thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-15 12:14:18 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'ClickHouse:master' into fix_logical_error_substrings

Browse Source
This commit is contained in:
Yarik Briukhovetskyi 2024-10-24 15:42:52 +02:00 committed by GitHub
commit 6640d78879
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
338 changed files with 16081 additions and 4962 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

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;
}
}

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);

105
ci_v2/docker/fasttest/Dockerfile Normal file
View File

@ -0,0 +1,105 @@
# 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-${LLVM_VERSION} \
&& apt-get clean \
&& rm -rf /var/lib/apt/lists/* /var/cache/debconf /tmp/*
# moreutils - provides ts fo FT
# expect, bzip2 - requried by FT
# bsdmainutils - provides hexdump for FT
RUN apt-get update \
&& apt-get install \
clang-${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 \
--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_v2/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

3
ci_v2/docker/style-test/requirements.txt
View File

@ -1,4 +1,5 @@
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
#use praktika from CH repo
#https://clickhouse-builds.s3.amazonaws.com/packages/praktika-0.1-py3-none-any.whl

92
ci_v2/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:
@ -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 = [
@ -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()

329
ci_v2/jobs/fast_test.py Normal file
View File

@ -0,0 +1,329 @@
import threading
from pathlib import Path
from ci_v2.jobs.scripts.functional_tests_results import FTResultsProcessor
from praktika.environment import Environment
from praktika.result import Result
from praktika.settings import Settings
from praktika.utils import MetaClasses, Shell, Utils
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 main():
stop_watch = Utils.Stopwatch()
stages = list(JobStages)
stage = Environment.LOCAL_RUN_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()

82
ci_v2/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+\*|\$\(\()| \)"'
@ -39,12 +41,12 @@ 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
@ -52,26 +54,26 @@ find -L $ROOT_PATH -type l 2>/dev/null | grep -v contrib && echo "^ Broken symli
# Duplicated or incorrect setting declarations
SETTINGS_FILE=$(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}
ALL_DECLARATION_FILES="
$ROOT_PATH/src/Core/Settings.cpp
$ROOT_PATH/src/Storages/MergeTree/MergeTreeSettings.cpp
$ROOT_PATH/src/Core/FormatFactorySettingsDeclaration.h"
# Duplicate extern declarations for settings
awk '{if (seen[$0]++) print $3 " -> " $1 ;}' ${SETTINGS_FILE} | while read line;
cat $ROOT_PATH/src/Core/Settings.cpp $ROOT_PATH/src/Core/FormatFactorySettingsDeclaration.h | grep "M(" | awk '{print substr($2, 0, length($2) - 1) " Settings" substr($1, 3, length($1) - 3) " SettingsDeclaration" }' | sort | uniq > ${SETTINGS_FILE}
cat $ROOT_PATH/src/Storages/MergeTree/MergeTreeSettings.cpp | grep "M(" | awk '{print substr($2, 0, length($2) - 1) " MergeTreeSettings" substr($1, 3, length($1) - 3) " SettingsDeclaration" }' | sort | uniq >> ${SETTINGS_FILE}
# Check that if there are duplicated settings (declared in different objects) they all have the same type (it's simpler to validate style with that assert)
for setting in $(awk '{print $1 " " $2}' ${SETTINGS_FILE} | sed -e 's/MergeTreeSettings//g' -e 's/Settings//g' | sort | uniq | awk '{ print $1 }' | uniq -d);
do
echo "Found duplicated setting declaration in: $line"
echo "# Found multiple definitions of setting ${setting} with different types: "
grep --line-number " ${setting}," ${ALL_DECLARATION_FILES} | awk '{print " > " $0 }'
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
# We append all uses of extern found in implementation files to validate them in a single pass and avoid reading the same files over and over
find $ROOT_PATH/{src,base,programs,utils} -name '*.h' -or -name '*.cpp' | xargs grep -e "^\s*extern const Settings" -e "^\s**extern const MergeTreeSettings" -T | awk '{print substr($5, 0, length($5) -1) " " $4 " " substr($1, 0, length($1) - 1)}' >> ${SETTINGS_FILE}
rm ${SETTINGS_FILE}
# Duplicated or incorrect setting declarations
bash $ROOT_PATH/utils/check-style/check-settings-style
# Unused/Undefined/Duplicates ErrorCodes/ProfileEvents/CurrentMetrics
declare -A EXTERN_TYPES
@ -91,12 +93,14 @@ 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 +112,7 @@ EXTERN_TYPES_EXCLUDES=(
CurrentMetrics::values
CurrentMetrics::Value
CurrentMetrics::keeper_metrics
CurrentMetrics::size
ErrorCodes::ErrorCode
ErrorCodes::getName
@ -130,7 +135,7 @@ 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 +153,7 @@ 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 +166,7 @@ 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 +174,32 @@ 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 +209,7 @@ 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 +236,10 @@ 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 +284,23 @@ 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 +327,15 @@ 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"

