiguration"
)
PRE_JOB_FAILED = (
"Failed to do a job pre-run step, it's praktika bug or misconfiguration"
)
KILLED = "Job killed or terminated, no Result provided"
NOT_FOUND_IMPOSSIBLE = (
"No Result file (bug, or job misbehaviour, must not ever happen)"
)
SKIPPED_DUE_TO_PREVIOUS_FAILURE = "Skipped due to previous failure"
TIMEOUT = "Timeout"
GH_STATUS_ERROR = "Failed to set GH commit status"
NOT_FINALIZED = (
"Job did not not provide Result: job script bug, died CI runner or praktika bug"
)
S3_ERROR = "S3 call failure"

348
ci/praktika/runner.py Normal file
View File

@ -0,0 +1,348 @@
import os
import re
import sys
import traceback
from pathlib import Path
from praktika._environment import _Environment
from praktika.artifact import Artifact
from praktika.cidb import CIDB
from praktika.digest import Digest
from praktika.hook_cache import CacheRunnerHooks
from praktika.hook_html import HtmlRunnerHooks
from praktika.result import Result, ResultInfo
from praktika.runtime import RunConfig
from praktika.s3 import S3
from praktika.settings import Settings
from praktika.utils import Shell, TeePopen, Utils
class Runner:
@staticmethod
def generate_dummy_environment(workflow, job):
print("WARNING: Generate dummy env for local test")
Shell.check(
f"mkdir -p {Settings.TEMP_DIR} {Settings.INPUT_DIR} {Settings.OUTPUT_DIR}"
)
_Environment(
WORKFLOW_NAME=workflow.name,
JOB_NAME=job.name,
REPOSITORY="",
BRANCH="",
SHA="",
PR_NUMBER=-1,
EVENT_TYPE="",
JOB_OUTPUT_STREAM="",
EVENT_FILE_PATH="",
CHANGE_URL="",
COMMIT_URL="",
BASE_BRANCH="",
RUN_URL="",
RUN_ID="",
INSTANCE_ID="",
INSTANCE_TYPE="",
INSTANCE_LIFE_CYCLE="",
LOCAL_RUN=True,
).dump()
workflow_config = RunConfig(
name=workflow.name,
digest_jobs={},
digest_dockers={},
sha="",
cache_success=[],
cache_success_base64=[],
cache_artifacts={},
)
for docker in workflow.dockers:
workflow_config.digest_dockers[docker.name] = Digest().calc_docker_digest(
docker, workflow.dockers
)
workflow_config.dump()
Result.generate_pending(job.name).dump()
def _setup_env(self, _workflow, job):
# source env file to write data into fs (workflow config json, workflow status json)
Shell.check(f". {Settings.ENV_SETUP_SCRIPT}", verbose=True, strict=True)
# parse the same env script and apply envs from python so that this process sees them
with open(Settings.ENV_SETUP_SCRIPT, "r") as f:
content = f.read()
export_pattern = re.compile(
r"export (\w+)=\$\(cat<<\'EOF\'\n(.*?)EOF\n\)", re.DOTALL
)
matches = export_pattern.findall(content)
for key, value in matches:
value = value.strip()
os.environ[key] = value
print(f"Set environment variable {key}.")
print("Read GH Environment")
env = _Environment.from_env()
env.JOB_NAME = job.name
env.PARAMETER = job.parameter
env.dump()
print(env)
return 0
def _pre_run(self, workflow, job):
env = _Environment.get()
result = Result(
name=job.name,
status=Result.Status.RUNNING,
start_time=Utils.timestamp(),
)
result.dump()
if workflow.enable_report and job.name != Settings.CI_CONFIG_JOB_NAME:
print("Update Job and Workflow Report")
HtmlRunnerHooks.pre_run(workflow, job)
print("Download required artifacts")
required_artifacts = []
if job.requires and workflow.artifacts:
for requires_artifact_name in job.requires:
for artifact in workflow.artifacts:
if (
artifact.name == requires_artifact_name
and artifact.type == Artifact.Type.S3
):
required_artifacts.append(artifact)
print(f"--- Job requires s3 artifacts [{required_artifacts}]")
if workflow.enable_cache:
prefixes = CacheRunnerHooks.pre_run(
_job=job, _workflow=workflow, _required_artifacts=required_artifacts
)
else:
prefixes = [env.get_s3_prefix()] * len(required_artifacts)
for artifact, prefix in zip(required_artifacts, prefixes):
s3_path = f"{Settings.S3_ARTIFACT_PATH}/{prefix}/{Utils.normalize_string(artifact._provided_by)}/{Path(artifact.path).name}"
assert S3.copy_file_from_s3(s3_path=s3_path, local_path=Settings.INPUT_DIR)
return 0
def _run(self, workflow, job, docker="", no_docker=False, param=None):
if param:
if not isinstance(param, str):
Utils.raise_with_error(
f"Custom param for local tests must be of type str, got [{type(param)}]"
)
env = _Environment.get()
env.dump()
if job.run_in_docker and not no_docker:
# TODO: add support for any image, including not from ci config (e.g. ubuntu:latest)
docker_tag = RunConfig.from_fs(workflow.name).digest_dockers[
job.run_in_docker
]
docker = docker or f"{job.run_in_docker}:{docker_tag}"
cmd = f"docker run --rm --user \"$(id -u):$(id -g)\" -e PYTHONPATH='{Settings.DOCKER_WD}:{Settings.DOCKER_WD}/ci' --volume ./:{Settings.DOCKER_WD} --volume {Settings.TEMP_DIR}:{Settings.TEMP_DIR} --workdir={Settings.DOCKER_WD} {docker} {job.command}"
else:
cmd = job.command
if param:
print(f"Custom --param [{param}] will be passed to job's script")
cmd += f" --param {param}"
print(f"--- Run command [{cmd}]")
with TeePopen(cmd, timeout=job.timeout) as process:
exit_code = process.wait()
result = Result.from_fs(job.name)
if exit_code != 0:
if not result.is_completed():
if process.timeout_exceeded:
print(
f"WARNING: Job timed out: [{job.name}], timeout [{job.timeout}], exit code [{exit_code}]"
)
result.set_status(Result.Status.ERROR).set_info(
ResultInfo.TIMEOUT
)
elif result.is_running():
info = f"ERROR: Job terminated with an error, exit code [{exit_code}] - set status to [{Result.Status.ERROR}]"
print(info)
result.set_status(Result.Status.ERROR).set_info(info)
else:
info = f"ERROR: Invalid status [{result.status}] for exit code [{exit_code}] - switch to [{Result.Status.ERROR}]"
print(info)
result.set_status(Result.Status.ERROR).set_info(info)
result.dump()
return exit_code
def _post_run(
self, workflow, job, setup_env_exit_code, prerun_exit_code, run_exit_code
):
info_errors = []
env = _Environment.get()
result_exist = Result.exist(job.name)
if setup_env_exit_code != 0:
info = f"ERROR: {ResultInfo.SETUP_ENV_JOB_FAILED}"
print(info)
# set Result with error and logs
Result(
name=job.name,
status=Result.Status.ERROR,
start_time=Utils.timestamp(),
duration=0.0,
info=info,
).dump()
elif prerun_exit_code != 0:
info = f"ERROR: {ResultInfo.PRE_JOB_FAILED}"
print(info)
# set Result with error and logs
Result(
name=job.name,
status=Result.Status.ERROR,
start_time=Utils.timestamp(),
duration=0.0,
info=info,
).dump()
elif not result_exist:
info = f"ERROR: {ResultInfo.NOT_FOUND_IMPOSSIBLE}"
print(info)
Result(
name=job.name,
start_time=Utils.timestamp(),
duration=None,
status=Result.Status.ERROR,
info=ResultInfo.NOT_FOUND_IMPOSSIBLE,
).dump()
result = Result.from_fs(job.name)
if not result.is_completed():
info = f"ERROR: {ResultInfo.KILLED}"
print(info)
result.set_info(info).set_status(Result.Status.ERROR).dump()
result.set_files(files=[Settings.RUN_LOG])
result.update_duration().dump()
if result.info and result.status != Result.Status.SUCCESS:
# provide job info to workflow level
info_errors.append(result.info)
if run_exit_code == 0:
providing_artifacts = []
if job.provides and workflow.artifacts:
for provides_artifact_name in job.provides:
for artifact in workflow.artifacts:
if (
artifact.name == provides_artifact_name
and artifact.type == Artifact.Type.S3
):
providing_artifacts.append(artifact)
if providing_artifacts:
print(f"Job provides s3 artifacts [{providing_artifacts}]")
for artifact in providing_artifacts:
try:
assert Shell.check(
f"ls -l {artifact.path}", verbose=True
), f"Artifact {artifact.path} not found"
s3_path = f"{Settings.S3_ARTIFACT_PATH}/{env.get_s3_prefix()}/{Utils.normalize_string(env.JOB_NAME)}"
link = S3.copy_file_to_s3(
s3_path=s3_path, local_path=artifact.path
)
result.set_link(link)
except Exception as e:
error = (
f"ERROR: Failed to upload artifact [{artifact}], ex [{e}]"
)
print(error)
info_errors.append(error)
result.set_status(Result.Status.ERROR)
if workflow.enable_cidb:
print("Insert results to CIDB")
try:
CIDB(
url=workflow.get_secret(Settings.SECRET_CI_DB_URL).get_value(),
passwd=workflow.get_secret(
Settings.SECRET_CI_DB_PASSWORD
).get_value(),
).insert(result)
except Exception as ex:
error = f"ERROR: Failed to insert data into CI DB, exception [{ex}]"
print(error)
info_errors.append(error)
result.dump()
# always in the end
if workflow.enable_cache:
print(f"Run CI cache hook")
if result.is_ok():
CacheRunnerHooks.post_run(workflow, job)
if workflow.enable_report:
print(f"Run html report hook")
HtmlRunnerHooks.post_run(workflow, job, info_errors)
return True
def run(
self, workflow, job, docker="", dummy_env=False, no_docker=False, param=None
):
res = True
setup_env_code = -10
prerun_code = -10
run_code = -10
if res and not dummy_env:
print(
f"\n\n=== Setup env script [{job.name}], workflow [{workflow.name}] ==="
)
try:
setup_env_code = self._setup_env(workflow, job)
# Source the bash script and capture the environment variables
res = setup_env_code == 0
if not res:
print(
f"ERROR: Setup env script failed with exit code [{setup_env_code}]"
)
except Exception as e:
print(f"ERROR: Setup env script failed with exception [{e}]")
traceback.print_exc()
print(f"=== Setup env finished ===\n\n")
else:
self.generate_dummy_environment(workflow, job)
if res and not dummy_env:
res = False
print(f"=== Pre run script [{job.name}], workflow [{workflow.name}] ===")
try:
prerun_code = self._pre_run(workflow, job)
res = prerun_code == 0
if not res:
print(f"ERROR: Pre-run failed with exit code [{prerun_code}]")
except Exception as e:
print(f"ERROR: Pre-run script failed with exception [{e}]")
traceback.print_exc()
print(f"=== Pre run finished ===\n\n")
if res:
res = False
print(f"=== Run script [{job.name}], workflow [{workflow.name}] ===")
try:
run_code = self._run(
workflow, job, docker=docker, no_docker=no_docker, param=param
)
res = run_code == 0
if not res:
print(f"ERROR: Run failed with exit code [{run_code}]")
except Exception as e:
print(f"ERROR: Run script failed with exception [{e}]")
traceback.print_exc()
print(f"=== Run scrip finished ===\n\n")
if not dummy_env:
print(f"=== Post run script [{job.name}], workflow [{workflow.name}] ===")
self._post_run(workflow, job, setup_env_code, prerun_code, run_code)
print(f"=== Post run scrip finished ===")
if not res:
sys.exit(1)

35
ci/praktika/runtime.py Normal file
View File

@ -0,0 +1,35 @@
from dataclasses import dataclass
from typing import Dict, List
from praktika.cache import Cache
from praktika.settings import Settings
from praktika.utils import MetaClasses, Utils
@dataclass
class RunConfig(MetaClasses.Serializable):
name: str
digest_jobs: Dict[str, str]
digest_dockers: Dict[str, str]
cache_success: List[str]
# there are might be issue with special characters in job names if used directly in yaml syntax - create base64 encoded list to avoid this
cache_success_base64: List[str]
cache_artifacts: Dict[str, Cache.CacheRecord]
sha: str
@classmethod
def from_dict(cls, obj):
cache_artifacts = obj["cache_artifacts"]
cache_artifacts_deserialized = {}
for artifact_name, cache_artifact in cache_artifacts.items():
cache_artifacts_deserialized[artifact_name] = Cache.CacheRecord.from_dict(
cache_artifact
)
obj["cache_artifacts"] = cache_artifacts_deserialized
return RunConfig(**obj)
@classmethod
def file_name_static(cls, name):
return (
f"{Settings.TEMP_DIR}/workflow_config_{Utils.normalize_string(name)}.json"
)

295
ci/praktika/s3.py Normal file
View File

@ -0,0 +1,295 @@
import dataclasses
import json
import time
from pathlib import Path
from typing import Dict
from praktika._environment import _Environment
from praktika.settings import Settings
from praktika.utils import Shell, Utils
class S3:
@dataclasses.dataclass
class Object:
AcceptRanges: str
Expiration: str
LastModified: str
ContentLength: int
ETag: str
ContentType: str
ServerSideEncryption: str
Metadata: Dict
def has_tags(self, tags):
meta = self.Metadata
for k, v in tags.items():
if k not in meta or meta[k] != v:
print(f"tag [{k}={v}] does not match meta [{meta}]")
return False
return True
@classmethod
def clean_s3_directory(cls, s3_path):
assert len(s3_path.split("/")) > 2, "check to not delete too much"
cmd = f"aws s3 rm s3://{s3_path} --recursive"
cls.run_command_with_retries(cmd, retries=1)
return
@classmethod
def copy_file_to_s3(cls, s3_path, local_path, text=False):
assert Path(local_path).exists(), f"Path [{local_path}] does not exist"
assert Path(s3_path), f"Invalid S3 Path [{s3_path}]"
assert Path(
local_path
).is_file(), f"Path [{local_path}] is not file. Only files are supported"
file_name = Path(local_path).name
s3_full_path = s3_path
if not s3_full_path.endswith(file_name):
s3_full_path = f"{s3_path}/{Path(local_path).name}"
cmd = f"aws s3 cp {local_path} s3://{s3_full_path}"
if text:
cmd += " --content-type text/plain"
res = cls.run_command_with_retries(cmd)
if not res:
raise
bucket = s3_path.split("/")[0]
endpoint = Settings.S3_BUCKET_TO_HTTP_ENDPOINT[bucket]
assert endpoint
return f"https://{s3_full_path}".replace(bucket, endpoint)
@classmethod
def put(cls, s3_path, local_path, text=False, metadata=None):
assert Path(local_path).exists(), f"Path [{local_path}] does not exist"
assert Path(s3_path), f"Invalid S3 Path [{s3_path}]"
assert Path(
local_path
).is_file(), f"Path [{local_path}] is not file. Only files are supported"
file_name = Path(local_path).name
s3_full_path = s3_path
if not s3_full_path.endswith(file_name):
s3_full_path = f"{s3_path}/{Path(local_path).name}"
s3_full_path = str(s3_full_path).removeprefix("s3://")
bucket, key = s3_full_path.split("/", maxsplit=1)
command = (
f"aws s3api put-object --bucket {bucket} --key {key} --body {local_path}"
)
if metadata:
for k, v in metadata.items():
command += f" --metadata {k}={v}"
cmd = f"aws s3 cp {local_path} s3://{s3_full_path}"
if text:
cmd += " --content-type text/plain"
res = cls.run_command_with_retries(command)
assert res
@classmethod
def run_command_with_retries(cls, command, retries=Settings.MAX_RETRIES_S3):
i = 0
res = False
while not res and i < retries:
i += 1
ret_code, stdout, stderr = Shell.get_res_stdout_stderr(
command, verbose=True
)
if "aws sso login" in stderr:
print("ERROR: aws login expired")
break
elif "does not exist" in stderr:
print("ERROR: requested file does not exist")
break
if ret_code != 0:
print(
f"ERROR: aws s3 cp failed, stdout/stderr err: [{stderr}], out [{stdout}]"
)
res = ret_code == 0
return res
@classmethod
def get_link(cls, s3_path, local_path):
s3_full_path = f"{s3_path}/{Path(local_path).name}"
bucket = s3_path.split("/")[0]
endpoint = Settings.S3_BUCKET_TO_HTTP_ENDPOINT[bucket]
return f"https://{s3_full_path}".replace(bucket, endpoint)
@classmethod
def copy_file_from_s3(cls, s3_path, local_path):
assert Path(s3_path), f"Invalid S3 Path [{s3_path}]"
if Path(local_path).is_dir():
local_path = Path(local_path) / Path(s3_path).name
else:
assert Path(
local_path
).parent.is_dir(), f"Parent path for [{local_path}] does not exist"
cmd = f"aws s3 cp s3://{s3_path} {local_path}"
res = cls.run_command_with_retries(cmd)
return res
@classmethod
def head_object(cls, s3_path):
s3_path = str(s3_path).removeprefix("s3://")
bucket, key = s3_path.split("/", maxsplit=1)
output = Shell.get_output(
f"aws s3api head-object --bucket {bucket} --key {key}", verbose=True
)
if not output:
return None
else:
return cls.Object(**json.loads(output))
@classmethod
def delete(cls, s3_path):
assert Path(s3_path), f"Invalid S3 Path [{s3_path}]"
return Shell.check(
f"aws s3 rm s3://{s3_path}",
verbose=True,
)
# TODO: apparently should be placed into separate file to be used only inside praktika
# keeping this module clean from importing Settings, Environment and etc, making it easy for use externally
@classmethod
def copy_result_to_s3(cls, result, unlock=True):
result.dump()
env = _Environment.get()
s3_path = f"{Settings.HTML_S3_PATH}/{env.get_s3_prefix()}"
s3_path_full = f"{s3_path}/{Path(result.file_name()).name}"
url = S3.copy_file_to_s3(s3_path=s3_path, local_path=result.file_name())
if env.PR_NUMBER:
print("Duplicate Result for latest commit alias in PR")
s3_path = f"{Settings.HTML_S3_PATH}/{env.get_s3_prefix(latest=True)}"
url = S3.copy_file_to_s3(s3_path=s3_path, local_path=result.file_name())
if unlock:
if not cls.unlock(s3_path_full):
print(f"ERROR: File [{s3_path_full}] unlock failure")
assert False # TODO: investigate
return url
@classmethod
def copy_result_from_s3(cls, local_path, lock=True):
env = _Environment.get()
file_name = Path(local_path).name
s3_path = f"{Settings.HTML_S3_PATH}/{env.get_s3_prefix()}/{file_name}"
if lock:
cls.lock(s3_path)
if not S3.copy_file_from_s3(s3_path=s3_path, local_path=local_path):
print(f"ERROR: failed to cp file [{s3_path}] from s3")
raise
@classmethod
def lock(cls, s3_path, level=0):
assert level < 3, "Never"
env = _Environment.get()
s3_path_lock = s3_path + f".lock"
file_path_lock = f"{Settings.TEMP_DIR}/{Path(s3_path_lock).name}"
assert Shell.check(
f"echo '''{env.JOB_NAME}''' > {file_path_lock}", verbose=True
), "Never"
i = 20
meta = S3.head_object(s3_path_lock)
while meta:
print(f"WARNING: Failed to acquire lock, meta [{meta}] - wait")
i -= 5
if i < 0:
info = f"ERROR: lock acquire failure - unlock forcefully"
print(info)
env.add_info(info)
break
time.sleep(5)
metadata = {"job": Utils.to_base64(env.JOB_NAME)}
S3.put(
s3_path=s3_path_lock,
local_path=file_path_lock,
metadata=metadata,
)
time.sleep(1)
obj = S3.head_object(s3_path_lock)
if not obj or not obj.has_tags(tags=metadata):
print(f"WARNING: locked by another job [{obj}]")
env.add_info("S3 lock file failure")
cls.lock(s3_path, level=level + 1)
print("INFO: lock acquired")
@classmethod
def unlock(cls, s3_path):
s3_path_lock = s3_path + ".lock"
env = _Environment.get()
obj = S3.head_object(s3_path_lock)
if not obj:
print("ERROR: lock file is removed")
assert False # investigate
elif not obj.has_tags({"job": Utils.to_base64(env.JOB_NAME)}):
print("ERROR: lock file was acquired by another job")
assert False # investigate
if not S3.delete(s3_path_lock):
print(f"ERROR: File [{s3_path_lock}] delete failure")
print("INFO: lock released")
return True
@classmethod
def get_result_link(cls, result):
env = _Environment.get()
s3_path = f"{Settings.HTML_S3_PATH}/{env.get_s3_prefix(latest=True if env.PR_NUMBER else False)}"
return S3.get_link(s3_path=s3_path, local_path=result.file_name())
@classmethod
def clean_latest_result(cls):
env = _Environment.get()
env.SHA = "latest"
assert env.PR_NUMBER
s3_path = f"{Settings.HTML_S3_PATH}/{env.get_s3_prefix()}"
S3.clean_s3_directory(s3_path=s3_path)
@classmethod
def _upload_file_to_s3(
cls, local_file_path, upload_to_s3: bool, text: bool = False, s3_subprefix=""
) -> str:
if upload_to_s3:
env = _Environment.get()
s3_path = f"{Settings.HTML_S3_PATH}/{env.get_s3_prefix()}"
if s3_subprefix:
s3_subprefix.removeprefix("/").removesuffix("/")
s3_path += f"/{s3_subprefix}"
html_link = S3.copy_file_to_s3(
s3_path=s3_path, local_path=local_file_path, text=text
)
return html_link
return f"file://{Path(local_file_path).absolute()}"
@classmethod
def upload_result_files_to_s3(cls, result):
if result.results:
for result_ in result.results:
cls.upload_result_files_to_s3(result_)
for file in result.files:
if not Path(file).is_file():
print(f"ERROR: Invalid file [{file}] in [{result.name}] - skip upload")
result.info += f"\nWARNING: Result file [{file}] was not found"
file_link = cls._upload_file_to_s3(file, upload_to_s3=False)
else:
is_text = False
for text_file_suffix in Settings.TEXT_CONTENT_EXTENSIONS:
if file.endswith(text_file_suffix):
print(
f"File [{file}] matches Settings.TEXT_CONTENT_EXTENSIONS [{Settings.TEXT_CONTENT_EXTENSIONS}] - add text attribute for s3 object"
)
is_text = True
break
file_link = cls._upload_file_to_s3(
file,
upload_to_s3=True,
text=is_text,
s3_subprefix=Utils.normalize_string(result.name),
)
result.links.append(file_link)
if result.files:
print(
f"Result files [{result.files}] uploaded to s3 [{result.links[-len(result.files):]}] - clean files list"
)
result.files = []
result.dump()

61
ci/praktika/secret.py Normal file
View File

@ -0,0 +1,61 @@
import dataclasses
import os
from praktika.utils import Shell
class Secret:
class Type:
AWS_SSM_VAR = "aws parameter"
AWS_SSM_SECRET = "aws secret"
GH_SECRET = "gh secret"
@dataclasses.dataclass
class Config:
name: str
type: str
def is_gh(self):
return self.type == Secret.Type.GH_SECRET
def get_value(self):
if self.type == Secret.Type.AWS_SSM_VAR:
return self.get_aws_ssm_var()
if self.type == Secret.Type.AWS_SSM_SECRET:
return self.get_aws_ssm_secret()
elif self.type == Secret.Type.GH_SECRET:
return self.get_gh_secret()
else:
assert False, f"Not supported secret type, secret [{self}]"
def get_aws_ssm_var(self):
res = Shell.get_output(
f"aws ssm get-parameter --name {self.name} --with-decryption --output text --query Parameter.Value",
)
if not res:
print(f"ERROR: Failed to get secret [{self.name}]")
raise RuntimeError()
return res
def get_aws_ssm_secret(self):
name, secret_key_name = self.name, ""
if "." in self.name:
name, secret_key_name = self.name.split(".")
cmd = f"aws secretsmanager get-secret-value --secret-id {name} --query SecretString --output text"
if secret_key_name:
cmd += f" | jq -r '.[\"{secret_key_name}\"]'"
res = Shell.get_output(cmd, verbose=True)
if not res:
print(f"ERROR: Failed to get secret [{self.name}]")
raise RuntimeError()
return res
def get_gh_secret(self):
res = os.getenv(f"{self.name}")
if not res:
print(f"ERROR: Failed to get secret [{self.name}]")
raise RuntimeError()
return res
def __repr__(self):
return self.name

8
ci/praktika/settings.py Normal file
View File

@ -0,0 +1,8 @@
from praktika._settings import _Settings
from praktika.mangle import _get_user_settings
Settings = _Settings()
user_settings = _get_user_settings()
for setting, value in user_settings.items():
Settings.__setattr__(setting, value)