284
ci_v2/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")

115
ci_v2/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"
@ -29,142 +30,122 @@ SECRETS = [
DOCKERS = [
# Docker.Config(
# name="clickhouse/binary-builder",
# path="./docker/packager/binary-builder",
# arm64=True,
# amd64=True,
# path="./ci_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/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_v2/docker/test/integration/runner",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
Docker.Config(
@ -175,9 +156,8 @@ DOCKERS = [
),
# Docker.Config(
# name="clickhouse/docs-builder",
# path="./docker/docs/builder",
# arm64=True,
# amd64=True,
# path="./ci_v2/docker/docs/builder",
# platforms=Docker.Platforms.arm_amd,
# depends_on=["clickhouse/test-base"],
# ),
]
@ -249,3 +229,4 @@ DOCKERS = [
class JobNames:
STYLE_CHECK = "Style Check"
FAST_TEST = "Fast test"

14
ci_v2/workflows/pull_request.py
View File

@ -16,12 +16,20 @@ style_check_job = Job.Config(
run_in_docker="clickhouse/style-test",
)
fast_test_job = Job.Config(
name=JobNames.FAST_TEST,
runs_on=[RunnerLabels.BUILDER],
command="python3 ./ci_v2/jobs/fast_test.py",
run_in_docker="clickhouse/fasttest",
)
workflow = Workflow.Config(
name="PR",
event=Workflow.Event.PULL_REQUEST,
base_branches=[BASE_BRANCH],
jobs=[
style_check_job,
fast_test_job,
],
dockers=DOCKERS,
secrets=SECRETS,
@ -36,9 +44,7 @@ WORKFLOWS = [
if __name__ == "__main__":
# example: local job test inside praktika environment
# 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)
Runner().run(workflow, fast_test_job, docker="fasttest", dummy_env=True)