597
ci/praktika/utils.py Normal file
View File

@ -0,0 +1,597 @@
import base64
import dataclasses
import glob
import json
import multiprocessing
import os
import re
import signal
import subprocess
import sys
import time
from abc import ABC, abstractmethod
from contextlib import contextmanager
from datetime import datetime
from pathlib import Path
from threading import Thread
from types import SimpleNamespace
from typing import Any, Dict, Iterator, List, Optional, Type, TypeVar, Union
from praktika._settings import _Settings
T = TypeVar("T", bound="Serializable")
class MetaClasses:
class WithIter(type):
def __iter__(cls):
return (v for k, v in cls.__dict__.items() if not k.startswith("_"))
@dataclasses.dataclass
class Serializable(ABC):
@classmethod
def to_dict(cls, obj):
if dataclasses.is_dataclass(obj):
return {k: cls.to_dict(v) for k, v in dataclasses.asdict(obj).items()}
elif isinstance(obj, SimpleNamespace):
return {k: cls.to_dict(v) for k, v in vars(obj).items()}
elif isinstance(obj, list):
return [cls.to_dict(i) for i in obj]
elif isinstance(obj, dict):
return {k: cls.to_dict(v) for k, v in obj.items()}
else:
return obj
@classmethod
def from_dict(cls: Type[T], obj: Dict[str, Any]) -> T:
return cls(**obj)
@classmethod
def from_fs(cls: Type[T], name) -> T:
with open(cls.file_name_static(name), "r", encoding="utf8") as f:
try:
return cls.from_dict(json.load(f))
except json.decoder.JSONDecodeError as ex:
print(f"ERROR: failed to parse json, ex [{ex}]")
print(f"JSON content [{cls.file_name_static(name)}]")
Shell.check(f"cat {cls.file_name_static(name)}")
raise ex
@classmethod
@abstractmethod
def file_name_static(cls, name):
pass
def file_name(self):
return self.file_name_static(self.name)
def dump(self):
with open(self.file_name(), "w", encoding="utf8") as f:
json.dump(self.to_dict(self), f, indent=4)
return self
@classmethod
def exist(cls, name):
return Path(cls.file_name_static(name)).is_file()
def to_json(self, pretty=False):
return json.dumps(dataclasses.asdict(self), indent=4 if pretty else None)
class ContextManager:
@staticmethod
@contextmanager
def cd(to: Optional[Union[Path, str]] = None) -> Iterator[None]:
"""
changes current working directory to @path or `git root` if @path is None
:param to:
:return:
"""
if not to:
try:
to = Shell.get_output_or_raise("git rev-parse --show-toplevel")
except:
pass
if not to:
if Path(_Settings.DOCKER_WD).is_dir():
to = _Settings.DOCKER_WD
if not to:
assert False, "FIX IT"
assert to
old_pwd = os.getcwd()
os.chdir(to)
try:
yield
finally:
os.chdir(old_pwd)
class Shell:
@classmethod
def get_output_or_raise(cls, command, verbose=False):
return cls.get_output(command, verbose=verbose, strict=True).strip()
@classmethod
def get_output(cls, command, strict=False, verbose=False):
if verbose:
print(f"Run command [{command}]")
res = subprocess.run(
command,
shell=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
)
if res.stderr:
print(f"WARNING: stderr: {res.stderr.strip()}")
if strict and res.returncode != 0:
raise RuntimeError(f"command failed with {res.returncode}")
return res.stdout.strip()
@classmethod
def get_res_stdout_stderr(cls, command, verbose=True):
if verbose:
print(f"Run command [{command}]")
res = subprocess.run(
command,
shell=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
)
return res.returncode, res.stdout.strip(), res.stderr.strip()
@classmethod
def check(
cls,
command,
log_file=None,
strict=False,
verbose=False,
dry_run=False,
stdin_str=None,
timeout=None,
retries=0,
**kwargs,
):
return (
cls.run(
command,
log_file,
strict,
verbose,
dry_run,
stdin_str,
retries=retries,
timeout=timeout,
**kwargs,
)
== 0
)
@classmethod
def run(
cls,
command,
log_file=None,
strict=False,
verbose=False,
dry_run=False,
stdin_str=None,
timeout=None,
retries=0,
**kwargs,
):
def _check_timeout(timeout, process) -> None:
if not timeout:
return
time.sleep(timeout)
print(
f"WARNING: Timeout exceeded [{timeout}], sending SIGTERM to process group [{process.pid}]"
)
try:
os.killpg(process.pid, signal.SIGTERM)
except ProcessLookupError:
print("Process already terminated.")
return
time_wait = 0
wait_interval = 5
# Wait for process to terminate
while process.poll() is None and time_wait < 100:
print("Waiting for process to exit...")
time.sleep(wait_interval)
time_wait += wait_interval
# Force kill if still running
if process.poll() is None:
print(f"WARNING: Process still running after SIGTERM, sending SIGKILL")
try:
os.killpg(process.pid, signal.SIGKILL)
except ProcessLookupError:
print("Process already terminated.")
# Dry-run
if dry_run:
print(f"Dry-run. Would run command [{command}]")
return 0 # Return success for dry-run
if verbose:
print(f"Run command: [{command}]")
log_file = log_file or "/dev/null"
proc = None
for retry in range(retries + 1):
try:
with open(log_file, "w") as log_fp:
proc = subprocess.Popen(
command,
shell=True,
stderr=subprocess.STDOUT,
stdout=subprocess.PIPE,
stdin=subprocess.PIPE if stdin_str else None,
universal_newlines=True,
start_new_session=True, # Start a new process group for signal handling
bufsize=1, # Line-buffered
errors="backslashreplace",
**kwargs,
)
# Start the timeout thread if specified
if timeout:
t = Thread(target=_check_timeout, args=(timeout, proc))
t.daemon = True
t.start()
# Write stdin if provided
if stdin_str:
proc.stdin.write(stdin_str)
proc.stdin.close()
# Process output in real-time
if proc.stdout:
for line in proc.stdout:
sys.stdout.write(line)
log_fp.write(line)
proc.wait() # Wait for the process to finish
if proc.returncode == 0:
break # Exit retry loop if success
else:
if verbose:
print(
f"ERROR: command [{command}] failed, exit code: {proc.returncode}, retry: {retry}/{retries}"
)
except Exception as e:
if verbose:
print(
f"ERROR: command failed, exception: {e}, retry: {retry}/{retries}"
)
if proc:
proc.kill()
# Handle strict mode (ensure process success or fail)
if strict:
assert (
proc and proc.returncode == 0
), f"Command failed with return code {proc.returncode}"
return proc.returncode if proc else 1 # Return 1 if process never started
@classmethod
def run_async(
cls,
command,
stdin_str=None,
verbose=False,
suppress_output=False,
**kwargs,
):
if verbose:
print(f"Run command in background [{command}]")
proc = subprocess.Popen(
command,
shell=True,
stderr=subprocess.STDOUT if not suppress_output else subprocess.DEVNULL,
stdout=subprocess.PIPE if not suppress_output else subprocess.DEVNULL,
stdin=subprocess.PIPE if stdin_str else None,
universal_newlines=True,
start_new_session=True,
bufsize=1,
errors="backslashreplace",
**kwargs,
)
if proc.stdout:
for line in proc.stdout:
print(line, end="")
return proc
class Utils:
@staticmethod
def terminate_process_group(pid, force=False):
if not force:
os.killpg(os.getpgid(pid), signal.SIGTERM)
else:
os.killpg(os.getpgid(pid), signal.SIGKILL)
@staticmethod
def set_env(key, val):
os.environ[key] = val
@staticmethod
def print_formatted_error(error_message, stdout="", stderr=""):
stdout_lines = stdout.splitlines() if stdout else []
stderr_lines = stderr.splitlines() if stderr else []
print(f"ERROR: {error_message}")
if stdout_lines:
print(" Out:")
for line in stdout_lines:
print(f" | {line}")
if stderr_lines:
print(" Err:")
for line in stderr_lines:
print(f" | {line}")
@staticmethod
def sleep(seconds):
time.sleep(seconds)
@staticmethod
def cwd():
return Path.cwd()
@staticmethod
def cpu_count():
return multiprocessing.cpu_count()
@staticmethod
def raise_with_error(error_message, stdout="", stderr="", ex=None):
Utils.print_formatted_error(error_message, stdout, stderr)
raise ex or RuntimeError()
@staticmethod
def timestamp():
return datetime.utcnow().timestamp()
@staticmethod
def timestamp_to_str(timestamp):
return datetime.utcfromtimestamp(timestamp).strftime("%Y-%m-%d %H:%M:%S")
@staticmethod
def get_failed_tests_number(description: str) -> Optional[int]:
description = description.lower()
pattern = r"fail:\s*(\d+)\s*(?=,|$)"
match = re.search(pattern, description)
if match:
return int(match.group(1))
return None
@staticmethod
def is_killed_with_oom():
if Shell.check(
"sudo dmesg -T | grep -q -e 'Out of memory: Killed process' -e 'oom_reaper: reaped process' -e 'oom-kill:constraint=CONSTRAINT_NONE'"
):
return True
return False
@staticmethod
def clear_dmesg():
Shell.check("sudo dmesg --clear", verbose=True)
@staticmethod
def to_base64(value):
assert isinstance(value, str), f"TODO: not supported for {type(value)}"
string_bytes = value.encode("utf-8")
base64_bytes = base64.b64encode(string_bytes)
base64_string = base64_bytes.decode("utf-8")
return base64_string
@staticmethod
def is_hex(s):
try:
int(s, 16)
return True
except ValueError:
return False
@staticmethod
def normalize_string(string: str) -> str:
res = string.lower()
for r in (
(" ", "_"),
("(", ""),
(")", ""),
("{", ""),
("}", ""),
("'", ""),
("[", ""),
("]", ""),
(",", ""),
("/", "_"),
("-", "_"),
(":", ""),
('"', ""),
):
res = res.replace(*r)
return res
@staticmethod
def traverse_path(path, file_suffixes=None, sorted=False, not_exists_ok=False):
res = []
def is_valid_file(file):
if file_suffixes is None:
return True
return any(file.endswith(suffix) for suffix in file_suffixes)
if os.path.isfile(path):
if is_valid_file(path):
res.append(path)
elif os.path.isdir(path):
for root, dirs, files in os.walk(path):
for file in files:
full_path = os.path.join(root, file)
if is_valid_file(full_path):
res.append(full_path)
elif "*" in str(path):
res.extend(
[
f
for f in glob.glob(path, recursive=True)
if os.path.isfile(f) and is_valid_file(f)
]
)
else:
if not_exists_ok:
pass
else:
assert False, f"File does not exist or not valid [{path}]"
if sorted:
res.sort(reverse=True)
return res
@classmethod
def traverse_paths(
cls,
include_paths,
exclude_paths,
file_suffixes=None,
sorted=False,
not_exists_ok=False,
) -> List["str"]:
included_files_ = set()
for path in include_paths:
included_files_.update(cls.traverse_path(path, file_suffixes=file_suffixes))
excluded_files = set()
for path in exclude_paths:
res = cls.traverse_path(path, not_exists_ok=not_exists_ok)
if not res:
print(
f"WARNING: Utils.traverse_paths excluded 0 files by path [{path}] in exclude_paths"
)
else:
excluded_files.update(res)
res = [f for f in included_files_ if f not in excluded_files]
if sorted:
res.sort(reverse=True)
return res
@classmethod
def add_to_PATH(cls, path):
path_cur = os.getenv("PATH", "")
if path_cur:
path += ":" + path_cur
os.environ["PATH"] = path
class Stopwatch:
def __init__(self):
self.start_time = datetime.utcnow().timestamp()
@property
def duration(self) -> float:
return datetime.utcnow().timestamp() - self.start_time
class TeePopen:
def __init__(
self,
command: str,
log_file: Union[str, Path] = "",
env: Optional[dict] = None,
timeout: Optional[int] = None,
):
self.command = command
self.log_file_name = log_file
self.log_file = None
self.env = env or os.environ.copy()
self.process = None # type: Optional[subprocess.Popen]
self.timeout = timeout
self.timeout_exceeded = False
self.terminated_by_sigterm = False
self.terminated_by_sigkill = False
def _check_timeout(self) -> None:
if self.timeout is None:
return
time.sleep(self.timeout)
print(
f"WARNING: Timeout exceeded [{self.timeout}], send SIGTERM to [{self.process.pid}] and give a chance for graceful termination"
)
self.send_signal(signal.SIGTERM)
time_wait = 0
self.terminated_by_sigterm = True
self.timeout_exceeded = True
while self.process.poll() is None and time_wait < 100:
print("wait...")
wait = 5
time.sleep(wait)
time_wait += wait
while self.process.poll() is None:
print(f"WARNING: Still running, send SIGKILL to [{self.process.pid}]")
self.send_signal(signal.SIGKILL)
self.terminated_by_sigkill = True
time.sleep(2)
def __enter__(self) -> "TeePopen":
if self.log_file_name:
self.log_file = open(self.log_file_name, "w", encoding="utf-8")
self.process = subprocess.Popen(
self.command,
shell=True,
universal_newlines=True,
env=self.env,
start_new_session=True, # signall will be sent to all children
stderr=subprocess.STDOUT,
stdout=subprocess.PIPE,
bufsize=1,
errors="backslashreplace",
)
time.sleep(1)
print(f"Subprocess started, pid [{self.process.pid}]")
if self.timeout is not None and self.timeout > 0:
t = Thread(target=self._check_timeout)
t.daemon = True # does not block the program from exit
t.start()
return self
def __exit__(self, exc_type, exc_value, traceback):
self.wait()
if self.log_file:
self.log_file.close()
def wait(self) -> int:
if self.process.stdout is not None:
for line in self.process.stdout:
sys.stdout.write(line)
if self.log_file:
self.log_file.write(line)
return self.process.wait()
def poll(self):
return self.process.poll()
def send_signal(self, signal_num):
os.killpg(self.process.pid, signal_num)
if __name__ == "__main__":
@dataclasses.dataclass
class Test(MetaClasses.Serializable):
name: str
@staticmethod
def file_name_static(name):
return f"/tmp/{Utils.normalize_string(name)}.json"
Test(name="dsada").dump()
t = Test.from_fs("dsada")
print(t)

208
ci/praktika/validator.py Normal file
View File

@ -0,0 +1,208 @@
import glob
import sys
from itertools import chain
from pathlib import Path
from praktika import Workflow
from praktika._settings import GHRunners
from praktika.mangle import _get_workflows
from praktika.settings import Settings
from praktika.utils import ContextManager
class Validator:
@classmethod
def validate(cls):
print("---Start validating Pipeline and settings---")
workflows = _get_workflows()
for workflow in workflows:
print(f"Validating workflow [{workflow.name}]")
cls.validate_file_paths_in_run_command(workflow)
cls.validate_file_paths_in_digest_configs(workflow)
cls.validate_requirements_txt_files(workflow)
cls.validate_dockers(workflow)
if workflow.artifacts:
for artifact in workflow.artifacts:
if artifact.is_s3_artifact():
assert (
Settings.S3_ARTIFACT_PATH
), "Provide S3_ARTIFACT_PATH setting in any .py file in ./ci/settings/* to be able to use s3 for artifacts"
for job in workflow.jobs:
if job.requires and workflow.artifacts:
for require in job.requires:
if (
require in workflow.artifacts
and workflow.artifacts[require].is_s3_artifact()
):
assert not any(
[r in GHRunners for r in job.runs_on]
), f"GH runners [{job.name}:{job.runs_on}] must not be used with S3 as artifact storage"
if job.allow_merge_on_failure:
assert (
workflow.enable_merge_ready_status
), f"Job property allow_merge_on_failure must be used only with enabled workflow.enable_merge_ready_status, workflow [{workflow.name}], job [{job.name}]"
if workflow.enable_cache:
assert (
Settings.CI_CONFIG_RUNS_ON
), f"Runner label to run workflow config job must be provided via CACHE_CONFIG_RUNS_ON setting if enable_cache=True, workflow [{workflow.name}]"
assert (
Settings.CACHE_S3_PATH
), f"CACHE_S3_PATH Setting must be defined if enable_cache=True, workflow [{workflow.name}]"
if workflow.dockers:
cls.evaluate_check(
Settings.DOCKER_BUILD_RUNS_ON,
f"DOCKER_BUILD_RUNS_ON settings must be defined if workflow has dockers",
workflow_name=workflow.name,
)
if workflow.enable_report:
assert (
Settings.HTML_S3_PATH
), f"HTML_S3_PATH Setting must be defined if enable_html=True, workflow [{workflow.name}]"
assert (
Settings.S3_BUCKET_TO_HTTP_ENDPOINT
), f"S3_BUCKET_TO_HTTP_ENDPOINT Setting must be defined if enable_html=True, workflow [{workflow.name}]"
assert (
Settings.HTML_S3_PATH.split("/")[0]
in Settings.S3_BUCKET_TO_HTTP_ENDPOINT
), f"S3_BUCKET_TO_HTTP_ENDPOINT Setting must include bucket name [{Settings.HTML_S3_PATH}] from HTML_S3_PATH, workflow [{workflow.name}]"
if workflow.enable_cache:
for artifact in workflow.artifacts or []:
assert (
artifact.is_s3_artifact()
), f"All artifacts must be of S3 type if enable_cache|enable_html=True, artifact [{artifact.name}], type [{artifact.type}], workflow [{workflow.name}]"
if workflow.dockers:
assert (
Settings.DOCKERHUB_USERNAME
), f"Settings.DOCKERHUB_USERNAME must be provided if workflow has dockers, workflow [{workflow.name}]"
assert (
Settings.DOCKERHUB_SECRET
), f"Settings.DOCKERHUB_SECRET must be provided if workflow has dockers, workflow [{workflow.name}]"
assert workflow.get_secret(
Settings.DOCKERHUB_SECRET
), f"Secret [{Settings.DOCKERHUB_SECRET}] must have configuration in workflow.secrets, workflow [{workflow.name}]"
if (
workflow.enable_cache
or workflow.enable_report
or workflow.enable_merge_ready_status
):
for job in workflow.jobs:
assert not any(
job in ("ubuntu-latest",) for job in job.runs_on
), f"GitHub Runners must not be used for workflow with enabled: workflow.enable_cache, workflow.enable_html or workflow.enable_merge_ready_status as s3 access is required, workflow [{workflow.name}], job [{job.name}]"
if workflow.enable_cidb:
assert (
Settings.SECRET_CI_DB_URL
), f"Settings.CI_DB_URL_SECRET must be provided if workflow.enable_cidb=True, workflow [{workflow.name}]"
assert (
Settings.SECRET_CI_DB_PASSWORD
), f"Settings.CI_DB_PASSWORD_SECRET must be provided if workflow.enable_cidb=True, workflow [{workflow.name}]"
assert (
Settings.CI_DB_DB_NAME
), f"Settings.CI_DB_DB_NAME must be provided if workflow.enable_cidb=True, workflow [{workflow.name}]"
assert (
Settings.CI_DB_TABLE_NAME
), f"Settings.CI_DB_TABLE_NAME must be provided if workflow.enable_cidb=True, workflow [{workflow.name}]"
@classmethod
def validate_file_paths_in_run_command(cls, workflow: Workflow.Config) -> None:
if not Settings.VALIDATE_FILE_PATHS:
return
with ContextManager.cd():
for job in workflow.jobs:
run_command = job.command
command_parts = run_command.split(" ")
for part in command_parts:
if ">" in part:
return
if "/" in part:
assert (
Path(part).is_file() or Path(part).is_dir()
), f"Apparently run command [{run_command}] for job [{job}] has invalid path [{part}]. Setting to disable check: VALIDATE_FILE_PATHS"
@classmethod
def validate_file_paths_in_digest_configs(cls, workflow: Workflow.Config) -> None:
if not Settings.VALIDATE_FILE_PATHS:
return
with ContextManager.cd():
for job in workflow.jobs:
if not job.digest_config:
continue
for include_path in chain(
job.digest_config.include_paths, job.digest_config.exclude_paths
):
if "*" in include_path:
assert glob.glob(
include_path, recursive=True
), f"Apparently file glob [{include_path}] in job [{job.name}] digest_config [{job.digest_config}] invalid, workflow [{workflow.name}]. Setting to disable check: VALIDATE_FILE_PATHS"
else:
assert (
Path(include_path).is_file() or Path(include_path).is_dir()
), f"Apparently file path [{include_path}] in job [{job.name}] digest_config [{job.digest_config}] invalid, workflow [{workflow.name}]. Setting to disable check: VALIDATE_FILE_PATHS"
@classmethod
def validate_requirements_txt_files(cls, workflow: Workflow.Config) -> None:
with ContextManager.cd():
for job in workflow.jobs:
if job.job_requirements:
if job.job_requirements.python_requirements_txt:
path = Path(job.job_requirements.python_requirements_txt)
message = f"File with py requirement [{path}] does not exist"
if job.name in (
Settings.DOCKER_BUILD_JOB_NAME,
Settings.CI_CONFIG_JOB_NAME,
Settings.FINISH_WORKFLOW_JOB_NAME,
):
message += '\n If all requirements already installed on your runners - add setting INSTALL_PYTHON_REQS_FOR_NATIVE_JOBS""'
message += "\n If requirements needs to be installed - add requirements file (Settings.INSTALL_PYTHON_REQS_FOR_NATIVE_JOBS):"
message += "\n echo jwt==1.3.1 > ./ci/requirements.txt"
message += (
"\n echo requests==2.32.3 >> ./ci/requirements.txt"
)
message += "\n echo https://clickhouse-builds.s3.amazonaws.com/packages/praktika-0.1-py3-none-any.whl >> ./ci/requirements.txt"
cls.evaluate_check(
path.is_file(), message, job.name, workflow.name
)
@classmethod
def validate_dockers(cls, workflow: Workflow.Config):
names = []
for docker in workflow.dockers:
cls.evaluate_check(
docker.name not in names,
f"Non uniq docker name [{docker.name}]",
workflow_name=workflow.name,
)
names.append(docker.name)
for docker in workflow.dockers:
for docker_dep in docker.depends_on:
cls.evaluate_check(
docker_dep in names,
f"Docker [{docker.name}] has invalid dependency [{docker_dep}]",
workflow_name=workflow.name,
)
@classmethod
def evaluate_check(cls, check_ok, message, workflow_name, job_name=""):
message = message.split("\n")
messages = [message] if not isinstance(message, list) else message
if check_ok:
return
else:
print(
f"ERROR: Config validation failed: workflow [{workflow_name}], job [{job_name}]:"
)
for message in messages:
print(" || " + message)
sys.exit(1)

1
ci/praktika/version.py Normal file
View File

@ -0,0 +1 @@
VERSION = 1

68
ci/praktika/workflow.py Normal file
View File

@ -0,0 +1,68 @@
from dataclasses import dataclass, field
from typing import List, Optional
from praktika import Artifact, Job
from praktika.docker import Docker
from praktika.secret import Secret
from praktika.utils import Utils
class Workflow:
class Event:
PULL_REQUEST = "pull_request"
PUSH = "push"
@dataclass
class Config:
"""
branches - List of branch names or patterns, for push trigger only
base_branches - List of base branches (target branch), for pull_request trigger only
"""
name: str
event: str
jobs: List[Job.Config]
branches: List[str] = field(default_factory=list)
base_branches: List[str] = field(default_factory=list)
artifacts: List[Artifact.Config] = field(default_factory=list)
dockers: List[Docker.Config] = field(default_factory=list)
secrets: List[Secret.Config] = field(default_factory=list)
enable_cache: bool = False
enable_report: bool = False
enable_merge_ready_status: bool = False
enable_cidb: bool = False
def is_event_pull_request(self):
return self.event == Workflow.Event.PULL_REQUEST
def is_event_push(self):
return self.event == Workflow.Event.PUSH
def get_job(self, name):
job = self.find_job(name)
if not job:
Utils.raise_with_error(
f"Failed to find job [{name}], workflow [{self.name}]"
)
return job
def find_job(self, name, lazy=False):
name = str(name)
for job in self.jobs:
if lazy:
if name.lower() in job.name.lower():
return job
else:
if job.name == name:
return job
return None
def get_secret(self, name) -> Optional[Secret.Config]:
name = str(name)
names = []
for secret in self.secrets:
if secret.name == name:
return secret
names.append(secret.name)
print(f"ERROR: Failed to find secret [{name}], workflow secrets [{names}]")
raise

350
ci/praktika/yaml_generator.py Normal file
View File

@ -0,0 +1,350 @@
import dataclasses
from typing import List
from praktika import Artifact, Job, Workflow
from praktika.mangle import _get_workflows
from praktika.parser import WorkflowConfigParser
from praktika.runtime import RunConfig
from praktika.settings import Settings
from praktika.utils import ContextManager, Shell, Utils
class YamlGenerator:
class Templates:
TEMPLATE_PULL_REQUEST_0 = """\
# generated by praktika
name: {NAME}
on:
{EVENT}:
branches: [{BRANCHES}]
# Cancel the previous wf run in PRs.
concurrency:
group: ${{{{{{{{ github.workflow }}}}}}}}-${{{{{{{{ github.ref }}}}}}}}
cancel-in-progress: true
env:
# Force the stdout and stderr streams to be unbuffered
PYTHONUNBUFFERED: 1
GH_TOKEN: ${{{{{{{{ github.token }}}}}}}}
# Allow updating GH commit statuses and PR comments to post an actual job reports link
permissions: write-all
jobs:
{JOBS}\
"""
TEMPLATE_CALLABLE_WORKFLOW = """\
# generated by praktika
name: {NAME}
on:
workflow_call:
inputs:
config:
type: string
required: false
default: ''
secrets:
{SECRETS}
env:
PYTHONUNBUFFERED: 1
jobs:
{JOBS}\
"""
TEMPLATE_SECRET_CONFIG = """\
{SECRET_NAME}:
required: true
"""
TEMPLATE_MATRIX = """
strategy:
fail-fast: false
matrix:
params: {PARAMS_LIST}\
"""
TEMPLATE_JOB_0 = """
{JOB_NAME_NORMALIZED}:
runs-on: [{RUNS_ON}]
needs: [{NEEDS}]{IF_EXPRESSION}
name: "{JOB_NAME_GH}"
outputs:
data: ${{{{ steps.run.outputs.DATA }}}}
steps:
- name: Checkout code
uses: actions/checkout@v4
{JOB_ADDONS}
- name: Prepare env script
run: |
cat > {ENV_SETUP_SCRIPT} << 'ENV_SETUP_SCRIPT_EOF'
export PYTHONPATH=./ci:.
{SETUP_ENVS}
cat > {WORKFLOW_CONFIG_FILE} << 'EOF'
${{{{ needs.{WORKFLOW_CONFIG_JOB_NAME}.outputs.data }}}}
EOF
cat > {WORKFLOW_STATUS_FILE} << 'EOF'
${{{{ toJson(needs) }}}}
EOF
ENV_SETUP_SCRIPT_EOF
rm -rf {INPUT_DIR} {OUTPUT_DIR} {TEMP_DIR}
mkdir -p {TEMP_DIR} {INPUT_DIR} {OUTPUT_DIR}
{DOWNLOADS_GITHUB}
- name: Run
id: run
run: |
. /tmp/praktika_setup_env.sh
set -o pipefail
{PYTHON} -m praktika run --job '''{JOB_NAME}''' --workflow "{WORKFLOW_NAME}" --ci |& tee {RUN_LOG}
{UPLOADS_GITHUB}\
"""
TEMPLATE_SETUP_ENV_SECRETS = """\
export {SECRET_NAME}=$(cat<<'EOF'
${{{{ secrets.{SECRET_NAME} }}}}
EOF
)\
"""
TEMPLATE_PY_INSTALL = """
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: {PYTHON_VERSION}
"""
TEMPLATE_PY_WITH_REQUIREMENTS = """
- name: Install dependencies
run: |
sudo apt-get update && sudo apt install -y python3-pip
# TODO: --break-system-packages? otherwise ubuntu's apt/apt-get complains
{PYTHON} -m pip install --upgrade pip --break-system-packages
{PIP} install -r {REQUIREMENT_PATH} --break-system-packages
"""
TEMPLATE_GH_UPLOAD = """
- name: Upload artifact {NAME}
uses: actions/upload-artifact@v4
with:
name: {NAME}
path: {PATH}
"""
TEMPLATE_GH_DOWNLOAD = """
- name: Download artifact {NAME}
uses: actions/download-artifact@v4
with:
name: {NAME}
path: {PATH}
"""
TEMPLATE_IF_EXPRESSION = """
if: ${{{{ !failure() && !cancelled() && !contains(fromJson(needs.{WORKFLOW_CONFIG_JOB_NAME}.outputs.data).cache_success_base64, '{JOB_NAME_BASE64}') }}}}\
"""
TEMPLATE_IF_EXPRESSION_SKIPPED_OR_SUCCESS = """
if: ${{ !failure() && !cancelled() }}\
"""
TEMPLATE_IF_EXPRESSION_NOT_CANCELLED = """
if: ${{ !cancelled() }}\
"""
def __init__(self):
self.py_workflows = [] # type: List[Workflow.Config]
@classmethod
def _get_workflow_file_name(cls, workflow_name):
return f"{Settings.WORKFLOW_PATH_PREFIX}/{Utils.normalize_string(workflow_name)}.yaml"
def generate(self, workflow_file="", workflow_config=None):
print("---Start generating yaml pipelines---")
if workflow_config:
self.py_workflows = [workflow_config]
else:
self.py_workflows = _get_workflows(file=workflow_file)
assert self.py_workflows
for workflow_config in self.py_workflows:
print(f"Generate workflow [{workflow_config.name}]")
parser = WorkflowConfigParser(workflow_config).parse()
if (
workflow_config.is_event_pull_request()
or workflow_config.is_event_push()
):
yaml_workflow_str = PullRequestPushYamlGen(parser).generate()
else:
assert (
False
), f"Workflow event not yet supported [{workflow_config.event}]"
with ContextManager.cd():
with open(self._get_workflow_file_name(workflow_config.name), "w") as f:
f.write(yaml_workflow_str)
with ContextManager.cd():
Shell.check("git add ./.github/workflows/*.yaml")
class PullRequestPushYamlGen:
def __init__(self, parser: WorkflowConfigParser):
self.workflow_config = parser.workflow_yaml_config
self.parser = parser
def generate(self):
job_items = []
for i, job in enumerate(self.workflow_config.jobs):
job_name_normalized = Utils.normalize_string(job.name)
needs = ", ".join(map(Utils.normalize_string, job.needs))
job_name = job.name
job_addons = []
for addon in job.addons:
if addon.install_python:
job_addons.append(
YamlGenerator.Templates.TEMPLATE_PY_INSTALL.format(
PYTHON_VERSION=Settings.PYTHON_VERSION
)
)
if addon.requirements_txt_path:
job_addons.append(
YamlGenerator.Templates.TEMPLATE_PY_WITH_REQUIREMENTS.format(
PYTHON=Settings.PYTHON_INTERPRETER,
PIP=Settings.PYTHON_PACKET_MANAGER,
PYTHON_VERSION=Settings.PYTHON_VERSION,
REQUIREMENT_PATH=addon.requirements_txt_path,
)
)
uploads_github = []
for artifact in job.artifacts_gh_provides:
uploads_github.append(
YamlGenerator.Templates.TEMPLATE_GH_UPLOAD.format(
NAME=artifact.name, PATH=artifact.path
)
)
downloads_github = []
for artifact in job.artifacts_gh_requires:
downloads_github.append(
YamlGenerator.Templates.TEMPLATE_GH_DOWNLOAD.format(
NAME=artifact.name, PATH=Settings.INPUT_DIR
)
)
config_job_name_normalized = Utils.normalize_string(
Settings.CI_CONFIG_JOB_NAME
)
if_expression = ""
if (
self.workflow_config.enable_cache
and job_name_normalized != config_job_name_normalized
):
if_expression = YamlGenerator.Templates.TEMPLATE_IF_EXPRESSION.format(
WORKFLOW_CONFIG_JOB_NAME=config_job_name_normalized,
JOB_NAME_BASE64=Utils.to_base64(job_name),
)
if job.run_unless_cancelled:
if_expression = (
YamlGenerator.Templates.TEMPLATE_IF_EXPRESSION_NOT_CANCELLED
)
secrets_envs = []
for secret in self.workflow_config.secret_names_gh:
secrets_envs.append(
YamlGenerator.Templates.TEMPLATE_SETUP_ENV_SECRETS.format(
SECRET_NAME=secret
)
)
job_item = YamlGenerator.Templates.TEMPLATE_JOB_0.format(
JOB_NAME_NORMALIZED=job_name_normalized,
WORKFLOW_CONFIG_JOB_NAME=config_job_name_normalized,
IF_EXPRESSION=if_expression,
RUNS_ON=", ".join(job.runs_on),
NEEDS=needs,
JOB_NAME_GH=job_name.replace('"', '\\"'),
JOB_NAME=job_name.replace(
"'", "'\\''"
), # ' must be escaped so that yaml commands are properly parsed
WORKFLOW_NAME=self.workflow_config.name,
ENV_SETUP_SCRIPT=Settings.ENV_SETUP_SCRIPT,
SETUP_ENVS="\n".join(secrets_envs),
WORKFLOW_CONFIG_FILE=RunConfig.file_name_static(
self.workflow_config.name
),
JOB_ADDONS="".join(job_addons),
DOWNLOADS_GITHUB="\n".join(downloads_github),
UPLOADS_GITHUB="\n".join(uploads_github),
RUN_LOG=Settings.RUN_LOG,
PYTHON=Settings.PYTHON_INTERPRETER,
WORKFLOW_STATUS_FILE=Settings.WORKFLOW_STATUS_FILE,
TEMP_DIR=Settings.TEMP_DIR,
INPUT_DIR=Settings.INPUT_DIR,
OUTPUT_DIR=Settings.OUTPUT_DIR,
)
job_items.append(job_item)
base_template = YamlGenerator.Templates.TEMPLATE_PULL_REQUEST_0
template_1 = base_template.strip().format(
NAME=self.workflow_config.name,
BRANCHES=", ".join(
[f"'{branch}'" for branch in self.workflow_config.branches]
),
EVENT=self.workflow_config.event,
JOBS="{}" * len(job_items),
)
res = template_1.format(*job_items)
return res
@dataclasses.dataclass
class AuxConfig:
# defines aux step to install dependencies
addon: Job.Requirements
# defines aux step(s) to upload GH artifacts
uploads_gh: List[Artifact.Config]
# defines aux step(s) to download GH artifacts
downloads_gh: List[Artifact.Config]
def get_aux_workflow_name(self):
suffix = ""
if self.addon.python_requirements_txt:
suffix += "_py"
for _ in self.uploads_gh:
suffix += "_uplgh"
for _ in self.downloads_gh:
suffix += "_dnlgh"
return f"{Settings.WORKFLOW_PATH_PREFIX}/aux_job{suffix}.yaml"
def get_aux_workflow_input(self):
res = ""
if self.addon.python_requirements_txt:
res += f" requirements_txt: {self.addon.python_requirements_txt}"
return res
if __name__ == "__main__":
WFS = [
Workflow.Config(
name="PR",
event=Workflow.Event.PULL_REQUEST,
jobs=[
Job.Config(
name="Hello World",
runs_on=["foo"],
command="bar",
job_requirements=Job.Requirements(
python_requirements_txt="./requirement.txt"
),
)
],
enable_cache=True,
)
]
YamlGenerator().generate(workflow_config=WFS)

118
ci_v2/settings/definitions.py → ci/settings/definitions.py
View File

@ -7,6 +7,7 @@ S3_BUCKET_HTTP_ENDPOINT = "clickhouse-builds.s3.amazonaws.com"
class RunnerLabels:
CI_SERVICES = "ci_services"
CI_SERVICES_EBS = "ci_services_ebs"
BUILDER = "builder"
BASE_BRANCH = "master"
@ -29,155 +30,134 @@ SECRETS = [
DOCKERS = [
# Docker.Config(
# name="clickhouse/binary-builder",
# path="./docker/packager/binary-builder",
# arm64=True,
# amd64=True,
# path="./ci/docker/packager/binary-builder",
# platforms=Docker.Platforms.arm_amd,
# depends_on=[],
# ),
# Docker.Config(
# name="clickhouse/cctools",
# path="./docker/packager/cctools",
# arm64=True,
# amd64=True,
# path="./ci/docker/packager/cctools",
# platforms=Docker.Platforms.arm_amd,
# depends_on=[],
# ),
# Docker.Config(
# name="clickhouse/test-old-centos",
# path="./docker/test/compatibility/centos",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/compatibility/centos",
# platforms=Docker.Platforms.arm_amd,
# depends_on=[],
# ),
# Docker.Config(
# name="clickhouse/test-old-ubuntu",
# path="./docker/test/compatibility/ubuntu",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/compatibility/ubuntu",
# platforms=Docker.Platforms.arm_amd,
# depends_on=[],
# ),
# Docker.Config(
# name="clickhouse/test-util",
# path="./docker/test/util",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/util",
# platforms=Docker.Platforms.arm_amd,
# depends_on=[],
# ),
# Docker.Config(
# name="clickhouse/integration-test",
# path="./docker/test/integration/base",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/integration/base",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
# Docker.Config(
# name="clickhouse/fuzzer",
# path="./docker/test/fuzzer",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/fuzzer",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
# Docker.Config(
# name="clickhouse/performance-comparison",
# path="./docker/test/performance-comparison",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/performance-comparison",
# platforms=Docker.Platforms.arm_amd,
# depends_on=[],
# ),
# Docker.Config(
# name="clickhouse/fasttest",
# path="./docker/test/fasttest",
# arm64=True,
# amd64=True,
# depends_on=["clickhouse/test-util"],
# ),
Docker.Config(
name="clickhouse/fasttest",
path="./ci/docker/fasttest",
platforms=Docker.Platforms.arm_amd,
depends_on=[],
),
# Docker.Config(
# name="clickhouse/test-base",
# path="./docker/test/base",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/base",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-util"],
# ),
# Docker.Config(
# name="clickhouse/clickbench",
# path="./docker/test/clickbench",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/clickbench",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
# Docker.Config(
# name="clickhouse/keeper-jepsen-test",
# path="./docker/test/keeper-jepsen",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/keeper-jepsen",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
# Docker.Config(
# name="clickhouse/server-jepsen-test",
# path="./docker/test/server-jepsen",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/server-jepsen",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
# Docker.Config(
# name="clickhouse/sqllogic-test",
# path="./docker/test/sqllogic",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/sqllogic",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
# Docker.Config(
# name="clickhouse/sqltest",
# path="./docker/test/sqltest",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/sqltest",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
# Docker.Config(
# name="clickhouse/stateless-test",
# path="./docker/test/stateless",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/stateless",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
# Docker.Config(
# name="clickhouse/stateful-test",
# path="./docker/test/stateful",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/stateful",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/stateless-test"],
# ),
# Docker.Config(
# name="clickhouse/stress-test",
# path="./docker/test/stress",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/stress",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/stateful-test"],
# ),
# Docker.Config(
# name="clickhouse/unit-test",
# path="./docker/test/unit",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/unit",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
# Docker.Config(
# name="clickhouse/integration-tests-runner",
# path="./docker/test/integration/runner",
# arm64=True,
# amd64=True,
# path="./ci/docker/test/integration/runner",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
Docker.Config(
name="clickhouse/style-test",
path="./ci_v2/docker/style-test",
path="./ci/docker/style-test",
platforms=Docker.Platforms.arm_amd,
depends_on=[],
),
# Docker.Config(
# name="clickhouse/docs-builder",
# path="./docker/docs/builder",
# arm64=True,
# amd64=True,
# path="./ci/docker/docs/builder",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
]
@ -249,3 +229,5 @@ DOCKERS = [
class JobNames:
STYLE_CHECK = "Style Check"
FAST_TEST = "Fast test"
BUILD_AMD_DEBUG = "Build amd64 debug"

2
ci_v2/settings/settings.py → ci/settings/settings.py
View File

@ -1,4 +1,4 @@
from ci_v2.settings.definitions import (
from ci.settings.definitions import (
S3_BUCKET_HTTP_ENDPOINT,
S3_BUCKET_NAME,
RunnerLabels,

94
ci/workflows/pull_request.py Normal file
View File

@ -0,0 +1,94 @@
from typing import List
from praktika import Artifact, Job, Workflow
from praktika.settings import Settings
from ci.settings.definitions import (
BASE_BRANCH,
DOCKERS,
SECRETS,
JobNames,
RunnerLabels,
)
class ArtifactNames:
ch_debug_binary = "clickhouse_debug_binary"
style_check_job = Job.Config(
name=JobNames.STYLE_CHECK,
runs_on=[RunnerLabels.CI_SERVICES],
command="python3 ./ci/jobs/check_style.py",
run_in_docker="clickhouse/style-test",
)
fast_test_job = Job.Config(
name=JobNames.FAST_TEST,
runs_on=[RunnerLabels.BUILDER],
command="python3 ./ci/jobs/fast_test.py",
run_in_docker="clickhouse/fasttest",
digest_config=Job.CacheDigestConfig(
include_paths=[
"./ci/jobs/fast_test.py",
"./tests/queries/0_stateless/",
"./src",
],
),
)
job_build_amd_debug = Job.Config(
name=JobNames.BUILD_AMD_DEBUG,
runs_on=[RunnerLabels.BUILDER],
command="python3 ./ci/jobs/build_clickhouse.py amd_debug",
run_in_docker="clickhouse/fasttest",
digest_config=Job.CacheDigestConfig(
include_paths=[
"./src",
"./contrib/",
"./CMakeLists.txt",
"./PreLoad.cmake",
"./cmake",
"./base",
"./programs",
"./docker/packager/packager",
"./rust",
"./tests/ci/version_helper.py",
],
),
provides=[ArtifactNames.ch_debug_binary],
)
workflow = Workflow.Config(
name="PR",
event=Workflow.Event.PULL_REQUEST,
base_branches=[BASE_BRANCH],
jobs=[
style_check_job,
fast_test_job,
job_build_amd_debug,
],
artifacts=[
Artifact.Config(
name=ArtifactNames.ch_debug_binary,
type=Artifact.Type.S3,
path=f"{Settings.TEMP_DIR}/build/programs/clickhouse",
)
],
dockers=DOCKERS,
secrets=SECRETS,
enable_cache=True,
enable_report=True,
enable_merge_ready_status=True,
)
WORKFLOWS = [
workflow,
] # type: List[Workflow.Config]
if __name__ == "__main__":
# local job test inside praktika environment
from praktika.runner import Runner
Runner().run(workflow, fast_test_job, docker="fasttest", dummy_env=True)

4
ci_v2/docker/style-test/requirements.txt
View File

@ -1,4 +0,0 @@
requests==2.32.3
yamllint==1.26.3
codespell==2.2.1
https://clickhouse-builds.s3.amazonaws.com/packages/praktika-0.1-py3-none-any.whl

44
ci_v2/workflows/pull_request.py
View File

@ -1,44 +0,0 @@
from typing import List
from ci_v2.settings.definitions import (
BASE_BRANCH,
DOCKERS,
SECRETS,
JobNames,
RunnerLabels,
)
from praktika import Job, Workflow
style_check_job = Job.Config(
name=JobNames.STYLE_CHECK,
runs_on=[RunnerLabels.CI_SERVICES],
command="python3 ./ci_v2/jobs/check_style.py",
run_in_docker="clickhouse/style-test",
)
workflow = Workflow.Config(
name="PR",
event=Workflow.Event.PULL_REQUEST,
base_branches=[BASE_BRANCH],
jobs=[
style_check_job,
],
dockers=DOCKERS,
secrets=SECRETS,
enable_cache=True,
enable_report=True,
enable_merge_ready_status=True,
)
WORKFLOWS = [
workflow,
] # type: List[Workflow.Config]
if __name__ == "__main__":
# example: local job test inside praktika environment
from praktika.runner import Runner
Runner.generate_dummy_environment(workflow, style_check_job)
Runner().run(workflow, style_check_job)

10
cmake/autogenerated_versions.txt
View File

@ -2,11 +2,11 @@
# NOTE: VERSION_REVISION has nothing common with DBMS_TCP_PROTOCOL_VERSION,
# only DBMS_TCP_PROTOCOL_VERSION should be incremented on protocol changes.
SET(VERSION_REVISION 54491)
SET(VERSION_REVISION 54492)
SET(VERSION_MAJOR 24)
SET(VERSION_MINOR 10)
SET(VERSION_MINOR 11)
SET(VERSION_PATCH 1)
SET(VERSION_GITHASH b12a367741812f9e5fe754d19ebae600e2a2614c)
SET(VERSION_DESCRIBE v24.10.1.1-testing)
SET(VERSION_STRING 24.10.1.1)
SET(VERSION_GITHASH c82cf25b3e5864bcc153cbe45adb8c6527e1ec6e)
SET(VERSION_DESCRIBE v24.11.1.1-testing)
SET(VERSION_STRING 24.11.1.1)
# end of autochange

24
contrib/CMakeLists.txt vendored
View File

@ -178,35 +178,13 @@ add_contrib (sqlite-cmake sqlite-amalgamation)
add_contrib (s2geometry-cmake s2geometry)
add_contrib (c-ares-cmake c-ares)
if (OS_LINUX AND ARCH_AMD64 AND ENABLE_SSE42)
option (ENABLE_QPL "Enable Intel® Query Processing Library (QPL)" ${ENABLE_LIBRARIES})
elseif(ENABLE_QPL)
message (${RECONFIGURE_MESSAGE_LEVEL} "QPL library is only supported on x86_64 with SSE 4.2 or higher")
endif()
if (ENABLE_QPL)
add_contrib (idxd-config-cmake idxd-config)
add_contrib (qpl-cmake qpl) # requires: idxd-config
else()
message(STATUS "Not using QPL")
endif ()
if (OS_LINUX AND ARCH_AMD64 AND NOT NO_SSE3_OR_HIGHER)
option (ENABLE_QATLIB "Enable Intel® QuickAssist Technology Library (QATlib)" ${ENABLE_LIBRARIES})
elseif(ENABLE_QATLIB)
message (${RECONFIGURE_MESSAGE_LEVEL} "QATLib is only supported on x86_64")
endif()
if (ENABLE_QATLIB)
option (ENABLE_QAT_USDM_DRIVER "A User Space DMA-able Memory (USDM) component which allocates/frees DMA-able memory" OFF)
option (ENABLE_QAT_OUT_OF_TREE_BUILD "Using out-of-tree driver, user needs to customize ICP_ROOT variable" OFF)
set(ICP_ROOT "" CACHE STRING "ICP_ROOT variable to define the path of out-of-tree driver package")
if (ENABLE_QAT_OUT_OF_TREE_BUILD)
if (ICP_ROOT STREQUAL "")
message(FATAL_ERROR "Please define the path of out-of-tree driver package with -DICP_ROOT=xxx or disable out-of-tree build with -DENABLE_QAT_OUT_OF_TREE_BUILD=OFF; \
If you want out-of-tree build but have no package available, please download and build ICP package from: https://www.intel.com/content/www/us/en/download/765501.html")
endif ()
else()
add_contrib (qatlib-cmake qatlib) # requires: isa-l
endif ()
add_contrib (qatlib-cmake qatlib) # requires: isa-l
add_contrib (QAT-ZSTD-Plugin-cmake QAT-ZSTD-Plugin)
else()
message(STATUS "Not using QATLib")

126
contrib/QAT-ZSTD-Plugin-cmake/CMakeLists.txt
View File

@ -1,85 +1,53 @@
# Intel® QuickAssist Technology ZSTD Plugin (QAT ZSTD Plugin) is a plugin to Zstandard*(ZSTD*) for accelerating compression by QAT.
# ENABLE_QAT_OUT_OF_TREE_BUILD = 1 means kernel don't have native support, user will build and install driver from external package: https://www.intel.com/content/www/us/en/download/765501.html
# meanwhile, user need to set ICP_ROOT environment variable which point to the root directory of QAT driver source tree.
# ENABLE_QAT_OUT_OF_TREE_BUILD = 0 means kernel has built-in qat driver, QAT-ZSTD-PLUGIN just has dependency on qatlib.
if (ENABLE_QAT_OUT_OF_TREE_BUILD)
message(STATUS "Intel QATZSTD out-of-tree build, ICP_ROOT:${ICP_ROOT}")
message(STATUS "Intel QATZSTD in-tree build")
set(QATZSTD_SRC_DIR "${ClickHouse_SOURCE_DIR}/contrib/QAT-ZSTD-Plugin/src")
set(QATZSTD_SRC "${QATZSTD_SRC_DIR}/qatseqprod.c")
set(ZSTD_LIBRARY_DIR "${ClickHouse_SOURCE_DIR}/contrib/zstd/lib")
set(QATZSTD_SRC_DIR "${ClickHouse_SOURCE_DIR}/contrib/QAT-ZSTD-Plugin/src")
set(QATZSTD_SRC "${QATZSTD_SRC_DIR}/qatseqprod.c")
set(ZSTD_LIBRARY_DIR "${ClickHouse_SOURCE_DIR}/contrib/zstd/lib")
set(QAT_INCLUDE_DIR "${ICP_ROOT}/quickassist/include")
set(QAT_DC_INCLUDE_DIR "${ICP_ROOT}/quickassist/include/dc")
set(QAT_AL_INCLUDE_DIR "${ICP_ROOT}/quickassist/lookaside/access_layer/include")
set(QAT_USDM_INCLUDE_DIR "${ICP_ROOT}/quickassist/utilities/libusdm_drv")
set(USDM_LIBRARY "${ICP_ROOT}/build/libusdm_drv_s.so")
set(QAT_S_LIBRARY "${ICP_ROOT}/build/libqat_s.so")
if (ENABLE_QAT_USDM_DRIVER)
add_definitions(-DENABLE_USDM_DRV)
endif()
add_library(_qatzstd_plugin ${QATZSTD_SRC})
target_link_libraries (_qatzstd_plugin PUBLIC ${USDM_LIBRARY} ${QAT_S_LIBRARY})
target_include_directories(_qatzstd_plugin
SYSTEM PUBLIC "${QATZSTD_SRC_DIR}"
PRIVATE ${QAT_INCLUDE_DIR}
${QAT_DC_INCLUDE_DIR}
${QAT_AL_INCLUDE_DIR}
${QAT_USDM_INCLUDE_DIR}
${ZSTD_LIBRARY_DIR})
target_compile_definitions(_qatzstd_plugin PRIVATE -DDEBUGLEVEL=0)
add_library (ch_contrib::qatzstd_plugin ALIAS _qatzstd_plugin)
else () # In-tree build
message(STATUS "Intel QATZSTD in-tree build")
set(QATZSTD_SRC_DIR "${ClickHouse_SOURCE_DIR}/contrib/QAT-ZSTD-Plugin/src")
set(QATZSTD_SRC "${QATZSTD_SRC_DIR}/qatseqprod.c")
set(ZSTD_LIBRARY_DIR "${ClickHouse_SOURCE_DIR}/contrib/zstd/lib")
# please download&build ICP package from: https://www.intel.com/content/www/us/en/download/765501.html
set(ICP_ROOT "${ClickHouse_SOURCE_DIR}/contrib/qatlib")
set(QAT_INCLUDE_DIR "${ICP_ROOT}/quickassist/include")
set(QAT_DC_INCLUDE_DIR "${ICP_ROOT}/quickassist/include/dc")
set(QAT_AL_INCLUDE_DIR "${ICP_ROOT}/quickassist/lookaside/access_layer/include")
set(QAT_USDM_INCLUDE_DIR "${ICP_ROOT}/quickassist/utilities/libusdm_drv")
set(USDM_LIBRARY "${ICP_ROOT}/build/libusdm_drv_s.so")
set(QAT_S_LIBRARY "${ICP_ROOT}/build/libqat_s.so")
set(LIBQAT_ROOT_DIR "${ClickHouse_SOURCE_DIR}/contrib/qatlib")
set(LIBQAT_HEADER_DIR "${CMAKE_CURRENT_BINARY_DIR}/include")
# please download&build ICP package from: https://www.intel.com/content/www/us/en/download/765501.html
set(ICP_ROOT "${ClickHouse_SOURCE_DIR}/contrib/qatlib")
set(QAT_INCLUDE_DIR "${ICP_ROOT}/quickassist/include")
set(QAT_DC_INCLUDE_DIR "${ICP_ROOT}/quickassist/include/dc")
set(QAT_AL_INCLUDE_DIR "${ICP_ROOT}/quickassist/lookaside/access_layer/include")
set(QAT_USDM_INCLUDE_DIR "${ICP_ROOT}/quickassist/utilities/libusdm_drv")
set(USDM_LIBRARY "${ICP_ROOT}/build/libusdm_drv_s.so")
set(QAT_S_LIBRARY "${ICP_ROOT}/build/libqat_s.so")
set(LIBQAT_ROOT_DIR "${ClickHouse_SOURCE_DIR}/contrib/qatlib")
set(LIBQAT_HEADER_DIR "${CMAKE_CURRENT_BINARY_DIR}/include")
file(MAKE_DIRECTORY
"${LIBQAT_HEADER_DIR}/qat"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/include/cpa.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/include/dc/cpa_dc.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/lookaside/access_layer/include/icp_sal_poll.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/lookaside/access_layer/include/icp_sal_user.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/utilities/libusdm_drv/qae_mem.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
file(MAKE_DIRECTORY
"${LIBQAT_HEADER_DIR}/qat"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/include/cpa.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/include/dc/cpa_dc.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/lookaside/access_layer/include/icp_sal_poll.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/lookaside/access_layer/include/icp_sal_user.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
file(COPY "${LIBQAT_ROOT_DIR}/quickassist/utilities/libusdm_drv/qae_mem.h"
DESTINATION "${LIBQAT_HEADER_DIR}/qat/"
)
if (ENABLE_QAT_USDM_DRIVER)
add_definitions(-DENABLE_USDM_DRV)
endif()
add_library(_qatzstd_plugin ${QATZSTD_SRC})
target_link_libraries (_qatzstd_plugin PUBLIC ch_contrib::qatlib ch_contrib::usdm)
target_include_directories(_qatzstd_plugin PRIVATE
${QAT_INCLUDE_DIR}
${QAT_DC_INCLUDE_DIR}
${QAT_AL_INCLUDE_DIR}
${QAT_USDM_INCLUDE_DIR}
${ZSTD_LIBRARY_DIR}
${LIBQAT_HEADER_DIR})
target_compile_definitions(_qatzstd_plugin PRIVATE -DDEBUGLEVEL=0 PUBLIC -DINTREE)
target_include_directories(_qatzstd_plugin SYSTEM PUBLIC $<BUILD_INTERFACE:${QATZSTD_SRC_DIR}> $<INSTALL_INTERFACE:include>)
add_library (ch_contrib::qatzstd_plugin ALIAS _qatzstd_plugin)
endif ()
if (ENABLE_QAT_USDM_DRIVER)
add_definitions(-DENABLE_USDM_DRV)
endif()
add_library(_qatzstd_plugin ${QATZSTD_SRC})
target_link_libraries (_qatzstd_plugin PUBLIC ch_contrib::qatlib ch_contrib::usdm)
target_include_directories(_qatzstd_plugin PRIVATE
${QAT_INCLUDE_DIR}
${QAT_DC_INCLUDE_DIR}
${QAT_AL_INCLUDE_DIR}
${QAT_USDM_INCLUDE_DIR}
${ZSTD_LIBRARY_DIR}
${LIBQAT_HEADER_DIR})
target_compile_definitions(_qatzstd_plugin PRIVATE -DDEBUGLEVEL=0 PUBLIC -DINTREE)
target_include_directories(_qatzstd_plugin SYSTEM PUBLIC $<BUILD_INTERFACE:${QATZSTD_SRC_DIR}> $<INSTALL_INTERFACE:include>)
add_library (ch_contrib::qatzstd_plugin ALIAS _qatzstd_plugin)