2
contrib/orc vendored

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

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 \

7
docs/en/development/build-osx.md
View File

@ -42,15 +42,14 @@ brew install ccache cmake ninja libtool gettext llvm gcc binutils grep findutils
## Checkout ClickHouse Sources {#checkout-clickhouse-sources}
:::note
The ClickHouse build assumes a case-sensitive file system. Case-insensitive file systems may cause errors during the build process. If necessary, please follow [these instructions](https://brianboyko.medium.com/a-case-sensitive-src-folder-for-mac-programmers-176cc82a3830) to create a new disk image and checkout the code into it.
:::
``` bash
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):

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/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:

4
docs/en/engines/table-engines/mergetree-family/annindexes.md
View File

@ -56,7 +56,7 @@ Parameters:
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`)
- `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: 16)
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)
@ -143,7 +143,7 @@ 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: 64), also known as `ef_search` in the
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>`.

711
docs/en/getting-started/example-datasets/tpch.md
View File

File diff suppressed because it is too large Load Diff

55
docs/en/interfaces/formats.md
View File

@ -2054,35 +2054,40 @@ ClickHouse Avro format supports reading and writing [Avro data files](https://av
The table below shows supported data types and how they match ClickHouse [data types](/docs/en/sql-reference/data-types/index.md) in `INSERT` and `SELECT` queries.
| Avro data type `INSERT` | ClickHouse data type | Avro data type `SELECT` |
|---------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------|-------------------------------|
| `boolean`, `int`, `long`, `float`, `double` | [Int(8\16\32)](/docs/en/sql-reference/data-types/int-uint.md), [UInt(8\16\32)](/docs/en/sql-reference/data-types/int-uint.md) | `int` |
| `boolean`, `int`, `long`, `float`, `double` | [Int64](/docs/en/sql-reference/data-types/int-uint.md), [UInt64](/docs/en/sql-reference/data-types/int-uint.md) | `long` |
| `boolean`, `int`, `long`, `float`, `double` | [Float32](/docs/en/sql-reference/data-types/float.md) | `float` |
| `boolean`, `int`, `long`, `float`, `double` | [Float64](/docs/en/sql-reference/data-types/float.md) | `double` |
| `bytes`, `string`, `fixed`, `enum` | [String](/docs/en/sql-reference/data-types/string.md) | `bytes` or `string` \* |
| `bytes`, `string`, `fixed` | [FixedString(N)](/docs/en/sql-reference/data-types/fixedstring.md) | `fixed(N)` |
| `enum` | [Enum(8\16)](/docs/en/sql-reference/data-types/enum.md) | `enum` |
| `array(T)` | [Array(T)](/docs/en/sql-reference/data-types/array.md) | `array(T)` |
| `map(V, K)` | [Map(V, K)](/docs/en/sql-reference/data-types/map.md) | `map(string, K)` |
| `union(null, T)`, `union(T, null)` | [Nullable(T)](/docs/en/sql-reference/data-types/date.md) | `union(null, T)` |
| `null` | [Nullable(Nothing)](/docs/en/sql-reference/data-types/special-data-types/nothing.md) | `null` |
| `int (date)` \** | [Date](/docs/en/sql-reference/data-types/date.md), [Date32](docs/en/sql-reference/data-types/date32.md) | `int (date)` \** |
| `long (timestamp-millis)` \** | [DateTime64(3)](/docs/en/sql-reference/data-types/datetime.md) | `long (timestamp-millis)` \** |
| `long (timestamp-micros)` \** | [DateTime64(6)](/docs/en/sql-reference/data-types/datetime.md) | `long (timestamp-micros)` \** |
| `bytes (decimal)` \** | [DateTime64(N)](/docs/en/sql-reference/data-types/datetime.md) | `bytes (decimal)` \** |
| `int` | [IPv4](/docs/en/sql-reference/data-types/ipv4.md) | `int` |
| `fixed(16)` | [IPv6](/docs/en/sql-reference/data-types/ipv6.md) | `fixed(16)` |
| `bytes (decimal)` \** | [Decimal(P, S)](/docs/en/sql-reference/data-types/decimal.md) | `bytes (decimal)` \** |
| `string (uuid)` \** | [UUID](/docs/en/sql-reference/data-types/uuid.md) | `string (uuid)` \** |
| `fixed(16)` | [Int128/UInt128](/docs/en/sql-reference/data-types/int-uint.md) | `fixed(16)` |
| `fixed(32)` | [Int256/UInt256](/docs/en/sql-reference/data-types/int-uint.md) | `fixed(32)` |
| `record` | [Tuple](/docs/en/sql-reference/data-types/tuple.md) | `record` |
| Avro data type `INSERT` | ClickHouse data type | Avro data type `SELECT` |
|---------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------|---------------------------------|
| `boolean`, `int`, `long`, `float`, `double` | [Int(8\16\32)](/docs/en/sql-reference/data-types/int-uint.md), [UInt(8\16\32)](/docs/en/sql-reference/data-types/int-uint.md) | `int` |
| `boolean`, `int`, `long`, `float`, `double` | [Int64](/docs/en/sql-reference/data-types/int-uint.md), [UInt64](/docs/en/sql-reference/data-types/int-uint.md) | `long` |
| `boolean`, `int`, `long`, `float`, `double` | [Float32](/docs/en/sql-reference/data-types/float.md) | `float` |
| `boolean`, `int`, `long`, `float`, `double` | [Float64](/docs/en/sql-reference/data-types/float.md) | `double` |
| `bytes`, `string`, `fixed`, `enum` | [String](/docs/en/sql-reference/data-types/string.md) | `bytes` or `string` \* |