1
contrib/idxd-config vendored

@ -1 +0,0 @@
Subproject commit a836ce0e42052a69bffbbc14239ab4097f3b77f1

23
contrib/idxd-config-cmake/CMakeLists.txt
View File

@ -1,23 +0,0 @@
## accel_config is the utility library required by QPL-Deflate codec for controlling and configuring Intel® In-Memory Analytics Accelerator (Intel® IAA).
set (LIBACCEL_SOURCE_DIR "${ClickHouse_SOURCE_DIR}/contrib/idxd-config")
set (UUID_DIR "${ClickHouse_SOURCE_DIR}/contrib/qpl-cmake")
set (LIBACCEL_HEADER_DIR "${ClickHouse_SOURCE_DIR}/contrib/idxd-config-cmake/include")
set (SRCS
"${LIBACCEL_SOURCE_DIR}/accfg/lib/libaccfg.c"
"${LIBACCEL_SOURCE_DIR}/util/log.c"
"${LIBACCEL_SOURCE_DIR}/util/sysfs.c"
)
add_library(_accel-config ${SRCS})
target_compile_options(_accel-config PRIVATE "-D_GNU_SOURCE")
target_include_directories(_accel-config BEFORE
PRIVATE ${UUID_DIR}
PRIVATE ${LIBACCEL_HEADER_DIR}
PRIVATE ${LIBACCEL_SOURCE_DIR})
target_include_directories(_accel-config SYSTEM BEFORE
PUBLIC ${LIBACCEL_SOURCE_DIR}/accfg)
add_library(ch_contrib::accel-config ALIAS _accel-config)

159
contrib/idxd-config-cmake/include/config.h
View File

@ -1,159 +0,0 @@
/* config.h. Generated from config.h.in by configure. */
/* config.h.in. Generated from configure.ac by autoheader. */
/* Define if building universal (internal helper macro) */
/* #undef AC_APPLE_UNIVERSAL_BUILD */
/* Debug messages. */
/* #undef ENABLE_DEBUG */
/* Documentation / man pages. */
/* #define ENABLE_DOCS */
/* System logging. */
#define ENABLE_LOGGING 1
/* accfg test support */
/* #undef ENABLE_TEST */
/* Define to 1 if big-endian-arch */
/* #undef HAVE_BIG_ENDIAN */
/* Define to 1 if you have the <dlfcn.h> header file. */
#define HAVE_DLFCN_H 1
/* Define to 1 if you have the <inttypes.h> header file. */
#define HAVE_INTTYPES_H 1
/* Define to 1 if you have the <linux/version.h> header file. */
#define HAVE_LINUX_VERSION_H 1
/* Define to 1 if little-endian-arch */
#define HAVE_LITTLE_ENDIAN 1
/* Define to 1 if you have the <memory.h> header file. */
#define HAVE_MEMORY_H 1
/* Define to 1 if you have the `secure_getenv' function. */
#define HAVE_SECURE_GETENV 1
/* Define to 1 if you have statement expressions. */
#define HAVE_STATEMENT_EXPR 1
/* Define to 1 if you have the <stdint.h> header file. */
#define HAVE_STDINT_H 1
/* Define to 1 if you have the <stdlib.h> header file. */
#define HAVE_STDLIB_H 1
/* Define to 1 if you have the <strings.h> header file. */
#define HAVE_STRINGS_H 1
/* Define to 1 if you have the <string.h> header file. */
#define HAVE_STRING_H 1
/* Define to 1 if you have the <sys/stat.h> header file. */
#define HAVE_SYS_STAT_H 1
/* Define to 1 if you have the <sys/types.h> header file. */
#define HAVE_SYS_TYPES_H 1
/* Define to 1 if typeof works with your compiler. */
#define HAVE_TYPEOF 1
/* Define to 1 if you have the <unistd.h> header file. */
#define HAVE_UNISTD_H 1
/* Define to 1 if using libuuid */
#define HAVE_UUID 1
/* Define to 1 if you have the `__secure_getenv' function. */
/* #undef HAVE___SECURE_GETENV */
/* Define to the sub-directory where libtool stores uninstalled libraries. */
#define LT_OBJDIR ".libs/"
/* Name of package */
#define PACKAGE "accel-config"
/* Define to the address where bug reports for this package should be sent. */
#define PACKAGE_BUGREPORT "linux-dsa@lists.01.org"
/* Define to the full name of this package. */
#define PACKAGE_NAME "accel-config"
/* Define to the full name and version of this package. */
#define PACKAGE_STRING "accel-config 3.5.2.gitf6605c41"
/* Define to the one symbol short name of this package. */
#define PACKAGE_TARNAME "accel-config"
/* Define to the home page for this package. */
#define PACKAGE_URL "https://github.com/xxx/accel-config"
/* Define to the version of this package. */
#define PACKAGE_VERSION "3.5.2.gitf6605c41"
/* Define to 1 if you have the ANSI C header files. */
#define STDC_HEADERS 1
/* Enable extensions on AIX 3, Interix. */
#ifndef _ALL_SOURCE
# define _ALL_SOURCE 1
#endif
/* Enable GNU extensions on systems that have them. */
#ifndef _GNU_SOURCE
# define _GNU_SOURCE 1
#endif
/* Enable threading extensions on Solaris. */
#ifndef _POSIX_PTHREAD_SEMANTICS
# define _POSIX_PTHREAD_SEMANTICS 1
#endif
/* Enable extensions on HP NonStop. */
#ifndef _TANDEM_SOURCE
# define _TANDEM_SOURCE 1
#endif
/* Enable general extensions on Solaris. */
#ifndef __EXTENSIONS__
# define __EXTENSIONS__ 1
#endif
/* Version number of package */
#define VERSION "3.5.2.gitf6605c41"
/* Define WORDS_BIGENDIAN to 1 if your processor stores words with the most
significant byte first (like Motorola and SPARC, unlike Intel). */
#if defined AC_APPLE_UNIVERSAL_BUILD
# if defined __BIG_ENDIAN__
# define WORDS_BIGENDIAN 1
# endif
#else
# ifndef WORDS_BIGENDIAN
/* # undef WORDS_BIGENDIAN */
# endif
#endif
/* Enable large inode numbers on Mac OS X 10.5. */
#ifndef _DARWIN_USE_64_BIT_INODE
# define _DARWIN_USE_64_BIT_INODE 1
#endif
/* Number of bits in a file offset, on hosts where this is settable. */
/* #undef _FILE_OFFSET_BITS */
/* Define for large files, on AIX-style hosts. */
/* #undef _LARGE_FILES */
/* Define to 1 if on MINIX. */
/* #undef _MINIX */
/* Define to 2 if the system does not provide POSIX.1 features except with
this defined. */
/* #undef _POSIX_1_SOURCE */
/* Define to 1 if you need to in order for `stat' and other things to work. */
/* #undef _POSIX_SOURCE */
/* Define to __typeof__ if your compiler spells it that way. */
/* #undef typeof */

2
contrib/libhdfs3 vendored

@ -1 +1 @@
Subproject commit 0d04201c45359f0d0701fb1e8297d25eff7cfecf
Subproject commit de6f1e0750aa3670a603cbfeddf5df3de1097687

2
contrib/numactl vendored

@ -1 +1 @@
Subproject commit 8d13d63a05f0c3cd88bf777cbb61541202b7da08
Subproject commit ff32c618d63ca7ac48cce366c5a04bb3563683a0

2
contrib/orc vendored

@ -1 +1 @@
Subproject commit bcc025c09828c556f54cfbdf83a66b9acae7d17f
Subproject commit 223e1957ff308f2aec7793884ffb0d7692d48484

1
contrib/qpl vendored

@ -1 +0,0 @@
Subproject commit c2ced94c53c1ee22191201a59878e9280bc9b9b8

738
contrib/qpl-cmake/CMakeLists.txt
View File

@ -1,738 +0,0 @@
## The Intel® QPL provides high performance implementations of data processing functions for existing hardware accelerator, and/or software path in case if hardware accelerator is not available.
set (UUID_DIR "${ClickHouse_SOURCE_DIR}/contrib/qpl-cmake")
set (QPL_PROJECT_DIR "${ClickHouse_SOURCE_DIR}/contrib/qpl")
set (QPL_SRC_DIR "${ClickHouse_SOURCE_DIR}/contrib/qpl/sources")
set (QPL_BINARY_DIR "${ClickHouse_BINARY_DIR}/build/contrib/qpl")
set (EFFICIENT_WAIT OFF)
set (LOG_HW_INIT OFF)
set (SANITIZE_MEMORY OFF)
set (SANITIZE_THREADS OFF)
set (LIB_FUZZING_ENGINE OFF)
set (DYNAMIC_LOADING_LIBACCEL_CONFIG OFF)
function(GetLibraryVersion _content _outputVar)
string(REGEX MATCHALL "QPL VERSION (.+) LANGUAGES" VERSION_REGEX "${_content}")
SET(${_outputVar} ${CMAKE_MATCH_1} PARENT_SCOPE)
endfunction()
set (QPL_VERSION 1.6.0)
message(STATUS "Intel QPL version: ${QPL_VERSION}")
# There are 5 source subdirectories under $QPL_SRC_DIR: c_api, core-iaa, core-sw, middle-layer and isal.
# Generate 8 library targets: qpl_c_api, core_iaa, qplcore_px, qplcore_avx512, qplcore_sw_dispatcher, middle_layer_lib, isal and isal_asm,
# which are then combined into static or shared qpl.
# Output ch_contrib::qpl by linking with 8 library targets.
# Note, QPL has integrated a customized version of ISA-L to meet specific needs.
# This version has been significantly modified and there are no plans to maintain compatibility with the upstream version
# or upgrade the current copy.
## cmake/CompileOptions.cmake and automatic wrappers generation
# ==========================================================================
# Copyright (C) 2022 Intel Corporation
#
# SPDX-License-Identifier: MIT
# ==========================================================================
set(QPL_LINUX_TOOLCHAIN_CPP_EMBEDDED_FLAGS "-fno-exceptions;-fno-rtti")
function(modify_standard_language_flag)
# Declaring function parameters
set(OPTIONS "")
set(ONE_VALUE_ARGS
LANGUAGE_NAME
FLAG_NAME
NEW_FLAG_VALUE)
set(MULTI_VALUE_ARGS "")
# Parsing function parameters
cmake_parse_arguments(MODIFY
"${OPTIONS}"
"${ONE_VALUE_ARGS}"
"${MULTI_VALUE_ARGS}"
${ARGN})
# Variables
set(FLAG_REGULAR_EXPRESSION "${MODIFY_FLAG_NAME}.*[ ]*")
set(NEW_VALUE "${MODIFY_FLAG_NAME}${MODIFY_NEW_FLAG_VALUE}")
# Replacing specified flag with new value
string(REGEX REPLACE
${FLAG_REGULAR_EXPRESSION} ${NEW_VALUE}
NEW_COMPILE_FLAGS
"${CMAKE_${MODIFY_LANGUAGE_NAME}_FLAGS}")
# Returning the value
set(CMAKE_${MODIFY_LANGUAGE_NAME}_FLAGS ${NEW_COMPILE_FLAGS} PARENT_SCOPE)
endfunction()
function(get_function_name_with_default_bit_width in_function_name bit_width out_function_name)
if(in_function_name MATCHES ".*_i")
string(REPLACE "_i" "" in_function_name ${in_function_name})
set(${out_function_name} "${in_function_name}_${bit_width}_i" PARENT_SCOPE)
else()
set(${out_function_name} "${in_function_name}_${bit_width}" PARENT_SCOPE)
endif()
endfunction()
macro(get_list_of_supported_optimizations PLATFORMS_LIST)
list(APPEND PLATFORMS_LIST "")
list(APPEND PLATFORMS_LIST "px")
list(APPEND PLATFORMS_LIST "avx512")
endmacro(get_list_of_supported_optimizations)
function(generate_unpack_kernel_arrays current_directory PLATFORMS_LIST)
list(APPEND UNPACK_POSTFIX_LIST "")
list(APPEND UNPACK_PRLE_POSTFIX_LIST "")
list(APPEND PACK_POSTFIX_LIST "")
list(APPEND PACK_INDEX_POSTFIX_LIST "")
list(APPEND SCAN_POSTFIX_LIST "")
list(APPEND DEFAULT_BIT_WIDTH_FUNCTIONS_LIST "")
list(APPEND DEFAULT_BIT_WIDTH_LIST "")
#create list of functions that use only 8u 16u 32u postfixes
list(APPEND DEFAULT_BIT_WIDTH_FUNCTIONS_LIST "unpack_prle")
list(APPEND DEFAULT_BIT_WIDTH_FUNCTIONS_LIST "extract")
list(APPEND DEFAULT_BIT_WIDTH_FUNCTIONS_LIST "extract_i")
list(APPEND DEFAULT_BIT_WIDTH_FUNCTIONS_LIST "select")
list(APPEND DEFAULT_BIT_WIDTH_FUNCTIONS_LIST "select_i")
list(APPEND DEFAULT_BIT_WIDTH_FUNCTIONS_LIST "expand")
#create default bit width list
list(APPEND DEFAULT_BIT_WIDTH_LIST "8u")
list(APPEND DEFAULT_BIT_WIDTH_LIST "16u")
list(APPEND DEFAULT_BIT_WIDTH_LIST "32u")
#create scan kernel postfixes
list(APPEND SCAN_COMPARATOR_LIST "")
list(APPEND SCAN_COMPARATOR_LIST "eq")
list(APPEND SCAN_COMPARATOR_LIST "ne")
list(APPEND SCAN_COMPARATOR_LIST "lt")
list(APPEND SCAN_COMPARATOR_LIST "le")
list(APPEND SCAN_COMPARATOR_LIST "gt")
list(APPEND SCAN_COMPARATOR_LIST "ge")
list(APPEND SCAN_COMPARATOR_LIST "range")
list(APPEND SCAN_COMPARATOR_LIST "not_range")
foreach(SCAN_COMPARATOR IN LISTS SCAN_COMPARATOR_LIST)
list(APPEND SCAN_POSTFIX_LIST "_${SCAN_COMPARATOR}_8u")
list(APPEND SCAN_POSTFIX_LIST "_${SCAN_COMPARATOR}_16u8u")
list(APPEND SCAN_POSTFIX_LIST "_${SCAN_COMPARATOR}_32u8u")
endforeach()
# create unpack kernel postfixes
foreach(input_width RANGE 1 32 1)
if(input_width LESS 8 OR input_width EQUAL 8)
list(APPEND UNPACK_POSTFIX_LIST "_${input_width}u8u")
elseif(input_width LESS 16 OR input_width EQUAL 16)
list(APPEND UNPACK_POSTFIX_LIST "_${input_width}u16u")
else()
list(APPEND UNPACK_POSTFIX_LIST "_${input_width}u32u")
endif()
endforeach()
# create pack kernel postfixes
foreach(output_width RANGE 1 8 1)
list(APPEND PACK_POSTFIX_LIST "_8u${output_width}u")
endforeach()
foreach(output_width RANGE 9 16 1)
list(APPEND PACK_POSTFIX_LIST "_16u${output_width}u")
endforeach()
foreach(output_width RANGE 17 32 1)
list(APPEND PACK_POSTFIX_LIST "_32u${output_width}u")
endforeach()
list(APPEND PACK_POSTFIX_LIST "_8u16u")
list(APPEND PACK_POSTFIX_LIST "_8u32u")
list(APPEND PACK_POSTFIX_LIST "_16u32u")
# create pack index kernel postfixes
list(APPEND PACK_INDEX_POSTFIX_LIST "_nu")
list(APPEND PACK_INDEX_POSTFIX_LIST "_8u")
list(APPEND PACK_INDEX_POSTFIX_LIST "_8u16u")
list(APPEND PACK_INDEX_POSTFIX_LIST "_8u32u")
# write to file
file(MAKE_DIRECTORY ${current_directory}/generated)
foreach(PLATFORM_VALUE IN LISTS PLATFORMS_LIST)
set(directory "${current_directory}/generated")
set(PLATFORM_PREFIX "${PLATFORM_VALUE}_")
#
# Write unpack table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}unpack.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}unpack.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}unpack.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}unpack.cpp "unpack_table_t ${PLATFORM_PREFIX}unpack_table = {\n")
#write LE kernels
foreach(UNPACK_POSTFIX IN LISTS UNPACK_POSTFIX_LIST)
file(APPEND ${directory}/${PLATFORM_PREFIX}unpack.cpp "\t${PLATFORM_PREFIX}qplc_unpack${UNPACK_POSTFIX},\n")
endforeach()
#write BE kernels
#get last element of the list
set(LAST_ELEMENT "")
list(GET UNPACK_POSTFIX_LIST -1 LAST_ELEMENT)
foreach(UNPACK_POSTFIX IN LISTS UNPACK_POSTFIX_LIST)
if(UNPACK_POSTFIX STREQUAL LAST_ELEMENT)
file(APPEND ${directory}/${PLATFORM_PREFIX}unpack.cpp "\t${PLATFORM_PREFIX}qplc_unpack_be${UNPACK_POSTFIX}};\n")
else()
file(APPEND ${directory}/${PLATFORM_PREFIX}unpack.cpp "\t${PLATFORM_PREFIX}qplc_unpack_be${UNPACK_POSTFIX},\n")
endif()
endforeach()
file(APPEND ${directory}/${PLATFORM_PREFIX}unpack.cpp "}\n")
#
# Write pack table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}pack.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack.cpp "pack_table_t ${PLATFORM_PREFIX}pack_table = {\n")
#write LE kernels
foreach(PACK_POSTFIX IN LISTS PACK_POSTFIX_LIST)
file(APPEND ${directory}/${PLATFORM_PREFIX}pack.cpp "\t${PLATFORM_PREFIX}qplc_pack${PACK_POSTFIX},\n")
endforeach()
#write BE kernels
#get last element of the list
set(LAST_ELEMENT "")
list(GET PACK_POSTFIX_LIST -1 LAST_ELEMENT)
foreach(PACK_POSTFIX IN LISTS PACK_POSTFIX_LIST)
if(PACK_POSTFIX STREQUAL LAST_ELEMENT)
file(APPEND ${directory}/${PLATFORM_PREFIX}pack.cpp "\t${PLATFORM_PREFIX}qplc_pack_be${PACK_POSTFIX}};\n")
else()
file(APPEND ${directory}/${PLATFORM_PREFIX}pack.cpp "\t${PLATFORM_PREFIX}qplc_pack_be${PACK_POSTFIX},\n")
endif()
endforeach()
file(APPEND ${directory}/${PLATFORM_PREFIX}pack.cpp "}\n")
#
# Write scan table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}scan.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}scan.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}scan.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}scan.cpp "scan_table_t ${PLATFORM_PREFIX}scan_table = {\n")
#get last element of the list
set(LAST_ELEMENT "")
list(GET SCAN_POSTFIX_LIST -1 LAST_ELEMENT)
foreach(SCAN_POSTFIX IN LISTS SCAN_POSTFIX_LIST)
if(SCAN_POSTFIX STREQUAL LAST_ELEMENT)
file(APPEND ${directory}/${PLATFORM_PREFIX}scan.cpp "\t${PLATFORM_PREFIX}qplc_scan${SCAN_POSTFIX}};\n")
else()
file(APPEND ${directory}/${PLATFORM_PREFIX}scan.cpp "\t${PLATFORM_PREFIX}qplc_scan${SCAN_POSTFIX},\n")
endif()
endforeach()
file(APPEND ${directory}/${PLATFORM_PREFIX}scan.cpp "}\n")
#
# Write scan_i table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}scan_i.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}scan_i.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}scan_i.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}scan_i.cpp "scan_i_table_t ${PLATFORM_PREFIX}scan_i_table = {\n")
#get last element of the list
set(LAST_ELEMENT "")
list(GET SCAN_POSTFIX_LIST -1 LAST_ELEMENT)
foreach(SCAN_POSTFIX IN LISTS SCAN_POSTFIX_LIST)
if(SCAN_POSTFIX STREQUAL LAST_ELEMENT)
file(APPEND ${directory}/${PLATFORM_PREFIX}scan_i.cpp "\t${PLATFORM_PREFIX}qplc_scan${SCAN_POSTFIX}_i};\n")
else()
file(APPEND ${directory}/${PLATFORM_PREFIX}scan_i.cpp "\t${PLATFORM_PREFIX}qplc_scan${SCAN_POSTFIX}_i,\n")
endif()
endforeach()
file(APPEND ${directory}/${PLATFORM_PREFIX}scan_i.cpp "}\n")
#
# Write pack_index table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}pack_index.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "pack_index_table_t ${PLATFORM_PREFIX}pack_index_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "\t${PLATFORM_PREFIX}qplc_pack_bits_nu,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "\t${PLATFORM_PREFIX}qplc_pack_index_8u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "\t${PLATFORM_PREFIX}qplc_pack_index_8u16u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "\t${PLATFORM_PREFIX}qplc_pack_index_8u32u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "\t${PLATFORM_PREFIX}qplc_pack_bits_be_nu,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "\t${PLATFORM_PREFIX}qplc_pack_index_8u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "\t${PLATFORM_PREFIX}qplc_pack_index_be_8u16u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "\t${PLATFORM_PREFIX}qplc_pack_index_be_8u32u};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}pack_index.cpp "}\n")
#
# Write default bit width functions
#
foreach(DEAULT_BIT_WIDTH_FUNCTION IN LISTS DEFAULT_BIT_WIDTH_FUNCTIONS_LIST)
file(WRITE ${directory}/${PLATFORM_PREFIX}${DEAULT_BIT_WIDTH_FUNCTION}.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}${DEAULT_BIT_WIDTH_FUNCTION}.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}${DEAULT_BIT_WIDTH_FUNCTION}.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}${DEAULT_BIT_WIDTH_FUNCTION}.cpp "${DEAULT_BIT_WIDTH_FUNCTION}_table_t ${PLATFORM_PREFIX}${DEAULT_BIT_WIDTH_FUNCTION}_table = {\n")
#get last element of the list
set(LAST_ELEMENT "")
list(GET DEFAULT_BIT_WIDTH_LIST -1 LAST_ELEMENT)
foreach(BIT_WIDTH IN LISTS DEFAULT_BIT_WIDTH_LIST)
set(FUNCTION_NAME "")
get_function_name_with_default_bit_width(${DEAULT_BIT_WIDTH_FUNCTION} ${BIT_WIDTH} FUNCTION_NAME)
if(BIT_WIDTH STREQUAL LAST_ELEMENT)
file(APPEND ${directory}/${PLATFORM_PREFIX}${DEAULT_BIT_WIDTH_FUNCTION}.cpp "\t${PLATFORM_PREFIX}qplc_${FUNCTION_NAME}};\n")
else()
file(APPEND ${directory}/${PLATFORM_PREFIX}${DEAULT_BIT_WIDTH_FUNCTION}.cpp "\t${PLATFORM_PREFIX}qplc_${FUNCTION_NAME},\n")
endif()
endforeach()
file(APPEND ${directory}/${PLATFORM_PREFIX}${DEAULT_BIT_WIDTH_FUNCTION}.cpp "}\n")
endforeach()
#
# Write aggregates table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}aggregates.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}aggregates.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}aggregates.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}aggregates.cpp "aggregates_table_t ${PLATFORM_PREFIX}aggregates_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}aggregates.cpp "\t${PLATFORM_PREFIX}qplc_bit_aggregates_8u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}aggregates.cpp "\t${PLATFORM_PREFIX}qplc_aggregates_8u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}aggregates.cpp "\t${PLATFORM_PREFIX}qplc_aggregates_16u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}aggregates.cpp "\t${PLATFORM_PREFIX}qplc_aggregates_32u};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}aggregates.cpp "}\n")
#
# Write mem_copy functions table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}memory_copy.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}memory_copy.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}memory_copy.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}memory_copy.cpp "memory_copy_table_t ${PLATFORM_PREFIX}memory_copy_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}memory_copy.cpp "\t${PLATFORM_PREFIX}qplc_copy_8u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}memory_copy.cpp "\t${PLATFORM_PREFIX}qplc_copy_16u,\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}memory_copy.cpp "\t${PLATFORM_PREFIX}qplc_copy_32u};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}memory_copy.cpp "}\n")
#
# Write mem_copy functions table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}zero.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}zero.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}zero.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}zero.cpp "zero_table_t ${PLATFORM_PREFIX}zero_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}zero.cpp "\t${PLATFORM_PREFIX}qplc_zero_8u};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}zero.cpp "}\n")
#
# Write move functions table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}move.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}move.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}move.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}move.cpp "move_table_t ${PLATFORM_PREFIX}move_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}move.cpp "\t${PLATFORM_PREFIX}qplc_move_8u};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}move.cpp "}\n")
#
# Write crc64 function table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}crc64.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}crc64.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}crc64.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}crc64.cpp "crc64_table_t ${PLATFORM_PREFIX}crc64_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}crc64.cpp "\t${PLATFORM_PREFIX}qplc_crc64};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}crc64.cpp "}\n")
#
# Write xor_checksum function table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}xor_checksum.cpp "#include \"qplc_api.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}xor_checksum.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}xor_checksum.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}xor_checksum.cpp "xor_checksum_table_t ${PLATFORM_PREFIX}xor_checksum_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}xor_checksum.cpp "\t${PLATFORM_PREFIX}qplc_xor_checksum_8u};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}xor_checksum.cpp "}\n")
#
# Write deflate functions table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}deflate.cpp "#include \"deflate_slow_icf.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate.cpp "#include \"deflate_hash_table.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate.cpp "#include \"deflate_histogram.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate.cpp "deflate_table_t ${PLATFORM_PREFIX}deflate_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate.cpp "\t reinterpret_cast<void *>(&${PLATFORM_PREFIX}slow_deflate_icf_body),\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate.cpp "\t reinterpret_cast<void *>(&${PLATFORM_PREFIX}deflate_histogram_reset),\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate.cpp "\t reinterpret_cast<void *>(&${PLATFORM_PREFIX}deflate_hash_table_reset)};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate.cpp "}\n")
#
# Write deflate fix functions table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}deflate_fix.cpp "#include \"deflate_slow.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate_fix.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate_fix.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate_fix.cpp "deflate_fix_table_t ${PLATFORM_PREFIX}deflate_fix_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate_fix.cpp "\t reinterpret_cast<void *>(&${PLATFORM_PREFIX}slow_deflate_body)};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}deflate_fix.cpp "}\n")
#
# Write setup_dictionary functions table
#
file(WRITE ${directory}/${PLATFORM_PREFIX}setup_dictionary.cpp "#include \"deflate_slow_utils.h\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}setup_dictionary.cpp "#include \"dispatcher/dispatcher.hpp\"\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}setup_dictionary.cpp "namespace qpl::core_sw::dispatcher\n{\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}setup_dictionary.cpp "setup_dictionary_table_t ${PLATFORM_PREFIX}setup_dictionary_table = {\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}setup_dictionary.cpp "\t reinterpret_cast<void *>(&${PLATFORM_PREFIX}setup_dictionary)};\n")
file(APPEND ${directory}/${PLATFORM_PREFIX}setup_dictionary.cpp "}\n")
endforeach()
endfunction()
# [SUBDIR]isal
enable_language(ASM_NASM)
set(ISAL_C_SRC ${QPL_SRC_DIR}/isal/igzip/adler32_base.c
${QPL_SRC_DIR}/isal/igzip/huff_codes.c
${QPL_SRC_DIR}/isal/igzip/hufftables_c.c
${QPL_SRC_DIR}/isal/igzip/igzip.c
${QPL_SRC_DIR}/isal/igzip/igzip_base.c
${QPL_SRC_DIR}/isal/igzip/flatten_ll.c
${QPL_SRC_DIR}/isal/igzip/encode_df.c
${QPL_SRC_DIR}/isal/igzip/igzip_icf_base.c
${QPL_SRC_DIR}/isal/igzip/igzip_inflate.c
${QPL_SRC_DIR}/isal/igzip/igzip_icf_body.c
${QPL_SRC_DIR}/isal/crc/crc_base.c
${QPL_SRC_DIR}/isal/crc/crc64_base.c)
set(ISAL_ASM_SRC ${QPL_SRC_DIR}/isal/igzip/igzip_body.asm
${QPL_SRC_DIR}/isal/igzip/igzip_gen_icf_map_lh1_04.asm
${QPL_SRC_DIR}/isal/igzip/igzip_gen_icf_map_lh1_06.asm
${QPL_SRC_DIR}/isal/igzip/igzip_decode_block_stateless_04.asm
${QPL_SRC_DIR}/isal/igzip/igzip_finish.asm
${QPL_SRC_DIR}/isal/igzip/encode_df_04.asm
${QPL_SRC_DIR}/isal/igzip/encode_df_06.asm
${QPL_SRC_DIR}/isal/igzip/igzip_decode_block_stateless_01.asm
${QPL_SRC_DIR}/isal/igzip/proc_heap.asm
${QPL_SRC_DIR}/isal/igzip/igzip_icf_body_h1_gr_bt.asm
${QPL_SRC_DIR}/isal/igzip/igzip_icf_finish.asm
${QPL_SRC_DIR}/isal/igzip/igzip_inflate_multibinary.asm
${QPL_SRC_DIR}/isal/igzip/igzip_update_histogram_01.asm
${QPL_SRC_DIR}/isal/igzip/igzip_update_histogram_04.asm
${QPL_SRC_DIR}/isal/igzip/rfc1951_lookup.asm
${QPL_SRC_DIR}/isal/igzip/adler32_sse.asm
${QPL_SRC_DIR}/isal/igzip/adler32_avx2_4.asm
${QPL_SRC_DIR}/isal/igzip/igzip_deflate_hash.asm
${QPL_SRC_DIR}/isal/igzip/igzip_set_long_icf_fg_04.asm
${QPL_SRC_DIR}/isal/igzip/igzip_set_long_icf_fg_06.asm
${QPL_SRC_DIR}/isal/igzip/igzip_multibinary.asm
${QPL_SRC_DIR}/isal/crc/crc_multibinary.asm
${QPL_SRC_DIR}/isal/crc/crc32_gzip_refl_by8.asm
${QPL_SRC_DIR}/isal/crc/crc32_gzip_refl_by8_02.asm
${QPL_SRC_DIR}/isal/crc/crc32_gzip_refl_by16_10.asm
${QPL_SRC_DIR}/isal/crc/crc32_ieee_01.asm
${QPL_SRC_DIR}/isal/crc/crc32_ieee_02.asm
${QPL_SRC_DIR}/isal/crc/crc32_ieee_by4.asm
${QPL_SRC_DIR}/isal/crc/crc32_ieee_by16_10.asm
${QPL_SRC_DIR}/isal/crc/crc32_iscsi_00.asm
${QPL_SRC_DIR}/isal/crc/crc32_iscsi_01.asm
${QPL_SRC_DIR}/isal/crc/crc32_iscsi_by16_10.asm)
# Adding ISA-L library target
add_library(isal OBJECT ${ISAL_C_SRC})
add_library(isal_asm OBJECT ${ISAL_ASM_SRC})
set_property(GLOBAL APPEND PROPERTY QPL_LIB_DEPS
$<TARGET_OBJECTS:isal>)
set_property(GLOBAL APPEND PROPERTY QPL_LIB_DEPS
$<TARGET_OBJECTS:isal_asm>)
# Setting external and internal interfaces for ISA-L library
target_include_directories(isal
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/isal/include>
PUBLIC ${QPL_SRC_DIR}/isal/igzip)
set_target_properties(isal PROPERTIES
CXX_STANDARD 11
C_STANDARD 99)
# AS_FEATURE_LEVEL=10 means "Check SIMD capabilities of the target system at runtime and use up to AVX512 if available".
# HAVE_KNOWS_AVX512 means rely on AVX512 being available on the target system.
target_compile_options(isal_asm PRIVATE "-I${QPL_SRC_DIR}/isal/include/"
PRIVATE "-I${QPL_SRC_DIR}/isal/igzip/"
PRIVATE "-I${QPL_SRC_DIR}/isal/crc/"
PRIVATE "-DHAVE_AS_KNOWS_AVX512"
PRIVATE "-DAS_FEATURE_LEVEL=10"
PRIVATE "-DQPL_LIB")
# Here must remove "-fno-sanitize=undefined" from COMPILE_OPTIONS.
# Otherwise nasm compiler would fail to proceed due to unrecognition of "-fno-sanitize=undefined"
if (SANITIZE STREQUAL "undefined")
get_target_property(target_options isal_asm COMPILE_OPTIONS)
list(REMOVE_ITEM target_options "-fno-sanitize=undefined")
set_property(TARGET isal_asm PROPERTY COMPILE_OPTIONS ${target_options})
endif()
target_compile_definitions(isal PUBLIC
QPL_LIB
NDEBUG)
# [SUBDIR]core-sw
# Create set of libraries corresponding to supported platforms for SW fallback which are implemented by AVX512 and non-AVX512 instructions respectively.
# The upper level QPL API will check SIMD capabilities of the target system at runtime and decide to call AVX512 function or non-AVX512 function.
# Hence, here we don't need put ENABLE_AVX512 CMake switch.
get_list_of_supported_optimizations(PLATFORMS_LIST)
foreach(PLATFORM_ID IN LISTS PLATFORMS_LIST)
# Find Core Sources
file(GLOB SOURCES
${QPL_SRC_DIR}/core-sw/src/checksums/*.c
${QPL_SRC_DIR}/core-sw/src/filtering/*.c
${QPL_SRC_DIR}/core-sw/src/other/*.c
${QPL_SRC_DIR}/core-sw/src/compression/*.c)
file(GLOB DATA_SOURCES
${QPL_SRC_DIR}/core-sw/src/data/*.c)
# Create library
add_library(qplcore_${PLATFORM_ID} OBJECT ${SOURCES})
set_property(GLOBAL APPEND PROPERTY QPL_LIB_DEPS
$<TARGET_OBJECTS:qplcore_${PLATFORM_ID}>)
target_include_directories(qplcore_${PLATFORM_ID}
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/core-sw>
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/core-sw/include>
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/core-sw/src/include>
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/core-sw/src/compression/include>
PRIVATE $<TARGET_PROPERTY:isal,INTERFACE_INCLUDE_DIRECTORIES>)
# Set specific compiler options and/or definitions based on a platform
if (${PLATFORM_ID} MATCHES "avx512")
target_compile_definitions(qplcore_${PLATFORM_ID} PRIVATE PLATFORM=2)
target_compile_options(qplcore_${PLATFORM_ID} PRIVATE -march=skylake-avx512)
else() # Create default px library
target_compile_definitions(qplcore_${PLATFORM_ID} PRIVATE PLATFORM=0)
endif()
target_link_libraries(qplcore_${PLATFORM_ID} isal)
endforeach()
#
# Create dispatcher between platforms and auto-generated wrappers
#
file(GLOB SW_DISPATCHER_SOURCES ${QPL_SRC_DIR}/core-sw/dispatcher/*.cpp)
add_library(qplcore_sw_dispatcher OBJECT ${SW_DISPATCHER_SOURCES})
set_property(GLOBAL APPEND PROPERTY QPL_LIB_DEPS
$<TARGET_OBJECTS:qplcore_sw_dispatcher>)
target_include_directories(qplcore_sw_dispatcher
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/core-sw/dispatcher>)
# Generate kernel wrappers
generate_unpack_kernel_arrays(${QPL_BINARY_DIR} "${PLATFORMS_LIST}")
foreach(PLATFORM_ID IN LISTS PLATFORMS_LIST)
file(GLOB GENERATED_${PLATFORM_ID}_TABLES_SRC ${QPL_BINARY_DIR}/generated/${PLATFORM_ID}_*.cpp)
target_sources(qplcore_sw_dispatcher PRIVATE ${GENERATED_${PLATFORM_ID}_TABLES_SRC})
# Set specific compiler options and/or definitions based on a platform
if (${PLATFORM_ID} MATCHES "avx512")
set_source_files_properties(${GENERATED_${PLATFORM_ID}_TABLES_SRC} PROPERTIES COMPILE_DEFINITIONS PLATFORM=2)
else()
set_source_files_properties(${GENERATED_${PLATFORM_ID}_TABLES_SRC} PROPERTIES COMPILE_DEFINITIONS PLATFORM=0)
endif()
target_include_directories(qplcore_sw_dispatcher
PUBLIC $<TARGET_PROPERTY:qplcore_${PLATFORM_ID},INTERFACE_INCLUDE_DIRECTORIES>)
endforeach()
set_target_properties(qplcore_sw_dispatcher PROPERTIES CXX_STANDARD 17)
# w/a for build compatibility with ISAL codebase
target_compile_definitions(qplcore_sw_dispatcher PUBLIC -DQPL_LIB)
target_compile_options(qplcore_sw_dispatcher
PRIVATE ${QPL_LINUX_TOOLCHAIN_CPP_EMBEDDED_FLAGS})
# [SUBDIR]core-iaa
file(GLOB HW_PATH_SRC ${QPL_SRC_DIR}/core-iaa/sources/aecs/*.c
${QPL_SRC_DIR}/core-iaa/sources/driver_loader/*.c
${QPL_SRC_DIR}/core-iaa/sources/descriptors/*.c
${QPL_SRC_DIR}/core-iaa/sources/*.c)
# Create library
add_library(core_iaa OBJECT ${HW_PATH_SRC})
set_property(GLOBAL APPEND PROPERTY QPL_LIB_DEPS
$<TARGET_OBJECTS:core_iaa>)
target_include_directories(core_iaa
PRIVATE ${UUID_DIR}
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/core-iaa/include>
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/core-iaa/sources/include>
PRIVATE $<BUILD_INTERFACE:${QPL_PROJECT_DIR}/include> # status.h in own_checkers.h
PRIVATE $<TARGET_PROPERTY:qpl_c_api,INTERFACE_INCLUDE_DIRECTORIES> # for own_checkers.h
PRIVATE $<TARGET_PROPERTY:qplcore_sw_dispatcher,INTERFACE_INCLUDE_DIRECTORIES>)
target_compile_features(core_iaa PRIVATE c_std_11)
target_compile_definitions(core_iaa PRIVATE QPL_BADARG_CHECK
PRIVATE $<$<BOOL:${LOG_HW_INIT}>:LOG_HW_INIT>
PRIVATE $<$<BOOL:${DYNAMIC_LOADING_LIBACCEL_CONFIG}>:DYNAMIC_LOADING_LIBACCEL_CONFIG>)
# [SUBDIR]middle-layer
file(GLOB MIDDLE_LAYER_SRC
${QPL_SRC_DIR}/middle-layer/accelerator/*.cpp
${QPL_SRC_DIR}/middle-layer/analytics/*.cpp
${QPL_SRC_DIR}/middle-layer/common/*.cpp
${QPL_SRC_DIR}/middle-layer/compression/*.cpp
${QPL_SRC_DIR}/middle-layer/compression/*/*.cpp
${QPL_SRC_DIR}/middle-layer/compression/*/*/*.cpp
${QPL_SRC_DIR}/middle-layer/dispatcher/*.cpp
${QPL_SRC_DIR}/middle-layer/other/*.cpp
${QPL_SRC_DIR}/middle-layer/util/*.cpp)
add_library(middle_layer_lib OBJECT
${MIDDLE_LAYER_SRC})
set_property(GLOBAL APPEND PROPERTY QPL_LIB_DEPS
$<TARGET_OBJECTS:middle_layer_lib>)
target_compile_options(middle_layer_lib
PRIVATE $<$<C_COMPILER_ID:GNU,Clang>:$<$<CONFIG:Release>:-O3;-U_FORTIFY_SOURCE;-D_FORTIFY_SOURCE=2>>
PRIVATE ${QPL_LINUX_TOOLCHAIN_CPP_EMBEDDED_FLAGS})
target_compile_definitions(middle_layer_lib
PUBLIC QPL_VERSION="${QPL_VERSION}"
PUBLIC $<$<BOOL:${LOG_HW_INIT}>:LOG_HW_INIT>
PUBLIC $<$<BOOL:${EFFICIENT_WAIT}>:QPL_EFFICIENT_WAIT>
PUBLIC QPL_BADARG_CHECK
PUBLIC $<$<BOOL:${DYNAMIC_LOADING_LIBACCEL_CONFIG}>:DYNAMIC_LOADING_LIBACCEL_CONFIG>)
set_target_properties(middle_layer_lib PROPERTIES CXX_STANDARD 17)
target_include_directories(middle_layer_lib
PRIVATE ${UUID_DIR}
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/middle-layer>
PUBLIC $<TARGET_PROPERTY:_qpl,INTERFACE_INCLUDE_DIRECTORIES>
PRIVATE $<TARGET_PROPERTY:qpl_c_api,INTERFACE_INCLUDE_DIRECTORIES>
PUBLIC $<TARGET_PROPERTY:qplcore_sw_dispatcher,INTERFACE_INCLUDE_DIRECTORIES>
PUBLIC $<TARGET_PROPERTY:isal,INTERFACE_INCLUDE_DIRECTORIES>
PUBLIC $<TARGET_PROPERTY:core_iaa,INTERFACE_INCLUDE_DIRECTORIES>)
target_compile_definitions(middle_layer_lib PUBLIC -DQPL_LIB)
# [SUBDIR]c_api
file(GLOB QPL_C_API_SRC
${QPL_SRC_DIR}/c_api/compression_operations/*.c
${QPL_SRC_DIR}/c_api/compression_operations/*.cpp
${QPL_SRC_DIR}/c_api/filter_operations/*.cpp
${QPL_SRC_DIR}/c_api/legacy_hw_path/*.c
${QPL_SRC_DIR}/c_api/legacy_hw_path/*.cpp
${QPL_SRC_DIR}/c_api/other_operations/*.cpp
${QPL_SRC_DIR}/c_api/serialization/*.cpp
${QPL_SRC_DIR}/c_api/*.cpp)
add_library(qpl_c_api OBJECT ${QPL_C_API_SRC})
target_include_directories(qpl_c_api
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/c_api/>
PUBLIC $<BUILD_INTERFACE:${QPL_SRC_DIR}/include/> $<INSTALL_INTERFACE:include>
PRIVATE $<TARGET_PROPERTY:middle_layer_lib,INTERFACE_INCLUDE_DIRECTORIES>)
set_target_properties(qpl_c_api PROPERTIES
$<$<C_COMPILER_ID:GNU,Clang>:C_STANDARD 17
CXX_STANDARD 17)
target_compile_options(qpl_c_api
PRIVATE $<$<C_COMPILER_ID:GNU,Clang>:$<$<CONFIG:Release>:-O3;-U_FORTIFY_SOURCE;-D_FORTIFY_SOURCE=2>>
PRIVATE $<$<COMPILE_LANG_AND_ID:CXX,GNU,Clang>:${QPL_LINUX_TOOLCHAIN_CPP_EMBEDDED_FLAGS}>)
target_compile_definitions(qpl_c_api
PUBLIC -DQPL_BADARG_CHECK # own_checkers.h
PUBLIC -DQPL_LIB # needed for middle_layer_lib
PUBLIC $<$<BOOL:${LOG_HW_INIT}>:LOG_HW_INIT>) # needed for middle_layer_lib
set_property(GLOBAL APPEND PROPERTY QPL_LIB_DEPS
$<TARGET_OBJECTS:qpl_c_api>)
# Final _qpl target
get_property(LIB_DEPS GLOBAL PROPERTY QPL_LIB_DEPS)
add_library(_qpl STATIC ${LIB_DEPS})
target_include_directories(_qpl
PUBLIC $<BUILD_INTERFACE:${QPL_PROJECT_DIR}/include/> $<INSTALL_INTERFACE:include>)
target_link_libraries(_qpl
PRIVATE ch_contrib::accel-config)
target_include_directories(_qpl SYSTEM BEFORE
PUBLIC "${QPL_PROJECT_DIR}/include"
PUBLIC ${UUID_DIR})
add_library (ch_contrib::qpl ALIAS _qpl)

4
contrib/qpl-cmake/uuid/uuid.h
View File

@ -1,4 +0,0 @@
#ifndef _QPL_UUID_UUID_H
#define _QPL_UUID_UUID_H
typedef unsigned char uuid_t[16];
#endif /* _QPL_UUID_UUID_H */

2
contrib/usearch vendored

@ -1 +1 @@
Subproject commit d1d33eac94acd3b628e0b446c927ec3295ef63c7
Subproject commit 1706420acafbd83d852c512dcf343af0a4059e48

5
docker/test/integration/runner/Dockerfile
View File

@ -42,13 +42,12 @@ ENV TZ=Etc/UTC
RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
ENV DOCKER_CHANNEL stable
# Unpin the docker version after the release 24.0.3 is released
# https://github.com/moby/moby/issues/45770#issuecomment-1618255130
RUN curl -fsSL https://download.docker.com/linux/ubuntu/gpg | apt-key add - \
&& add-apt-repository "deb https://download.docker.com/linux/ubuntu $(lsb_release -c -s) ${DOCKER_CHANNEL}" \
&& apt-get update \
&& env DEBIAN_FRONTEND=noninteractive apt-get install --yes \
docker-ce='5:23.*' docker-compose-plugin='2.29.*' \
docker-ce="5:27.0.3*" \
docker-compose-plugin='2.29.*' \
&& rm -rf \
/var/lib/apt/lists/* \
/var/cache/debconf \

5
docker/test/integration/runner/misc/openldap/initialized.sh Executable file
View File

@ -0,0 +1,5 @@
#!/bin/bash
set -e
# workaround for https://github.com/bitnami/containers/issues/73310
touch /tmp/.openldap-initialized

2
docker/test/integration/runner/requirements.txt
View File

@ -23,6 +23,7 @@ charset-normalizer==3.3.2
click==8.1.7
confluent-kafka==2.3.0
cryptography==42.0.0
datacompy==0.7.3
dbus-python==1.2.18
delta-spark==2.3.0
deltalake==0.16.0
@ -60,6 +61,7 @@ oauthlib==3.2.0
packaging==24.0
paramiko==3.4.0
pika==1.2.0
pandas==2.2.3
pip==24.1.1
pluggy==1.5.0
protobuf==4.25.2

1
docker/test/style/Dockerfile
View File

@ -16,6 +16,7 @@ RUN apt-get update && env DEBIAN_FRONTEND=noninteractive apt-get install --yes \
libxml2-utils \
locales \
moreutils \
ripgrep \
python3-pip \
yamllint \
zstd \

14
docker/test/style/requirements.txt
View File

@ -1,4 +1,4 @@
aiohttp==3.9.5
aiohttp==3.10.2
aiosignal==1.3.1
astroid==3.1.0
async-timeout==4.0.3
@ -6,12 +6,12 @@ attrs==23.2.0
black==24.4.2
boto3==1.34.131
botocore==1.34.131
certifi==2024.6.2
certifi==2024.07.04
cffi==1.16.0
charset-normalizer==3.3.2
click==8.1.7
codespell==2.2.1
cryptography==42.0.8
cryptography==43.0.1
Deprecated==1.2.14
dill==0.3.8
flake8==4.0.1
@ -19,7 +19,6 @@ frozenlist==1.4.1
idna==3.7
isort==5.13.2
jmespath==1.0.1
jwt==1.3.1
mccabe==0.6.1
multidict==6.0.5
mypy==1.8.0
@ -27,13 +26,12 @@ mypy-extensions==1.0.0
packaging==24.1
pathspec==0.9.0
pip==24.1.1
pipdeptree==2.23.0
platformdirs==4.2.2
pycodestyle==2.8.0
pycparser==2.22
pyflakes==2.4.0
PyGithub==2.3.0
PyJWT==2.8.0
PyJWT==2.9.0
pylint==3.1.0
PyNaCl==1.5.0
python-dateutil==2.9.0.post0
@ -42,7 +40,7 @@ PyYAML==6.0.1
rapidfuzz==3.9.3
requests==2.32.3
s3transfer==0.10.1
setuptools==59.6.0
setuptools==70.0.0
six==1.16.0
thefuzz==0.22.1
tomli==2.0.1
@ -52,7 +50,7 @@ types-requests==2.32.0.20240622
typing_extensions==4.12.2
unidiff==0.7.5
urllib3==2.2.2
wheel==0.37.1
wheel==0.38.1
wrapt==1.16.0
yamllint==1.26.3
yarl==1.9.4

7
docs/_includes/install/universal.sh
View File

@ -14,7 +14,12 @@ then
HAS_SSE42=$(grep sse4_2 /proc/cpuinfo)
if [ "${HAS_SSE42}" ]
then
DIR="amd64"
if ldd --version 2>&1 | grep -q musl
then
DIR="amd64musl"
else
DIR="amd64"
fi
else
DIR="amd64compat"
fi

3
docs/en/development/build-osx.md
View File

@ -47,6 +47,9 @@ git clone --recursive git@github.com:ClickHouse/ClickHouse.git
# ...alternatively, you can use https://github.com/ClickHouse/ClickHouse.git as the repo URL.
```
Apple uses a case-insensitive file system by default. While this usually does not affect compilation (especially scratch makes will work), it can confuse file operations like `git mv`.
For serious development on macOS, make sure that the source code is stored on a case-sensitive disk volume, e.g. see [these instructions](https://brianboyko.medium.com/a-case-sensitive-src-folder-for-mac-programmers-176cc82a3830).
## Build ClickHouse {#build-clickhouse}
To build using Homebrew's vanilla Clang compiler (the only **recommended** way):

327
docs/en/development/building_and_benchmarking_deflate_qpl.md
View File

@ -1,327 +0,0 @@
---
slug: /en/development/building_and_benchmarking_deflate_qpl
sidebar_position: 73
sidebar_label: Building and Benchmarking DEFLATE_QPL
description: How to build Clickhouse and run benchmark with DEFLATE_QPL Codec
---
# Build Clickhouse with DEFLATE_QPL
- Make sure your host machine meet the QPL required [prerequisites](https://intel.github.io/qpl/documentation/get_started_docs/installation.html#prerequisites)
- deflate_qpl is enabled by default during cmake build. In case you accidentally change it, please double-check build flag: ENABLE_QPL=1
- For generic requirements, please refer to Clickhouse generic [build instructions](/docs/en/development/build.md)
# Run Benchmark with DEFLATE_QPL
## Files list
The folders `benchmark_sample` under [qpl-cmake](https://github.com/ClickHouse/ClickHouse/tree/master/contrib/qpl-cmake) give example to run benchmark with python scripts:
`client_scripts` contains python scripts for running typical benchmark, for example:
- `client_stressing_test.py`: The python script for query stress test with [1~4] server instances.
- `queries_ssb.sql`: The file lists all queries for [Star Schema Benchmark](https://clickhouse.com/docs/en/getting-started/example-datasets/star-schema/)
- `allin1_ssb.sh`: This shell script executes benchmark workflow all in one automatically.
`database_files` means it will store database files according to lz4/deflate/zstd codec.
## Run benchmark automatically for Star Schema:
``` bash
$ cd ./benchmark_sample/client_scripts
$ sh run_ssb.sh
```
After complete, please check all the results in this folder:`./output/`
In case you run into failure, please manually run benchmark as below sections.
## Definition
[CLICKHOUSE_EXE] means the path of clickhouse executable program.
## Environment
- CPU: Sapphire Rapid
- OS Requirements refer to [System Requirements for QPL](https://intel.github.io/qpl/documentation/get_started_docs/installation.html#system-requirements)
- IAA Setup refer to [Accelerator Configuration](https://intel.github.io/qpl/documentation/get_started_docs/installation.html#accelerator-configuration)
- Install python modules:
``` bash
pip3 install clickhouse_driver numpy
```
[Self-check for IAA]
``` bash
$ accel-config list | grep -P 'iax|state'
```
Expected output like this:
``` bash
"dev":"iax1",
"state":"enabled",
"state":"enabled",
```
If you see nothing output, it means IAA is not ready to work. Please check IAA setup again.
## Generate raw data
``` bash
$ cd ./benchmark_sample
$ mkdir rawdata_dir && cd rawdata_dir
```
Use [`dbgen`](https://clickhouse.com/docs/en/getting-started/example-datasets/star-schema) to generate 100 million rows data with the parameters:
-s 20
The files like `*.tbl` are expected to output under `./benchmark_sample/rawdata_dir/ssb-dbgen`:
## Database setup
Set up database with LZ4 codec
``` bash
$ cd ./database_dir/lz4
$ [CLICKHOUSE_EXE] server -C config_lz4.xml >&/dev/null&
$ [CLICKHOUSE_EXE] client
```
Here you should see the message `Connected to ClickHouse server` from console which means client successfully setup connection with server.
Complete below three steps mentioned in [Star Schema Benchmark](https://clickhouse.com/docs/en/getting-started/example-datasets/star-schema)
- Creating tables in ClickHouse
- Inserting data. Here should use `./benchmark_sample/rawdata_dir/ssb-dbgen/*.tbl` as input data.
- Converting “star schema” to de-normalized “flat schema”
Set up database with IAA Deflate codec
``` bash
$ cd ./database_dir/deflate
$ [CLICKHOUSE_EXE] server -C config_deflate.xml >&/dev/null&
$ [CLICKHOUSE_EXE] client
```
Complete three steps same as lz4 above
Set up database with ZSTD codec
``` bash
$ cd ./database_dir/zstd
$ [CLICKHOUSE_EXE] server -C config_zstd.xml >&/dev/null&
$ [CLICKHOUSE_EXE] client
```
Complete three steps same as lz4 above
[self-check]
For each codec(lz4/zstd/deflate), please execute below query to make sure the databases are created successfully:
```sql
select count() from lineorder_flat
```
You are expected to see below output:
```sql
┌───count()─┐
│ 119994608 │
└───────────┘
```
[Self-check for IAA Deflate codec]
At the first time you execute insertion or query from client, clickhouse server console is expected to print this log:
```text
Hardware-assisted DeflateQpl codec is ready!
```
If you never find this, but see another log as below:
```text
Initialization of hardware-assisted DeflateQpl codec failed
```
That means IAA devices is not ready, you need check IAA setup again.
## Benchmark with single instance
- Before start benchmark, Please disable C6 and set CPU frequency governor to be `performance`
``` bash
$ cpupower idle-set -d 3
$ cpupower frequency-set -g performance
```
- To eliminate impact of memory bound on cross sockets, we use `numactl` to bind server on one socket and client on another socket.
- Single instance means single server connected with single client
Now run benchmark for LZ4/Deflate/ZSTD respectively:
LZ4:
``` bash
$ cd ./database_dir/lz4
$ numactl -m 0 -N 0 [CLICKHOUSE_EXE] server -C config_lz4.xml >&/dev/null&
$ cd ./client_scripts
$ numactl -m 1 -N 1 python3 client_stressing_test.py queries_ssb.sql 1 > lz4.log
```
IAA deflate:
``` bash
$ cd ./database_dir/deflate
$ numactl -m 0 -N 0 [CLICKHOUSE_EXE] server -C config_deflate.xml >&/dev/null&
$ cd ./client_scripts
$ numactl -m 1 -N 1 python3 client_stressing_test.py queries_ssb.sql 1 > deflate.log
```
ZSTD:
``` bash
$ cd ./database_dir/zstd
$ numactl -m 0 -N 0 [CLICKHOUSE_EXE] server -C config_zstd.xml >&/dev/null&
$ cd ./client_scripts
$ numactl -m 1 -N 1 python3 client_stressing_test.py queries_ssb.sql 1 > zstd.log
```
Now three logs should be output as expected:
```text
lz4.log
deflate.log
zstd.log
```
How to check performance metrics:
We focus on QPS, please search the keyword: `QPS_Final` and collect statistics
## Benchmark with multi-instances
- To reduce impact of memory bound on too much threads, We recommend run benchmark with multi-instances.
- Multi-instance means multiple（2 or 4）servers connected with respective client.
- The cores of one socket need to be divided equally and assigned to the servers respectively.
- For multi-instances, must create new folder for each codec and insert dataset by following the similar steps as single instance.
There are 2 differences:
- For client side, you need launch clickhouse with the assigned port during table creation and data insertion.
- For server side, you need launch clickhouse with the specific xml config file in which port has been assigned. All customized xml config files for multi-instances has been provided under ./server_config.
Here we assume there are 60 cores per socket and take 2 instances for example.
Launch server for first instance
LZ4:
``` bash
$ cd ./database_dir/lz4
$ numactl -C 0-29,120-149 [CLICKHOUSE_EXE] server -C config_lz4.xml >&/dev/null&
```
ZSTD:
``` bash
$ cd ./database_dir/zstd
$ numactl -C 0-29,120-149 [CLICKHOUSE_EXE] server -C config_zstd.xml >&/dev/null&
```
IAA Deflate:
``` bash
$ cd ./database_dir/deflate
$ numactl -C 0-29,120-149 [CLICKHOUSE_EXE] server -C config_deflate.xml >&/dev/null&
```
[Launch server for second instance]
LZ4:
``` bash
$ cd ./database_dir && mkdir lz4_s2 && cd lz4_s2
$ cp ../../server_config/config_lz4_s2.xml ./
$ numactl -C 30-59,150-179 [CLICKHOUSE_EXE] server -C config_lz4_s2.xml >&/dev/null&
```
ZSTD:
``` bash
$ cd ./database_dir && mkdir zstd_s2 && cd zstd_s2
$ cp ../../server_config/config_zstd_s2.xml ./
$ numactl -C 30-59,150-179 [CLICKHOUSE_EXE] server -C config_zstd_s2.xml >&/dev/null&
```
IAA Deflate:
``` bash
$ cd ./database_dir && mkdir deflate_s2 && cd deflate_s2
$ cp ../../server_config/config_deflate_s2.xml ./
$ numactl -C 30-59,150-179 [CLICKHOUSE_EXE] server -C config_deflate_s2.xml >&/dev/null&
```
Creating tables && Inserting data for second instance
Creating tables:
``` bash
$ [CLICKHOUSE_EXE] client -m --port=9001
```
Inserting data:
``` bash
$ [CLICKHOUSE_EXE] client --query "INSERT INTO [TBL_FILE_NAME] FORMAT CSV" < [TBL_FILE_NAME].tbl --port=9001
```
- [TBL_FILE_NAME] represents the name of a file named with the regular expression: *. tbl under `./benchmark_sample/rawdata_dir/ssb-dbgen`.
- `--port=9001` stands for the assigned port for server instance which is also defined in config_lz4_s2.xml/config_zstd_s2.xml/config_deflate_s2.xml. For even more instances, you need replace it with the value: 9002/9003 which stand for s3/s4 instance respectively. If you don't assign it, the port is 9000 by default which has been used by first instance.
Benchmarking with 2 instances
LZ4:
``` bash
$ cd ./database_dir/lz4
$ numactl -C 0-29,120-149 [CLICKHOUSE_EXE] server -C config_lz4.xml >&/dev/null&
$ cd ./database_dir/lz4_s2
$ numactl -C 30-59,150-179 [CLICKHOUSE_EXE] server -C config_lz4_s2.xml >&/dev/null&
$ cd ./client_scripts
$ numactl -m 1 -N 1 python3 client_stressing_test.py queries_ssb.sql 2 > lz4_2insts.log
```
ZSTD:
``` bash
$ cd ./database_dir/zstd
$ numactl -C 0-29,120-149 [CLICKHOUSE_EXE] server -C config_zstd.xml >&/dev/null&
$ cd ./database_dir/zstd_s2
$ numactl -C 30-59,150-179 [CLICKHOUSE_EXE] server -C config_zstd_s2.xml >&/dev/null&
$ cd ./client_scripts
$ numactl -m 1 -N 1 python3 client_stressing_test.py queries_ssb.sql 2 > zstd_2insts.log
```
IAA deflate
``` bash
$ cd ./database_dir/deflate
$ numactl -C 0-29,120-149 [CLICKHOUSE_EXE] server -C config_deflate.xml >&/dev/null&
$ cd ./database_dir/deflate_s2
$ numactl -C 30-59,150-179 [CLICKHOUSE_EXE] server -C config_deflate_s2.xml >&/dev/null&
$ cd ./client_scripts
$ numactl -m 1 -N 1 python3 client_stressing_test.py queries_ssb.sql 2 > deflate_2insts.log
```
Here the last argument: `2` of client_stressing_test.py stands for the number of instances. For more instances, you need replace it with the value: 3 or 4. This script support up to 4 instances/
Now three logs should be output as expected:
``` text
lz4_2insts.log
deflate_2insts.log
zstd_2insts.log
```
How to check performance metrics:
We focus on QPS, please search the keyword: `QPS_Final` and collect statistics
Benchmark setup for 4 instances is similar with 2 instances above.
We recommend use 2 instances benchmark data as final report for review.
## Tips
Each time before launch new clickhouse server, please make sure no background clickhouse process running, please check and kill old one:
``` bash
$ ps -aux| grep clickhouse
$ kill -9 [PID]
```
By comparing the query list in ./client_scripts/queries_ssb.sql with official [Star Schema Benchmark](https://clickhouse.com/docs/en/getting-started/example-datasets/star-schema), you will find 3 queries are not included: Q1.2/Q1.3/Q3.4 . This is because cpu utilization% is very low <10% for these queries which means cannot demonstrate performance differences.

2
docs/en/development/contrib.md
View File

@ -18,7 +18,7 @@ SELECT library_name, license_type, license_path FROM system.licenses ORDER BY li
Note that the listed libraries are the ones located in the `contrib/` directory of the ClickHouse repository.
Depending on the build options, some of the libraries may have not been compiled, and, as a result, their functionality may not be available at runtime.
[Example](https://play.clickhouse.com/play?user=play#U0VMRUNUIGxpYnJhcnlfbmFtZSwgbGljZW5zZV90eXBlLCBsaWNlbnNlX3BhdGggRlJPTSBzeXN0ZW0ubGljZW5zZXMgT1JERVIgQlkgbGlicmFyeV9uYW1lIENPTExBVEUgJ2VuJw==)
[Example](https://sql.clickhouse.com?query_id=478GCPU7LRTSZJBNY3EJT3)
## Adding and maintaining third-party libraries

312
docs/en/engines/database-engines/materialized-mysql.md
View File

@ -7,315 +7,5 @@ sidebar_position: 70
# [experimental] MaterializedMySQL
:::note
This database engine is experimental. To use it, set `allow_experimental_database_materialized_mysql` to 1 in your configuration files or by using the `SET` command:
```sql
SET allow_experimental_database_materialized_mysql=1
```
This database engine is obsolete and cannot be used.
:::
Creates a ClickHouse database with all the tables existing in MySQL, and all the data in those tables. The ClickHouse server works as MySQL replica. It reads `binlog` and performs DDL and DML queries.
## Creating a Database {#creating-a-database}
``` sql
CREATE DATABASE [IF NOT EXISTS] db_name [ON CLUSTER cluster]
ENGINE = MaterializedMySQL('host:port', ['database' | database], 'user', 'password') [SETTINGS ...]
[TABLE OVERRIDE table1 (...), TABLE OVERRIDE table2 (...)]
```
**Engine Parameters**
- `host:port` — MySQL server endpoint.
- `database` — MySQL database name.
- `user` — MySQL user.
- `password` — User password.
## Engine Settings
### max_rows_in_buffer
`max_rows_in_buffer` — Maximum number of rows that data is allowed to cache in memory (for single table and the cache data unable to query). When this number is exceeded, the data will be materialized. Default: `65 505`.
### max_bytes_in_buffer
`max_bytes_in_buffer` — Maximum number of bytes that data is allowed to cache in memory (for single table and the cache data unable to query). When this number is exceeded, the data will be materialized. Default: `1 048 576`.
### max_flush_data_time
`max_flush_data_time` — Maximum number of milliseconds that data is allowed to cache in memory (for database and the cache data unable to query). When this time is exceeded, the data will be materialized. Default: `1000`.
### max_wait_time_when_mysql_unavailable
`max_wait_time_when_mysql_unavailable` — Retry interval when MySQL is not available (milliseconds). Negative value disables retry. Default: `1000`.
### allows_query_when_mysql_lost
`allows_query_when_mysql_lost` — Allows to query a materialized table when MySQL is lost. Default: `0` (`false`).
### allow_startup_database_without_connection_to_mysql
`allow_startup_database_without_connection_to_mysql` — Allow to create and attach database without available connection to MySQL. Default: `0` (`false`).
### materialized_mysql_tables_list
`materialized_mysql_tables_list` — a comma-separated list of mysql database tables, which will be replicated by MaterializedMySQL database engine. Default value: empty list — means whole tables will be replicated.
```sql
CREATE DATABASE mysql ENGINE = MaterializedMySQL('localhost:3306', 'db', 'user', '***')
SETTINGS
allows_query_when_mysql_lost=true,
max_wait_time_when_mysql_unavailable=10000;
```
## Settings on MySQL-server Side
For the correct work of `MaterializedMySQL`, there are few mandatory `MySQL`-side configuration settings that must be set:
### default_authentication_plugin
`default_authentication_plugin = mysql_native_password` since `MaterializedMySQL` can only authorize with this method.
### gtid_mode
`gtid_mode = on` since GTID based logging is a mandatory for providing correct `MaterializedMySQL` replication.
:::note
While turning on `gtid_mode` you should also specify `enforce_gtid_consistency = on`.
:::
## Virtual Columns {#virtual-columns}
When working with the `MaterializedMySQL` database engine, [ReplacingMergeTree](/docs/en/engines/table-engines/mergetree-family/replacingmergetree.md) tables are used with virtual `_sign` and `_version` columns.
### \_version
`_version` — Transaction counter. Type [UInt64](/docs/en/sql-reference/data-types/int-uint.md).
### \_sign
`_sign` — Deletion mark. Type [Int8](/docs/en/sql-reference/data-types/int-uint.md). Possible values:
- `1` — Row is not deleted,
- `-1` — Row is deleted.
## Data Types Support {#data_types-support}
| MySQL | ClickHouse |
|-------------------------|--------------------------------------------------------------|
| TINY | [Int8](/docs/en/sql-reference/data-types/int-uint.md) |
| SHORT | [Int16](/docs/en/sql-reference/data-types/int-uint.md) |
| INT24 | [Int32](/docs/en/sql-reference/data-types/int-uint.md) |
| LONG | [UInt32](/docs/en/sql-reference/data-types/int-uint.md) |
| LONGLONG | [UInt64](/docs/en/sql-reference/data-types/int-uint.md) |
| FLOAT | [Float32](/docs/en/sql-reference/data-types/float.md) |
| DOUBLE | [Float64](/docs/en/sql-reference/data-types/float.md) |
| DECIMAL, NEWDECIMAL | [Decimal](/docs/en/sql-reference/data-types/decimal.md) |
| DATE, NEWDATE | [Date](/docs/en/sql-reference/data-types/date.md) |
| DATETIME, TIMESTAMP | [DateTime](/docs/en/sql-reference/data-types/datetime.md) |
| DATETIME2, TIMESTAMP2 | [DateTime64](/docs/en/sql-reference/data-types/datetime64.md) |
| YEAR | [UInt16](/docs/en/sql-reference/data-types/int-uint.md) |
| TIME | [Int64](/docs/en/sql-reference/data-types/int-uint.md) |
| ENUM | [Enum](/docs/en/sql-reference/data-types/enum.md) |
| STRING | [String](/docs/en/sql-reference/data-types/string.md) |
| VARCHAR, VAR_STRING | [String](/docs/en/sql-reference/data-types/string.md) |
| BLOB | [String](/docs/en/sql-reference/data-types/string.md) |
| GEOMETRY | [String](/docs/en/sql-reference/data-types/string.md) |
| BINARY | [FixedString](/docs/en/sql-reference/data-types/fixedstring.md) |
| BIT | [UInt64](/docs/en/sql-reference/data-types/int-uint.md) |
| SET | [UInt64](/docs/en/sql-reference/data-types/int-uint.md) |
[Nullable](/docs/en/sql-reference/data-types/nullable.md) is supported.
The data of TIME type in MySQL is converted to microseconds in ClickHouse.
Other types are not supported. If MySQL table contains a column of such type, ClickHouse throws an exception and stops replication.
## Specifics and Recommendations {#specifics-and-recommendations}
### Compatibility Restrictions {#compatibility-restrictions}
Apart of the data types limitations there are few restrictions comparing to `MySQL` databases, that should be resolved before replication will be possible:
- Each table in `MySQL` should contain `PRIMARY KEY`.
- Replication for tables, those are containing rows with `ENUM` field values out of range (specified in `ENUM` signature) will not work.
### DDL Queries {#ddl-queries}
MySQL DDL queries are converted into the corresponding ClickHouse DDL queries ([ALTER](/docs/en/sql-reference/statements/alter/index.md), [CREATE](/docs/en/sql-reference/statements/create/index.md), [DROP](/docs/en/sql-reference/statements/drop.md), [RENAME](/docs/en/sql-reference/statements/rename.md)). If ClickHouse cannot parse some DDL query, the query is ignored.
### Data Replication {#data-replication}
`MaterializedMySQL` does not support direct `INSERT`, `DELETE` and `UPDATE` queries. However, they are supported in terms of data replication:
- MySQL `INSERT` query is converted into `INSERT` with `_sign=1`.
- MySQL `DELETE` query is converted into `INSERT` with `_sign=-1`.
- MySQL `UPDATE` query is converted into `INSERT` with `_sign=-1` and `INSERT` with `_sign=1` if the primary key has been changed, or
`INSERT` with `_sign=1` if not.
### Selecting from MaterializedMySQL Tables {#select}
`SELECT` query from `MaterializedMySQL` tables has some specifics:
- If `_version` is not specified in the `SELECT` query, the
[FINAL](/docs/en/sql-reference/statements/select/from.md/#select-from-final) modifier is used, so only rows with
`MAX(_version)` are returned for each primary key value.
- If `_sign` is not specified in the `SELECT` query, `WHERE _sign=1` is used by default. So the deleted rows are not
included into the result set.
- The result includes columns comments in case they exist in MySQL database tables.
### Index Conversion {#index-conversion}
MySQL `PRIMARY KEY` and `INDEX` clauses are converted into `ORDER BY` tuples in ClickHouse tables.
ClickHouse has only one physical order, which is determined by `ORDER BY` clause. To create a new physical order, use
[materialized views](/docs/en/sql-reference/statements/create/view.md/#materialized).
**Notes**
- Rows with `_sign=-1` are not deleted physically from the tables.
- Cascade `UPDATE/DELETE` queries are not supported by the `MaterializedMySQL` engine, as they are not visible in the
MySQL binlog.
- Replication can be easily broken.
- Manual operations on database and tables are forbidden.
- `MaterializedMySQL` is affected by the [optimize_on_insert](/docs/en/operations/settings/settings.md/#optimize-on-insert)
setting. Data is merged in the corresponding table in the `MaterializedMySQL` database when a table in the MySQL
server changes.
### Table Overrides {#table-overrides}
Table overrides can be used to customize the ClickHouse DDL queries, allowing you to make schema optimizations for your
application. This is especially useful for controlling partitioning, which is important for the overall performance of
MaterializedMySQL.
These are the schema conversion manipulations you can do with table overrides for MaterializedMySQL:
* Modify column type. Must be compatible with the original type, or replication will fail. For example,
you can modify a UInt32 column to UInt64, but you can not modify a String column to Array(String).
* Modify [column TTL](/docs/en/engines/table-engines/mergetree-family/mergetree.md/#mergetree-column-ttl).
* Modify [column compression codec](/docs/en/sql-reference/statements/create/table.md/#codecs).
* Add [ALIAS columns](/docs/en/sql-reference/statements/create/table.md/#alias).
* Add [skipping indexes](/docs/en/engines/table-engines/mergetree-family/mergetree.md/#table_engine-mergetree-data_skipping-indexes). Note that you need to enable `use_skip_indexes_if_final` setting to make them work (MaterializedMySQL is using `SELECT ... FINAL` by default)
* Add [projections](/docs/en/engines/table-engines/mergetree-family/mergetree.md/#projections). Note that projection optimizations are
disabled when using `SELECT ... FINAL` (which MaterializedMySQL does by default), so their utility is limited here.
`INDEX ... TYPE hypothesis` as [described in the v21.12 blog post]](https://clickhouse.com/blog/en/2021/clickhouse-v21.12-released/)
may be more useful in this case.
* Modify [PARTITION BY](/docs/en/engines/table-engines/mergetree-family/custom-partitioning-key/)
* Modify [ORDER BY](/docs/en/engines/table-engines/mergetree-family/mergetree.md/#mergetree-query-clauses)
* Modify [PRIMARY KEY](/docs/en/engines/table-engines/mergetree-family/mergetree.md/#mergetree-query-clauses)
* Add [SAMPLE BY](/docs/en/engines/table-engines/mergetree-family/mergetree.md/#mergetree-query-clauses)
* Add [table TTL](/docs/en/engines/table-engines/mergetree-family/mergetree.md/#mergetree-query-clauses)
```sql
CREATE DATABASE db_name ENGINE = MaterializedMySQL(...)
[SETTINGS ...]
[TABLE OVERRIDE table_name (
[COLUMNS (
[col_name [datatype] [ALIAS expr] [CODEC(...)] [TTL expr], ...]
[INDEX index_name expr TYPE indextype[(...)] GRANULARITY val, ...]
[PROJECTION projection_name (SELECT <COLUMN LIST EXPR> [GROUP BY] [ORDER BY]), ...]
)]
[ORDER BY expr]
[PRIMARY KEY expr]
[PARTITION BY expr]
[SAMPLE BY expr]
[TTL expr]
), ...]
```
Example:
```sql
CREATE DATABASE db_name ENGINE = MaterializedMySQL(...)
TABLE OVERRIDE table1 (
COLUMNS (
userid UUID,
category LowCardinality(String),
timestamp DateTime CODEC(Delta, Default)
)
PARTITION BY toYear(timestamp)
),
TABLE OVERRIDE table2 (
COLUMNS (
client_ip String TTL created + INTERVAL 72 HOUR
)
SAMPLE BY ip_hash
)
```
The `COLUMNS` list is sparse; existing columns are modified as specified, extra ALIAS columns are added. It is not
possible to add ordinary or MATERIALIZED columns. Modified columns with a different type must be assignable from the
original type. There is currently no validation of this or similar issues when the `CREATE DATABASE` query executes, so
extra care needs to be taken.
You may specify overrides for tables that do not exist yet.
:::important
It is easy to break replication with table overrides if not used with care. For example:
* If an ALIAS column is added with a table override, and a column with the same name is later added to the source
MySQL table, the converted ALTER TABLE query in ClickHouse will fail and replication stops.
* It is currently possible to add overrides that reference nullable columns where not-nullable are required, such as in
`ORDER BY` or `PARTITION BY`. This will cause CREATE TABLE queries that will fail, also causing replication to stop.
:::
## Examples of Use {#examples-of-use}
Queries in MySQL:
``` sql
mysql> CREATE DATABASE db;
mysql> CREATE TABLE db.test (a INT PRIMARY KEY, b INT);
mysql> INSERT INTO db.test VALUES (1, 11), (2, 22);
mysql> DELETE FROM db.test WHERE a=1;
mysql> ALTER TABLE db.test ADD COLUMN c VARCHAR(16);
mysql> UPDATE db.test SET c='Wow!', b=222;
mysql> SELECT * FROM test;
```
```text
┌─a─┬───b─┬─c────┐
│ 2 │ 222 │ Wow! │
└───┴─────┴──────┘
```
Database in ClickHouse, exchanging data with the MySQL server:
The database and the table created:
``` sql
CREATE DATABASE mysql ENGINE = MaterializedMySQL('localhost:3306', 'db', 'user', '***');
SHOW TABLES FROM mysql;
```
``` text
┌─name─┐
│ test │
└──────┘
```
After inserting data:
``` sql
SELECT * FROM mysql.test;
```
``` text
┌─a─┬──b─┐
│ 1 │ 11 │
│ 2 │ 22 │
└───┴────┘
```
After deleting data, adding the column and updating:
``` sql
SELECT * FROM mysql.test;
```
``` text
┌─a─┬───b─┬─c────┐
│ 2 │ 222 │ Wow! │
└───┴─────┴──────┘
```

1
docs/en/engines/table-engines/integrations/azure-queue.md
View File

@ -36,6 +36,7 @@ SETTINGS
## Settings {#settings}
The set of supported settings is the same as for `S3Queue` table engine, but without `s3queue_` prefix. See [full list of settings settings](../../../engines/table-engines/integrations/s3queue.md#settings).
To get a list of settings, configured for the table, use `system.s3_queue_settings` table. Available from `24.10`.
## Description {#description}

2
docs/en/engines/table-engines/integrations/mongodb.md
View File

@ -9,7 +9,7 @@ sidebar_label: MongoDB
MongoDB engine is read-only table engine which allows to read data from remote [MongoDB](https://www.mongodb.com/) collection.
Only MongoDB v3.6+ servers are supported.
[Seed list(`mongodb**+srv**`)](https://www.mongodb.com/docs/manual/reference/glossary/#std-term-seed-list) is not yet supported.
[Seed list(`mongodb+srv`)](https://www.mongodb.com/docs/manual/reference/glossary/#std-term-seed-list) is not yet supported.
:::note
If you're facing troubles, please report the issue, and try to use [the legacy implementation](../../../operations/server-configuration-parameters/settings.md#use_legacy_mongodb_integration).

4
docs/en/engines/table-engines/integrations/s3.md
View File

@ -331,6 +331,10 @@ CREATE TABLE big_table (name String, value UInt32)
ENGINE = S3('https://datasets-documentation.s3.eu-west-3.amazonaws.com/aapl_stock.csv', NOSIGN, 'CSVWithNames');
```
## Optimizing performance
For details on optimizing the performance of the s3 function see [our detailed guide](/docs/en/integrations/s3/performance).
## See also
- [s3 table function](../../../sql-reference/table-functions/s3.md)

2
docs/en/engines/table-engines/integrations/s3queue.md
View File

@ -69,6 +69,8 @@ SETTINGS
## Settings {#settings}
To get a list of settings, configured for the table, use `system.s3_queue_settings` table. Available from `24.10`.
### mode {#mode}
Possible values:

24
docs/en/engines/table-engines/mergetree-family/aggregatingmergetree.md
View File

@ -37,7 +37,7 @@ For a description of request parameters, see [request description](../../../sql-
**Query clauses**
When creating an `AggregatingMergeTree` table the same [clauses](../../../engines/table-engines/mergetree-family/mergetree.md) are required, as when creating a `MergeTree` table.
When creating an `AggregatingMergeTree` table, the same [clauses](../../../engines/table-engines/mergetree-family/mergetree.md) are required as when creating a `MergeTree` table.
<details markdown="1">
@ -62,19 +62,19 @@ All of the parameters have the same meaning as in `MergeTree`.
## SELECT and INSERT {#select-and-insert}
To insert data, use [INSERT SELECT](../../../sql-reference/statements/insert-into.md) query with aggregate -State- functions.
When selecting data from `AggregatingMergeTree` table, use `GROUP BY` clause and the same aggregate functions as when inserting data, but using `-Merge` suffix.
When selecting data from `AggregatingMergeTree` table, use `GROUP BY` clause and the same aggregate functions as when inserting data, but using the `-Merge` suffix.
In the results of `SELECT` query, the values of `AggregateFunction` type have implementation-specific binary representation for all of the ClickHouse output formats. If dump data into, for example, `TabSeparated` format with `SELECT` query then this dump can be loaded back using `INSERT` query.
In the results of `SELECT` query, the values of `AggregateFunction` type have implementation-specific binary representation for all of the ClickHouse output formats. For example, if you dump data into `TabSeparated` format with a `SELECT` query, then this dump can be loaded back using an `INSERT` query.
## Example of an Aggregated Materialized View {#example-of-an-aggregated-materialized-view}
The following examples assumes that you have a database named `test` so make sure you create that if it doesn't already exist:
The following example assumes that you have a database named `test`, so create it if it doesn't already exist:
```sql
CREATE DATABASE test;
```
We will create the table `test.visits` that contain the raw data:
Now create the table `test.visits` that contains the raw data:
``` sql
CREATE TABLE test.visits
@ -86,9 +86,9 @@ CREATE TABLE test.visits
) ENGINE = MergeTree ORDER BY (StartDate, CounterID);
```
Next, we need to create an `AggregatingMergeTree` table that will store `AggregationFunction`s that keep track of the total number of visits and the number of unique users.
Next, you need an `AggregatingMergeTree` table that will store `AggregationFunction`s that keep track of the total number of visits and the number of unique users.
`AggregatingMergeTree` materialized view that watches the `test.visits` table, and use the `AggregateFunction` type:
Create an `AggregatingMergeTree` materialized view that watches the `test.visits` table, and uses the `AggregateFunction` type:
``` sql
CREATE TABLE test.agg_visits (
@ -100,7 +100,7 @@ CREATE TABLE test.agg_visits (
ENGINE = AggregatingMergeTree() ORDER BY (StartDate, CounterID);
```
And then let's create a materialized view that populates `test.agg_visits` from `test.visits` :
Create a materialized view that populates `test.agg_visits` from `test.visits`:
```sql
CREATE MATERIALIZED VIEW test.visits_mv TO test.agg_visits
@ -113,7 +113,7 @@ FROM test.visits
GROUP BY StartDate, CounterID;
```
Inserting data into the `test.visits` table.
Insert data into the `test.visits` table:
``` sql
INSERT INTO test.visits (StartDate, CounterID, Sign, UserID)
@ -122,7 +122,7 @@ INSERT INTO test.visits (StartDate, CounterID, Sign, UserID)
The data is inserted in both `test.visits` and `test.agg_visits`.
To get the aggregated data, we need to execute a query such as `SELECT ... GROUP BY ...` from the materialized view `test.mv_visits`:
To get the aggregated data, execute a query such as `SELECT ... GROUP BY ...` from the materialized view `test.mv_visits`:
```sql
SELECT
@ -140,14 +140,14 @@ ORDER BY StartDate;
└─────────────────────────┴────────┴───────┘
```
And how about if we add another couple of records to `test.visits`, but this time we'll use a different timestamp for one of the records:
Add another couple of records to `test.visits`, but this time try using a different timestamp for one of the records:
```sql
INSERT INTO test.visits (StartDate, CounterID, Sign, UserID)
VALUES (1669446031000, 2, 5, 10), (1667446031000, 3, 7, 5);
```
If we then run the `SELECT` query again, we'll see the following output:
Run the `SELECT` query again, which will return the following output:
```text
┌───────────────StartDate─┬─Visits─┬─Users─┐

22
docs/en/engines/table-engines/mergetree-family/annindexes.md
View File

@ -43,7 +43,7 @@ CREATE TABLE table
(
id Int64,
vectors Array(Float32),
INDEX index_name vectors TYPE vector_similarity(method, distance_function[, quantization, connectivity, expansion_add, expansion_search]) [GRANULARITY N]
INDEX index_name vectors TYPE vector_similarity(method, distance_function[, quantization, hnsw_max_connections_per_layer, hnsw_candidate_list_size_for_construction]) [GRANULARITY N]
)
ENGINE = MergeTree
ORDER BY id;
@ -55,11 +55,13 @@ Parameters:
line between two points in Euclidean space), or `cosineDistance` (the [cosine
distance](https://en.wikipedia.org/wiki/Cosine_similarity#Cosine_distance)- the angle between two non-zero vectors).
- `quantization`: either `f64`, `f32`, `f16`, `bf16`, or `i8` for storing the vector with reduced precision (optional, default: `bf16`)
- `m`: the number of neighbors per graph node (optional, default: 16)
- `ef_construction`: (optional, default: 128)
- `ef_search`: (optional, default: 64)
- `hnsw_max_connections_per_layer`: the number of neighbors per HNSW graph node, also known as `M` in the [HNSW
paper](https://doi.org/10.1109/TPAMI.2018.2889473) (optional, default: 32)
- `hnsw_candidate_list_size_for_construction`: the size of the dynamic candidate list when constructing the HNSW graph, also known as
`ef_construction` in the original [HNSW paper](https://doi.org/10.1109/TPAMI.2018.2889473) (optional, default: 128)
Value 0 for parameters `m`, `ef_construction`, and `ef_search` refers to the default value.
Values 0 for parameters `hnsw_max_connections_per_layer` and `hnsw_candidate_list_size_for_construction` means using the default values of
these parameters.
Example:
@ -115,6 +117,11 @@ ANN indexes are built during column insertion and merge. As a result, `INSERT` a
tables. ANNIndexes are ideally used only with immutable or rarely changed data, respectively when are far more read requests than write
requests.
:::tip
To reduce the cost of building vector similarity indexes, consider setting `materialize_skip_indexes_on_insert` which disables the
construction of skipping indexes on newly inserted parts. Search would fall back to exact search but as inserted parts are typically small
compared to the total table size, the performance impact of that would be negligible.
ANN indexes support this type of query:
``` sql
@ -124,6 +131,7 @@ FROM table
WHERE ... -- WHERE clause is optional
ORDER BY Distance(vectors, reference_vector)
LIMIT N
SETTINGS enable_analyzer = 0; -- Temporary limitation, will be lifted
```
:::tip
@ -135,6 +143,10 @@ clickhouse-client --param_vec='hello' --query="SELECT * FROM table WHERE L2Dista
```
:::
To search using a different value of HNSW parameter `hnsw_candidate_list_size_for_search` (default: 256), also known as `ef_search` in the
original [HNSW paper](https://doi.org/10.1109/TPAMI.2018.2889473), run the `SELECT` query with `SETTINGS hnsw_candidate_list_size_for_search
= <value>`.
**Restrictions**: Approximate algorithms used to determine the nearest neighbors require a limit, hence queries without `LIMIT` clause
cannot utilize ANN indexes. Also, ANN indexes are only used if the query has a `LIMIT` value smaller than setting
`max_limit_for_ann_queries` (default: 1 million rows). This is a safeguard to prevent large memory allocations by external libraries for

52
docs/en/engines/table-engines/mergetree-family/replacingmergetree.md
View File

@ -12,6 +12,10 @@ Data deduplication occurs only during a merge. Merging occurs in the background
Thus, `ReplacingMergeTree` is suitable for clearing out duplicate data in the background in order to save space, but it does not guarantee the absence of duplicates.
:::note
A detailed guide on ReplacingMergeTree, including best practices and how to optimize performance, is available [here](/docs/en/guides/replacing-merge-tree).
:::
## Creating a Table {#creating-a-table}
``` sql
@ -162,3 +166,51 @@ All of the parameters excepting `ver` have the same meaning as in `MergeTree`.
- `ver` - column with the version. Optional parameter. For a description, see the text above.
</details>
## Query time de-duplication & FINAL
At merge time, the ReplacingMergeTree identifies duplicate rows, using the values of the `ORDER BY` columns (used to create the table) as a unique identifier, and retains only the highest version. This, however, offers eventual correctness only - it does not guarantee rows will be deduplicated, and you should not rely on it. Queries can, therefore, produce incorrect answers due to update and delete rows being considered in queries.
To obtain correct answers, users will need to complement background merges with query time deduplication and deletion removal. This can be achieved using the `FINAL` operator. For example, consider the following example:
```sql
CREATE TABLE rmt_example
(
`number` UInt16
)
ENGINE = ReplacingMergeTree
ORDER BY number
INSERT INTO rmt_example SELECT floor(randUniform(0, 100)) AS number
FROM numbers(1000000000)
0 rows in set. Elapsed: 19.958 sec. Processed 1.00 billion rows, 8.00 GB (50.11 million rows/s., 400.84 MB/s.)
```
Querying without `FINAL` produces an incorrect count (exact result will vary depending on merges):
```sql
SELECT count()
FROM rmt_example
┌─count()─┐
│ 200 │
└─────────┘
1 row in set. Elapsed: 0.002 sec.
```
Adding final produces a correct result:
```sql
SELECT count()
FROM rmt_example
FINAL
┌─count()─┐
│ 100 │
└─────────┘
1 row in set. Elapsed: 0.002 sec.
```
For further details on `FINAL`, including how to optimize `FINAL` performance, we recommend reading our [detailed guide on ReplacingMergeTree](/docs/en/guides/replacing-merge-tree).

2
docs/en/getting-started/example-datasets/brown-benchmark.md
View File

@ -453,4 +453,4 @@ ORDER BY yr,
mo;
```
The data is also available for interactive queries in the [Playground](https://play.clickhouse.com/play?user=play), [example](https://play.clickhouse.com/play?user=play#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).
The data is also available for interactive queries in the [Playground](https://sql.clickhouse.com), [example](https://sql.clickhouse.com?query_id=1MXMHASDLEQIP4P1D1STND).

4
docs/en/getting-started/example-datasets/cell-towers.md
View File

@ -360,9 +360,9 @@ This screenshot shows cell tower locations with LTE, UMTS, and GSM radios. The
![Dashboard of cell towers by radio type in mcc 204](@site/docs/en/getting-started/example-datasets/images/superset-cell-tower-dashboard.png)
:::tip
The data is also available for interactive queries in the [Playground](https://play.clickhouse.com/play?user=play).
The data is also available for interactive queries in the [Playground](https://sql.clickhouse.com).
This [example](https://play.clickhouse.com/play?user=play#U0VMRUNUIG1jYywgY291bnQoKSBGUk9NIGNlbGxfdG93ZXJzIEdST1VQIEJZIG1jYyBPUkRFUiBCWSBjb3VudCgpIERFU0M=) will populate the username and even the query for you.
This [example](https://sql.clickhouse.com?query_id=UV8M4MAGS2PWAUOAYAAARM) will populate the username and even the query for you.
Although you cannot create tables in the Playground, you can run all of the queries and even use Superset (adjust the host name and port number).
:::

88
docs/en/getting-started/example-datasets/github.md
View File

@ -244,13 +244,13 @@ FROM s3('https://datasets-documentation.s3.amazonaws.com/github/commits/clickhou
The tool suggests several queries via its help output. We have answered these in addition to some additional supplementary questions of interest. These queries are of approximately increasing complexity vs. the tool's arbitrary order.
This dataset is available in [play.clickhouse.com](https://play.clickhouse.com/play?user=play#U0hPVyBUQUJMRVMgSU4gZ2l0X2NsaWNraG91c2U=) in the `git_clickhouse` databases. We provide a link to this environment for all queries, adapting the database name as required. Note that play results may vary from the those presented here due to differences in time of data collection.
This dataset is available in [play.clickhouse.com](https://sql.clickhouse.com?query_id=DCQPNPAIMAQXRLHYURLKVJ) in the `git_clickhouse` databases. We provide a link to this environment for all queries, adapting the database name as required. Note that play results may vary from the those presented here due to differences in time of data collection.
## History of a single file
The simplest of queries. Here we look at all commit messages for the `StorageReplicatedMergeTree.cpp`. Since these are likely more interesting, we sort by the most recent messages first.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICB0aW1lLAogICAgc3Vic3RyaW5nKGNvbW1pdF9oYXNoLCAxLCAxMSkgQVMgY29tbWl0LAogICAgY2hhbmdlX3R5cGUsCiAgICBhdXRob3IsCiAgICBwYXRoLAogICAgb2xkX3BhdGgsCiAgICBsaW5lc19hZGRlZCwKICAgIGxpbmVzX2RlbGV0ZWQsCiAgICBjb21taXRfbWVzc2FnZQpGUk9NIGdpdF9jbGlja2hvdXNlLmZpbGVfY2hhbmdlcwpXSEVSRSBwYXRoID0gJ3NyYy9TdG9yYWdlcy9TdG9yYWdlUmVwbGljYXRlZE1lcmdlVHJlZS5jcHAnCk9SREVSIEJZIHRpbWUgREVTQwpMSU1JVCAxMA==)
[play](https://sql.clickhouse.com?query_id=COAZRFX2YFULDBXRQTCQ1S)
```sql
SELECT
@ -287,7 +287,7 @@ LIMIT 10
We can also review the line changes, excluding renames i.e. we won't show changes before a rename event when the file existed under a different name:
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICB0aW1lLAogICAgc3Vic3RyaW5nKGNvbW1pdF9oYXNoLCAxLCAxMSkgQVMgY29tbWl0LAogICAgc2lnbiwKICAgIGxpbmVfbnVtYmVyX29sZCwKICAgIGxpbmVfbnVtYmVyX25ldywKICAgIGF1dGhvciwKICAgIGxpbmUKRlJPTSBnaXRfY2xpY2tob3VzZS5saW5lX2NoYW5nZXMKV0hFUkUgcGF0aCA9ICdzcmMvU3RvcmFnZXMvU3RvcmFnZVJlcGxpY2F0ZWRNZXJnZVRyZWUuY3BwJwpPUkRFUiBCWSBsaW5lX251bWJlcl9uZXcgQVNDCkxJTUlUIDEw)
[play](https://sql.clickhouse.com?query_id=AKS9SYLARFMZCHGAAQNEBN)
```sql
SELECT
@ -327,7 +327,7 @@ This is important for later analysis when we only want to consider the current f
**Note there appears to have been a broken commit history in relation to files under the `dbms`, `libs`, `tests/testflows/` directories during their renames. We also thus exclude these.**
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=2HNFWPCFWEEY92WTAPMA7W)
```sql
SELECT path
@ -369,7 +369,7 @@ LIMIT 10
Note that this allows for files to be renamed and then re-renamed to their original values. First we aggregate `old_path` for a list of deleted files as a result of renaming. We union this with the last operation for every `path`. Finally, we filter this list to those where the final event is not a `Delete`.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=1OXCKMOH2JVMSHD3NS2WW6)
```sql
SELECT uniq(path)
@ -419,7 +419,7 @@ The difference here is caused by a few factors:
- A rename can occur alongside other modifications to the file. These are listed as separate events in file_changes but with the same time. The `argMax` function has no way of distinguishing these - it picks the first value. The natural ordering of the inserts (the only means of knowing the correct order) is not maintained across the union so modified events can be selected. For example, below the `src/Functions/geometryFromColumn.h` file has several modifications before being renamed to `src/Functions/geometryConverters.h`. Our current solution may pick a Modify event as the latest change causing `src/Functions/geometryFromColumn.h` to be retained.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICAgIGNoYW5nZV90eXBlLAogICAgICBwYXRoLAogICAgICBvbGRfcGF0aCwKICAgICAgdGltZSwKICAgICAgY29tbWl0X2hhc2gKICBGUk9NIGdpdF9jbGlja2hvdXNlLmZpbGVfY2hhbmdlcwogIFdIRVJFIChwYXRoID0gJ3NyYy9GdW5jdGlvbnMvZ2VvbWV0cnlGcm9tQ29sdW1uLmgnKSBPUiAob2xkX3BhdGggPSAnc3JjL0Z1bmN0aW9ucy9nZW9tZXRyeUZyb21Db2x1bW4uaCcpCg==)
[play](https://sql.clickhouse.com?query_id=SCXWMR9GBMJ9UNZYQXQBFA)
```sql
SELECT
@ -454,7 +454,7 @@ These differences shouldn't meaningfully impact our analysis. **We welcome impro
Limiting to current files, we consider the number of modifications to be the sum of deletes and additions.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=MHXPSBNPTDMJYR3OYSXVR7)
```sql
WITH current_files AS
@ -507,7 +507,7 @@ LIMIT 10
## What day of the week do commits usually occur?
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBkYXlfb2Zfd2VlaywKICAgIGNvdW50KCkgQVMgYwpGUk9NIGdpdF9jbGlja2hvdXNlLmNvbW1pdHMKR1JPVVAgQlkgZGF5T2ZXZWVrKHRpbWUpIEFTIGRheV9vZl93ZWVrCg==)
[play](https://sql.clickhouse.com?query_id=GED2STFSYJDRAA59H8RLIV)
```sql
SELECT
@ -534,7 +534,7 @@ This makes sense with some productivity drop-off on Fridays. Great to see people
This would produce a large query result that is unrealistic to show or visualize if unfiltered. We, therefore, allow a file or subdirectory to be filtered in the following example. Here we group by week using the `toStartOfWeek` function - adapt as required.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICB3ZWVrLAogICAgc3VtKGxpbmVzX2FkZGVkKSBBUyBsaW5lc19hZGRlZCwKICAgIHN1bShsaW5lc19kZWxldGVkKSBBUyBsaW5lc19kZWxldGVkLAogICAgdW5pcShjb21taXRfaGFzaCkgQVMgbnVtX2NvbW1pdHMsCiAgICB1bmlxKGF1dGhvcikgQVMgYXV0aG9ycwpGUk9NIGdpdF9jbGlja2hvdXNlLmZpbGVfY2hhbmdlcwpXSEVSRSBwYXRoIExJS0UgJ3NyYy9TdG9yYWdlcyUnCkdST1VQIEJZIHRvU3RhcnRPZldlZWsodGltZSkgQVMgd2VlawpPUkRFUiBCWSB3ZWVrIEFTQwpMSU1JVCAxMAo=)
[play](https://sql.clickhouse.com?query_id=REZRXDVU7CAWT5WKNJSTNY)
```sql
SELECT
@ -578,7 +578,7 @@ This data visualizes well. Below we use Superset.
Limit to current files only.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=CYQFNQNK9TAMPU2OZ8KG5Y)
```sql
WITH current_files AS
@ -633,7 +633,7 @@ LIMIT 10
Limited to current files only.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=VWPBPGRZVGTHOCQYWNQZNT)
```sql
WITH current_files AS
@ -690,7 +690,7 @@ LIMIT 10
Limited to current files only.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=VWPBPGRZVGTHOCQYWNQZNT)
```sql
WITH current_files AS
@ -750,7 +750,7 @@ Our core data structure, the Merge Tree, is obviously under constant evolution w
Do we write more docs at certain times of the month e.g., around release dates? We can use the `countIf` function to compute a simple ratio, visualizing the result using the `bar` function.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=BA4RZUXUHNQBH9YK7F2T9J)
```sql
SELECT
@ -811,7 +811,7 @@ Maybe a little more near the end of the month, but overall we keep a good even d
We consider diversity here to be the number of unique files an author has contributed to.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBhdXRob3IsCiAgICB1bmlxKHBhdGgpIEFTIG51bV9maWxlcwpGUk9NIGdpdF9jbGlja2hvdXNlLmZpbGVfY2hhbmdlcwpXSEVSRSAoY2hhbmdlX3R5cGUgSU4gKCdBZGQnLCAnTW9kaWZ5JykpIEFORCAoZmlsZV9leHRlbnNpb24gSU4gKCdoJywgJ2NwcCcsICdzcWwnKSkKR1JPVVAgQlkgYXV0aG9yCk9SREVSIEJZIG51bV9maWxlcyBERVNDCkxJTUlUIDEw)
[play](https://sql.clickhouse.com?query_id=MT8WBABUKYBYSBA78W5TML)
```sql
SELECT
@ -841,7 +841,7 @@ LIMIT 10
Let's see who has the most diverse commits in their recent work. Rather than limit by date, we'll restrict to an author's last N commits (in this case, we've used 3 but feel free to modify):
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=4Q3D67FWRIVWTY8EIDDE5U)
```sql
SELECT
@ -888,7 +888,7 @@ LIMIT 10
Here we select our founder [Alexey Milovidov](https://github.com/alexey-milovidov) and limit our analysis to current files.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=OKGZBACRHVGCRAGCZAJKMF)
```sql
WITH current_files AS
@ -941,7 +941,7 @@ LIMIT 10
This makes sense because Alexey has been responsible for maintaining the Change log. But what if we use the base name of the file to identify his popular files - this allows for renames and should focus on code contributions.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBiYXNlLAogICAgY291bnQoKSBBUyBjCkZST00gZ2l0X2NsaWNraG91c2UuZmlsZV9jaGFuZ2VzCldIRVJFIChhdXRob3IgPSAnQWxleGV5IE1pbG92aWRvdicpIEFORCAoZmlsZV9leHRlbnNpb24gSU4gKCdoJywgJ2NwcCcsICdzcWwnKSkKR1JPVVAgQlkgYmFzZW5hbWUocGF0aCkgQVMgYmFzZQpPUkRFUiBCWSBjIERFU0MKTElNSVQgMTA=)
[play](https://sql.clickhouse.com?query_id=P9PBDZGOSVTKXEXU73ZNAJ)
```sql
SELECT
@ -976,7 +976,7 @@ For this, we first need to identify the largest files. Estimating this via a ful
To estimate, assuming we restrict to current files, we sum line additions and subtract deletions. We can then compute a ratio of length to the number of authors.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=PVSDOHZYUMRDDUZFEYJC7J)
```sql
WITH current_files AS
@ -1031,7 +1031,7 @@ LIMIT 10
Text dictionaries aren't maybe realistic, so lets restrict to code only via a file extension filter!
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=BZHGWUIZMPZZUHS5XRBK2M)
```sql
WITH current_files AS
@ -1085,7 +1085,7 @@ LIMIT 10
There is some recency bias in this - newer files have fewer opportunities for commits. What about if we restrict to files at least 1 yr old?
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=RMHHZEDHFUCBGRQVQA2732)
```sql
WITH current_files AS
@ -1144,7 +1144,7 @@ LIMIT 10
We interpret this as the number of lines added and removed by the day of the week. In this case, we focus on the [Functions directory](https://github.com/ClickHouse/ClickHouse/tree/master/src/Functions)
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBkYXlPZldlZWssCiAgICB1bmlxKGNvbW1pdF9oYXNoKSBBUyBjb21taXRzLAogICAgc3VtKGxpbmVzX2FkZGVkKSBBUyBsaW5lc19hZGRlZCwKICAgIHN1bShsaW5lc19kZWxldGVkKSBBUyBsaW5lc19kZWxldGVkCkZST00gZ2l0X2NsaWNraG91c2UuZmlsZV9jaGFuZ2VzCldIRVJFIHBhdGggTElLRSAnc3JjL0Z1bmN0aW9ucyUnCkdST1VQIEJZIHRvRGF5T2ZXZWVrKHRpbWUpIEFTIGRheU9mV2Vlaw==)
[play](https://sql.clickhouse.com?query_id=PF3KEMYG5CVLJGCFYQEGB1)
```sql
SELECT
@ -1171,7 +1171,7 @@ GROUP BY toDayOfWeek(time) AS dayOfWeek
And by time of day,
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBob3VyT2ZEYXksCiAgICB1bmlxKGNvbW1pdF9oYXNoKSBBUyBjb21taXRzLAogICAgc3VtKGxpbmVzX2FkZGVkKSBBUyBsaW5lc19hZGRlZCwKICAgIHN1bShsaW5lc19kZWxldGVkKSBBUyBsaW5lc19kZWxldGVkCkZST00gZ2l0X2NsaWNraG91c2UuZmlsZV9jaGFuZ2VzCldIRVJFIHBhdGggTElLRSAnc3JjL0Z1bmN0aW9ucyUnCkdST1VQIEJZIHRvSG91cih0aW1lKSBBUyBob3VyT2ZEYXk=)
[play](https://sql.clickhouse.com?query_id=Q4VDVKEGHHRBCUJHNCVTF1)
```sql
SELECT
@ -1215,7 +1215,7 @@ GROUP BY toHour(time) AS hourOfDay
This distribution makes sense given most of our development team is in Amsterdam. The `bar` functions helps us visualize these distributions:
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=9AZ8CENV8N91YGW7T6IB68)
```sql
SELECT
@ -1269,7 +1269,7 @@ FROM
The `sign = -1` indicates a code deletion. We exclude punctuation and the insertion of empty lines.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBwcmV2X2F1dGhvciB8fCAnKGEpJyBhcyBhZGRfYXV0aG9yLAogICAgYXV0aG9yICB8fCAnKGQpJyBhcyBkZWxldGVfYXV0aG9yLAogICAgY291bnQoKSBBUyBjCkZST00gZ2l0X2NsaWNraG91c2UubGluZV9jaGFuZ2VzCldIRVJFIChzaWduID0gLTEpIEFORCAoZmlsZV9leHRlbnNpb24gSU4gKCdoJywgJ2NwcCcpKSBBTkQgKGxpbmVfdHlwZSBOT1QgSU4gKCdQdW5jdCcsICdFbXB0eScpKSBBTkQgKGF1dGhvciAhPSBwcmV2X2F1dGhvcikgQU5EIChwcmV2X2F1dGhvciAhPSAnJykKR1JPVVAgQlkKICAgIHByZXZfYXV0aG9yLAogICAgYXV0aG9yCk9SREVSIEJZIGMgREVTQwpMSU1JVCAxIEJZIHByZXZfYXV0aG9yCkxJTUlUIDEwMA==)
[play](https://sql.clickhouse.com?query_id=448O8GWAHY3EM6ZZ7AGLAM)
```sql
SELECT
@ -1325,7 +1325,7 @@ Alexey clearly likes removing other peoples code. Lets exclude him for a more ba
If we consider by just number of commits:
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBkYXlfb2Zfd2VlaywKICAgIGF1dGhvciwKICAgIGNvdW50KCkgQVMgYwpGUk9NIGdpdF9jbGlja2hvdXNlLmNvbW1pdHMKR1JPVVAgQlkKICAgIGRheU9mV2Vlayh0aW1lKSBBUyBkYXlfb2Zfd2VlaywKICAgIGF1dGhvcgpPUkRFUiBCWQogICAgZGF5X29mX3dlZWsgQVNDLAogICAgYyBERVNDCkxJTUlUIDEgQlkgZGF5X29mX3dlZWs=)
[play](https://sql.clickhouse.com?query_id=WXPKFJCAHOKYKEVTWNFVCY)
```sql
SELECT
@ -1356,7 +1356,7 @@ LIMIT 1 BY day_of_week
OK, some possible advantages here to the longest contributor - our founder Alexey. Lets limit our analysis to the last year.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBkYXlfb2Zfd2VlaywKICAgIGF1dGhvciwKICAgIGNvdW50KCkgQVMgYwpGUk9NIGdpdF9jbGlja2hvdXNlLmNvbW1pdHMKV0hFUkUgdGltZSA+IChub3coKSAtIHRvSW50ZXJ2YWxZZWFyKDEpKQpHUk9VUCBCWQogICAgZGF5T2ZXZWVrKHRpbWUpIEFTIGRheV9vZl93ZWVrLAogICAgYXV0aG9yCk9SREVSIEJZCiAgICBkYXlfb2Zfd2VlayBBU0MsCiAgICBjIERFU0MKTElNSVQgMSBCWSBkYXlfb2Zfd2Vlaw==)
[play](https://sql.clickhouse.com?query_id=8YRJGHFTNJAWJ96XCJKKEH)
```sql
SELECT
@ -1390,7 +1390,7 @@ This is still a little simple and doesn't reflect people's work.
A better metric might be who is the top contributor each day as a fraction of the total work performed in the last year. Note that we treat the deletion and adding code equally.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=VQF4KMRDSUEXGS1JFVDJHV)
```sql
SELECT
@ -1440,7 +1440,7 @@ INNER JOIN
We limit the analysis to the current files. For brevity, we restrict the results to a depth of 2 with 5 files per root folder. Adjust as required.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=6YWAUQYPZINZDJGBEZBNWG)
```sql
WITH current_files AS
@ -1523,7 +1523,7 @@ LIMIT 5 BY root
For this question, we need the number of lines written by an author divided by the total number of lines they have had removed by another contributor.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=T4DTWTB36WFSEYAZLMGRNF)
```sql
SELECT
@ -1627,7 +1627,7 @@ This doesn't capture the notion of a "re-write" however, where a large portion o
The query is limited to the current files only. We list all file changes by grouping by `path` and `commit_hash`, returning the number of lines added and removed. Using a window function, we estimate the file's total size at any moment in time by performing a cumulative sum and estimating the impact of any change on file size as `lines added - lines removed`. Using this statistic, we can calculate the percentage of the file that has been added or removed for each change. Finally, we count the number of file changes that constitute a rewrite per file i.e. `(percent_add >= 0.5) AND (percent_delete >= 0.5) AND current_size > 50`. Note we require files to be more than 50 lines to avoid early contributions to a file being counted as a rewrite. This also avoids a bias to very small files, which may be more likely to be rewritten.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=5PL1QLNSH6QQTR8H9HINNP)
```sql
WITH
@ -1719,7 +1719,7 @@ We query for lines added, joining this with the lines removed - filtering to cas
Finally, we aggregate across this dataset to compute the average number of days lines stay in the repository by the day of the week.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=GVF23LEZTNZI22BT8LZBBE)
```sql
SELECT
@ -1778,7 +1778,7 @@ GROUP BY dayOfWeek(added_day) AS day_of_week_added
This query uses the same principle as [What weekday does the code have the highest chance to stay in the repository](#what-weekday-does-the-code-have-the-highest-chance-to-stay-in-the-repository) - by aiming to uniquely identify a line of code using the path and line contents.
This allows us to identify the time between when a line was added and removed. We filter to current files and code only, however, and average the time for each file across lines.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=3CYYT7HEHWRFHVCM9JCKSU)
```sql
WITH
@ -1869,7 +1869,7 @@ There are a few ways we can address this question. Focusing on the code to test
Note we limit to users with more than 20 changes to focus on regular committers and avoid a bias to one-off contributions.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBhdXRob3IsCiAgICBjb3VudElmKChmaWxlX2V4dGVuc2lvbiBJTiAoJ2gnLCAnY3BwJywgJ3NxbCcsICdzaCcsICdweScsICdleHBlY3QnKSkgQU5EIChwYXRoIExJS0UgJyV0ZXN0cyUnKSkgQVMgdGVzdCwKICAgIGNvdW50SWYoKGZpbGVfZXh0ZW5zaW9uIElOICgnaCcsICdjcHAnLCAnc3FsJykpIEFORCAoTk9UIChwYXRoIExJS0UgJyV0ZXN0cyUnKSkpIEFTIGNvZGUsCiAgICBjb2RlIC8gKGNvZGUgKyB0ZXN0KSBBUyByYXRpb19jb2RlCkZST00gZ2l0X2NsaWNraG91c2UuZmlsZV9jaGFuZ2VzCkdST1VQIEJZIGF1dGhvcgpIQVZJTkcgY29kZSA+IDIwCk9SREVSIEJZIGNvZGUgREVTQwpMSU1JVCAyMA==)
[play](https://sql.clickhouse.com?query_id=JGKZSEQDPDTDKZXD3ZCGLE)
```sql
SELECT
@ -1911,7 +1911,7 @@ LIMIT 20
We can plot this distribution as a histogram.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=S5AJIIRGSUAY1JXEVHQDAK)
```sql
WITH (
@ -1954,7 +1954,7 @@ Most contributors write more code than tests, as you'd expect.
What about who adds the most comments when contributing code?
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=EXPHDIURBTOXXOK1TGNNYD)
```sql
SELECT
@ -2038,7 +2038,7 @@ To compute this, we first work out each author's comments ratio over time - simi
After calculating the average by-week offset across all authors, we sample these results by selecting every 10th week.
[play](https://play.clickhouse.com/play?user=play#V0lUSCBhdXRob3JfcmF0aW9zX2J5X29mZnNldCBBUwogICAgKAogICAgICAgIFNFTEVDVAogICAgICAgICAgICBhdXRob3IsCiAgICAgICAgICAgIGRhdGVEaWZmKCd3ZWVrJywgc3RhcnRfZGF0ZXMuc3RhcnRfZGF0ZSwgY29udHJpYnV0aW9ucy53ZWVrKSBBUyB3ZWVrX29mZnNldCwKICAgICAgICAgICAgcmF0aW9fY29kZQogICAgICAgIEZST00KICAgICAgICAoCiAgICAgICAgICAgIFNFTEVDVAogICAgICAgICAgICAgICAgYXV0aG9yLAogICAgICAgICAgICAgICAgdG9TdGFydE9mV2VlayhtaW4odGltZSkpIEFTIHN0YXJ0X2RhdGUKICAgICAgICAgICAgRlJPTSBnaXRfY2xpY2tob3VzZS5saW5lX2NoYW5nZXMKICAgICAgICAgICAgV0hFUkUgZmlsZV9leHRlbnNpb24gSU4gKCdoJywgJ2NwcCcsICdzcWwnKQogICAgICAgICAgICBHUk9VUCBCWSBhdXRob3IgQVMgc3RhcnRfZGF0ZXMKICAgICAgICApIEFTIHN0YXJ0X2RhdGVzCiAgICAgICAgSU5ORVIgSk9JTgogICAgICAgICgKICAgICAgICAgICAgU0VMRUNUCiAgICAgICAgICAgICAgICBhdXRob3IsCiAgICAgICAgICAgICAgICBjb3VudElmKGxpbmVfdHlwZSA9ICdDb2RlJykgQVMgY29kZSwKICAgICAgICAgICAgICAgIGNvdW50SWYoKGxpbmVfdHlwZSA9ICdDb21tZW50JykgT1IgKGxpbmVfdHlwZSA9ICdQdW5jdCcpKSBBUyBjb21tZW50cywKICAgICAgICAgICAgICAgIGNvbW1lbnRzIC8gKGNvbW1lbnRzICsgY29kZSkgQVMgcmF0aW9fY29kZSwKICAgICAgICAgICAgICAgIHRvU3RhcnRPZldlZWsodGltZSkgQVMgd2VlawogICAgICAgICAgICBGUk9NIGdpdF9jbGlja2hvdXNlLmxpbmVfY2hhbmdlcwogICAgICAgICAgICBXSEVSRSAoZmlsZV9leHRlbnNpb24gSU4gKCdoJywgJ2NwcCcsICdzcWwnKSkgQU5EIChzaWduID0gMSkKICAgICAgICAgICAgR1JPVVAgQlkKICAgICAgICAgICAgICAgIHRpbWUsCiAgICAgICAgICAgICAgICBhdXRob3IKICAgICAgICAgICAgSEFWSU5HIGNvZGUgPiAyMAogICAgICAgICAgICBPUkRFUiBCWQogICAgICAgICAgICAgICAgYXV0aG9yIEFTQywKICAgICAgICAgICAgICAgIHRpbWUgQVNDCiAgICAgICAgKSBBUyBjb250cmlidXRpb25zIFVTSU5HIChhdXRob3IpCiAgICApClNFTEVDVAogICAgd2Vla19vZmZzZXQsCiAgICBhdmcocmF0aW9fY29kZSkgQVMgYXZnX2NvZGVfcmF0aW8KRlJPTSBhdXRob3JfcmF0aW9zX2J5X29mZnNldApHUk9VUCBCWSB3ZWVrX29mZnNldApIQVZJTkcgKHdlZWtfb2Zmc2V0ICUgMTApID0gMApPUkRFUiBCWSB3ZWVrX29mZnNldCBBU0MKTElNSVQgMjAK)
[play](https://sql.clickhouse.com?query_id=SBHEWR8XC4PRHY13HPPKCN)
```sql
WITH author_ratios_by_offset AS
@ -2116,7 +2116,7 @@ Encouragingly, our comment % is pretty constant and doesn't degrade the longer a
We can use the same principle as [List files that were rewritten most number of time or by most of authors](#list-files-that-were-rewritten-most-number-of-time-or-by-most-of-authors) to identify rewrites but consider all files. A window function is used to compute the time between rewrites for each file. From this, we can calculate an average and median across all files.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=WSHUEPJP9TNJUH7QITWWOR)
```sql
WITH
@ -2176,7 +2176,7 @@ FROM rewrites
Similar to [What is the average time before code will be rewritten and the median (half-life of code decay)?](#what-is-the-average-time-before-code-will-be-rewritten-and-the-median-half-life-of-code-decay) and [List files that were rewritten most number of time or by most of authors](#list-files-that-were-rewritten-most-number-of-time-or-by-most-of-authors), except we aggregate by day of week. Adjust as required e.g. month of year.
[play](https://play.clickhouse.com/play?user=play#V0lUSAogICAgY2hhbmdlcyBBUwogICAgKAogICAgICAgIFNFTEVDVAogICAgICAgICAgICBwYXRoLAogICAgICAgICAgICBjb21taXRfaGFzaCwKICAgICAgICAgICAgbWF4X3RpbWUsCiAgICAgICAgICAgIHR5cGUsCiAgICAgICAgICAgIG51bV9hZGRlZCwKICAgICAgICAgICAgbnVtX2RlbGV0ZWQsCiAgICAgICAgICAgIHN1bShudW1fYWRkZWQgLSBudW1fZGVsZXRlZCkgT1ZFUiAoUEFSVElUSU9OIEJZIHBhdGggT1JERVIgQlkgbWF4X3RpbWUgQVNDKSBBUyBjdXJyZW50X3NpemUsCiAgICAgICAgICAgIGlmKGN1cnJlbnRfc2l6ZSA+IDAsIG51bV9hZGRlZCAvIGN1cnJlbnRfc2l6ZSwgMCkgQVMgcGVyY2VudF9hZGQsCiAgICAgICAgICAgIGlmKGN1cnJlbnRfc2l6ZSA+IDAsIG51bV9kZWxldGVkIC8gY3VycmVudF9zaXplLCAwKSBBUyBwZXJjZW50X2RlbGV0ZQogICAgICAgIEZST00KICAgICAgICAoCiAgICAgICAgICAgIFNFTEVDVAogICAgICAgICAgICAgICAgcGF0aCwKICAgICAgICAgICAgICAgIG1heCh0aW1lKSBBUyBtYXhfdGltZSwKICAgICAgICAgICAgICAgIGNvbW1pdF9oYXNoLAogICAgICAgICAgICAgICAgYW55KGZpbGVfbGluZXNfYWRkZWQpIEFTIG51bV9hZGRlZCwKICAgICAgICAgICAgICAgIGFueShmaWxlX2xpbmVzX2RlbGV0ZWQpIEFTIG51bV9kZWxldGVkLAogICAgICAgICAgICAgICAgYW55KGZpbGVfY2hhbmdlX3R5cGUpIEFTIHR5cGUKICAgICAgICAgICAgRlJPTSBnaXRfY2xpY2tob3VzZS5saW5lX2NoYW5nZXMKICAgICAgICAgICAgV0hFUkUgKGZpbGVfY2hhbmdlX3R5cGUgSU4gKCdBZGQnLCAnTW9kaWZ5JykpIEFORCAoZmlsZV9leHRlbnNpb24gSU4gKCdoJywgJ2NwcCcsICdzcWwnKSkKICAgICAgICAgICAgR1JPVVAgQlkKICAgICAgICAgICAgICAgIHBhdGgsCiAgICAgICAgICAgICAgICBjb21taXRfaGFzaAogICAgICAgICAgICBPUkRFUiBCWQogICAgICAgICAgICAgICAgcGF0aCBBU0MsCiAgICAgICAgICAgICAgICBtYXhfdGltZSBBU0MKICAgICAgICApCiAgICApLAogICAgcmV3cml0ZXMgQVMKICAgICgKICAgICAgICBTRUxFQ1QgYW55KG1heF90aW1lKSBPVkVSIChQQVJUSVRJT04gQlkgcGF0aCBPUkRFUiBCWSBtYXhfdGltZSBBU0MgUk9XUyBCRVRXRUVOIDEgUFJFQ0VESU5HIEFORCBDVVJSRU5UIFJPVykgQVMgcHJldmlvdXNfcmV3cml0ZQogICAgICAgIEZST00gY2hhbmdlcwogICAgICAgIFdIRVJFICh0eXBlID0gJ01vZGlmeScpIEFORCAocGVyY2VudF9hZGQgPj0gMC41KSBBTkQgKHBlcmNlbnRfZGVsZXRlID49IDAuNSkgQU5EIChjdXJyZW50X3NpemUgPiA1MCkKICAgICkKU0VMRUNUCiAgICBkYXlPZldlZWsocHJldmlvdXNfcmV3cml0ZSkgQVMgZGF5T2ZXZWVrLAogICAgY291bnQoKSBBUyBudW1fcmVfd3JpdGVzCkZST00gcmV3cml0ZXMKR1JPVVAgQlkgZGF5T2ZXZWVr)
[play](https://sql.clickhouse.com?query_id=8PQNWEWHAJTGN6FTX59KH2)
```sql
WITH
@ -2240,7 +2240,7 @@ GROUP BY dayOfWeek
We define "sticky" as how long does an author's code stay before its rewritten. Similar to the previous question [What is the average time before code will be rewritten and the median (half-life of code decay)?](#what-is-the-average-time-before-code-will-be-rewritten-and-the-median-half-life-of-code-decay) - using the same metric for rewrites i.e. 50% additions and 50% deletions to the file. We compute the average rewrite time per author and only consider contributors with more than two files.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=BKHLVVWN5SET1VTIFQ8JVK)
```sql
WITH
@ -2319,7 +2319,7 @@ This query first requires us to calculate the days when an author has committed.
Our subsequent array functions compute each author's longest sequence of consecutive ones. First, the `groupArray` function is used to collate all `consecutive_day` values for an author. This array of 1s and 0s, is then split on 0 values into subarrays. Finally, we calculate the longest subarray.
[play](https://play.clickhouse.com/play?user=play#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)
[play](https://sql.clickhouse.com?query_id=S3E64UYCAMDAYJRSXINVFR)
```sql
WITH commit_days AS
@ -2372,7 +2372,7 @@ LIMIT 10
Files can be renamed. When this occurs, we get a rename event, where the `path` column is set to the new path of the file and the `old_path` represents the previous location e.g.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICB0aW1lLAogICAgcGF0aCwKICAgIG9sZF9wYXRoLAogICAgY29tbWl0X2hhc2gsCiAgICBjb21taXRfbWVzc2FnZQpGUk9NIGdpdF9jbGlja2hvdXNlLmZpbGVfY2hhbmdlcwpXSEVSRSAocGF0aCA9ICdzcmMvU3RvcmFnZXMvU3RvcmFnZVJlcGxpY2F0ZWRNZXJnZVRyZWUuY3BwJykgQU5EIChjaGFuZ2VfdHlwZSA9ICdSZW5hbWUnKQ==)
[play](https://sql.clickhouse.com?query_id=AKTW3Z8JZAPQ4H9BH2ZFRX)
```sql
SELECT
@ -2410,8 +2410,6 @@ By calling `file_path_history('src/Storages/StorageReplicatedMergeTree.cpp')` we
For example,
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUIGZpbGVfcGF0aF9oaXN0b3J5KCdzcmMvU3RvcmFnZXMvU3RvcmFnZVJlcGxpY2F0ZWRNZXJnZVRyZWUuY3BwJykgQVMgcGF0aHMK)
```sql
SELECT file_path_history('src/Storages/StorageReplicatedMergeTree.cpp') AS paths
@ -2424,8 +2422,6 @@ SELECT file_path_history('src/Storages/StorageReplicatedMergeTree.cpp') AS paths
We can use this capability to now assemble the commits for the entire history of a file. In this example, we show one commit for each of the `path` values.
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICB0aW1lLAogICAgc3Vic3RyaW5nKGNvbW1pdF9oYXNoLCAxLCAxMSkgQVMgY29tbWl0LAogICAgY2hhbmdlX3R5cGUsCiAgICBhdXRob3IsCiAgICBwYXRoLAogICAgY29tbWl0X21lc3NhZ2UKRlJPTSBnaXRfY2xpY2tob3VzZS5maWxlX2NoYW5nZXMKV0hFUkUgcGF0aCBJTiBmaWxlX3BhdGhfaGlzdG9yeSgnc3JjL1N0b3JhZ2VzL1N0b3JhZ2VSZXBsaWNhdGVkTWVyZ2VUcmVlLmNwcCcpCk9SREVSIEJZIHRpbWUgREVTQwpMSU1JVCAxIEJZIHBhdGgKRk9STUFUIFByZXR0eUNvbXBhY3RNb25vQmxvY2s=)
```sql
SELECT
time,
@ -2457,8 +2453,6 @@ This is particularly difficult to get an exact result due to the inability to cu
An approximate solution, sufficient for a high-level analysis, may look something like this:
[play](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICBsaW5lX251bWJlcl9uZXcsCiAgICBhcmdNYXgoYXV0aG9yLCB0aW1lKSwKICAgIGFyZ01heChsaW5lLCB0aW1lKQpGUk9NIGdpdF9jbGlja2hvdXNlLmxpbmVfY2hhbmdlcwpXSEVSRSBwYXRoIElOIGZpbGVfcGF0aF9oaXN0b3J5KCdzcmMvU3RvcmFnZXMvU3RvcmFnZVJlcGxpY2F0ZWRNZXJnZVRyZWUuY3BwJykKR1JPVVAgQlkgbGluZV9udW1iZXJfbmV3Ck9SREVSIEJZIGxpbmVfbnVtYmVyX25ldyBBU0MKTElNSVQgMjA=)
```sql
SELECT
line_number_new,

2
docs/en/getting-started/example-datasets/menus.md
View File

@ -354,4 +354,4 @@ At least they have caviar with vodka. Very nice.
## Online Playground {#playground}
The data is uploaded to ClickHouse Playground, [example](https://play.clickhouse.com/play?user=play#U0VMRUNUCiAgICByb3VuZCh0b1VJbnQzMk9yWmVybyhleHRyYWN0KG1lbnVfZGF0ZSwgJ15cXGR7NH0nKSksIC0xKSBBUyBkLAogICAgY291bnQoKSwKICAgIHJvdW5kKGF2ZyhwcmljZSksIDIpLAogICAgYmFyKGF2ZyhwcmljZSksIDAsIDUwLCAxMDApLAogICAgYW55KGRpc2hfbmFtZSkKRlJPTSBtZW51X2l0ZW1fZGVub3JtCldIRVJFIChtZW51X2N1cnJlbmN5IElOICgnRG9sbGFycycsICcnKSkgQU5EIChkID4gMCkgQU5EIChkIDwgMjAyMikgQU5EIChkaXNoX25hbWUgSUxJS0UgJyVjYXZpYXIlJykKR1JPVVAgQlkgZApPUkRFUiBCWSBkIEFTQw==).
The data is uploaded to ClickHouse Playground, [example](https://sql.clickhouse.com?query_id=KB5KQJJFNBKHE5GBUJCP1B).

2
docs/en/getting-started/example-datasets/ontime.md
View File

@ -386,7 +386,7 @@ ORDER BY c DESC
LIMIT 10;
```
You can also play with the data in Playground, [example](https://play.clickhouse.com/play?user=play#U0VMRUNUIERheU9mV2VlaywgY291bnQoKikgQVMgYwpGUk9NIG9udGltZQpXSEVSRSBZZWFyPj0yMDAwIEFORCBZZWFyPD0yMDA4CkdST1VQIEJZIERheU9mV2VlawpPUkRFUiBCWSBjIERFU0M7Cg==).
You can also play with the data in Playground, [example](https://sql.clickhouse.com?query_id=M4FSVBVMSHY98NKCQP8N4K).
This performance test was created by Vadim Tkachenko. See:

Some files were not shown because too many files have changed in this diff Show More