| `bytes`, `string`, `fixed` | [FixedString(N)](/docs/en/sql-reference/data-types/fixedstring.md) | `fixed(N)` |
| `enum` | [Enum(8\16)](/docs/en/sql-reference/data-types/enum.md) | `enum` |
| `array(T)` | [Array(T)](/docs/en/sql-reference/data-types/array.md) | `array(T)` |
| `map(V, K)` | [Map(V, K)](/docs/en/sql-reference/data-types/map.md) | `map(string, K)` |
| `union(null, T)`, `union(T, null)` | [Nullable(T)](/docs/en/sql-reference/data-types/date.md) | `union(null, T)` |
| `union(T1, T2, …)` \** | [Variant(T1, T2, …)](/docs/en/sql-reference/data-types/variant.md) | `union(T1, T2, …)` \** |
| `null` | [Nullable(Nothing)](/docs/en/sql-reference/data-types/special-data-types/nothing.md) | `null` |
| `int (date)` \**\* | [Date](/docs/en/sql-reference/data-types/date.md), [Date32](docs/en/sql-reference/data-types/date32.md) | `int (date)` \**\* |
| `long (timestamp-millis)` \**\* | [DateTime64(3)](/docs/en/sql-reference/data-types/datetime.md) | `long (timestamp-millis)` \**\* |
| `long (timestamp-micros)` \**\* | [DateTime64(6)](/docs/en/sql-reference/data-types/datetime.md) | `long (timestamp-micros)` \**\* |
| `bytes (decimal)` \**\* | [DateTime64(N)](/docs/en/sql-reference/data-types/datetime.md) | `bytes (decimal)` \**\* |
| `int` | [IPv4](/docs/en/sql-reference/data-types/ipv4.md) | `int` |
| `fixed(16)` | [IPv6](/docs/en/sql-reference/data-types/ipv6.md) | `fixed(16)` |
| `bytes (decimal)` \**\* | [Decimal(P, S)](/docs/en/sql-reference/data-types/decimal.md) | `bytes (decimal)` \**\* |
| `string (uuid)` \**\* | [UUID](/docs/en/sql-reference/data-types/uuid.md) | `string (uuid)` \**\* |
| `fixed(16)` | [Int128/UInt128](/docs/en/sql-reference/data-types/int-uint.md) | `fixed(16)` |
| `fixed(32)` | [Int256/UInt256](/docs/en/sql-reference/data-types/int-uint.md) | `fixed(32)` |
| `record` | [Tuple](/docs/en/sql-reference/data-types/tuple.md) | `record` |
\* `bytes` is default, controlled by [output_format_avro_string_column_pattern](/docs/en/operations/settings/settings-formats.md/#output_format_avro_string_column_pattern)
\** [Avro logical types](https://avro.apache.org/docs/current/spec.html#Logical+Types)
\** [Variant type](/docs/en/sql-reference/data-types/variant) implicitly accepts `null` as a field value, so for example the Avro `union(T1, T2, null)` will be converted to `Variant(T1, T2)`.
As a result, when producing Avro from ClickHouse, we have to always include the `null` type to the Avro `union` type set as we don't know if any value is actually `null` during the schema inference.
\**\* [Avro logical types](https://avro.apache.org/docs/current/spec.html#Logical+Types)
Unsupported Avro logical data types: `time-millis`, `time-micros`, `duration`

2
docs/en/operations/opentelemetry.md
View File

@ -33,7 +33,7 @@ The tags or attributes are saved as two parallel arrays, containing the keys and
## Log-query-settings
ClickHouse allows you to log changes to query settings during query execution. When enabled, any modifications made to query settings will be recorded in the OpenTelemetry span log. This feature is particularly useful in production environments for tracking configuration changes that may affect query performance.
Setting [log_query_settings](settings/settings.md) allows log changes to query settings during query execution. When enabled, any modifications made to query settings will be recorded in the OpenTelemetry span log. This feature is particularly useful in production environments for tracking configuration changes that may affect query performance.
## Integration with monitoring systems

24
docs/en/operations/server-configuration-parameters/settings.md
View File

@ -1488,6 +1488,8 @@ Keys:
- `formatting` – Log format for console output. Currently, only `json` is supported).
- `use_syslog` - Also forward log output to syslog.
- `syslog_level` - Log level for logging to syslog.
- `message_regexp` - Only log messages that match this regular expression. Defaults to `""`, indicating no filtering.
- `message_regexp_negative` - Only log messages that don't match this regular expression. Defaults to `""`, indicating no filtering.
**Log format specifiers**
@ -1576,6 +1578,28 @@ The log level of individual log names can be overridden. For example, to mute al
</logger>
```
**Regular Expression Filtering**
The messages logged can be filtered using regular expressions using `message_regexp` and `message_regexp_negative`. This can be done on a per-level basis or globally. If both a global and logger-specific pattern is specified, the global pattern is overridden (ignored) and only the logger-specific pattern applies. The positive and negative patterns are considered independently for this situation. Note: Using this feature may cause a slight slowdown in performance.
```xml
<logger>
<level>trace</level>
<!-- Global: Don't log Trace messages -->
<message_regexp_negative>.*Trace.*</message_regexp_negative>
<message_regexps>
<logger>
<!-- For the executeQuery logger, only log if message has "Read", but not "from" -->
<name>executeQuery</name>
<message_regexp>.*Read.*</message_regexp>
<message_regexp_negative>.*from.*</message_regexp_negative>
</logger>
</message_regexps>
</logger>
```
### syslog
To write log messages additionally to syslog:

2
docs/en/operations/settings/merge-tree-settings.md
View File

@ -1079,6 +1079,8 @@ Possible values:
Default value: 0 bytes.
Note that if both `min_free_disk_bytes_to_perform_insert` and `min_free_disk_ratio_to_perform_insert` are specified, ClickHouse will count on the value that will allow to perform inserts on a bigger amount of free memory.
## min_free_disk_ratio_to_perform_insert
The minimum free to total disk space ratio to perform an `INSERT`. Must be a floating point value between 0 and 1. Note that this setting:

20
docs/en/operations/system-tables/azure_queue_settings.md Normal file
View File

@ -0,0 +1,20 @@
---
slug: /en/operations/system-tables/azure_queue_settings
---
# azure_queue_settings
Contains information about settings of [AzureQueue](../../engines/table-engines/integrations/azure-queue.md) tables.
Available from `24.10` server version.
Columns:
- `database` ([String](../../sql-reference/data-types/string.md)) — Table name.
- `table` ([String](../../sql-reference/data-types/string.md)) — Database name.
- `name` ([String](../../sql-reference/data-types/string.md)) — Setting name.
- `value` ([String](../../sql-reference/data-types/string.md)) — Setting value.
- `changed` ([UInt8](../../sql-reference/data-types/int-uint.md#uint-ranges)) — Whether the setting was explicitly defined in the config or explicitly changed.
- `description` ([String](../../sql-reference/data-types/string.md)) — Setting description.
- `alterable` ([UInt8](../../sql-reference/data-types/int-uint.md#uint-ranges)) — Shows whether the setting can be changes via `ALTER TABLE ... MODIFY SETTING`.
- `0` — Current user can alter the setting.
- `1` — Current user can’t alter the setting.
- `type` ([String](../../sql-reference/data-types/string.md)) — Setting type (implementation specific string value).

6
docs/en/operations/system-tables/part_log.md
View File

@ -13,10 +13,12 @@ The `system.part_log` table contains the following columns:
- `query_id` ([String](../../sql-reference/data-types/string.md)) — Identifier of the `INSERT` query that created this data part.
- `event_type` ([Enum8](../../sql-reference/data-types/enum.md)) — Type of the event that occurred with the data part. Can have one of the following values:
- `NewPart` — Inserting of a new data part.
- `MergeParts` — Merging of data parts.
- `MergePartsStart` — Merging of data parts has started.
- `MergeParts` — Merging of data parts has finished.
- `DownloadPart` — Downloading a data part.
- `RemovePart` — Removing or detaching a data part using [DETACH PARTITION](../../sql-reference/statements/alter/partition.md#alter_detach-partition).
- `MutatePart` — Mutating of a data part.
- `MutatePartStart` — Mutating of a data part has started.
- `MutatePart` — Mutating of a data part has finished.
- `MovePart` — Moving the data part from the one disk to another one.
- `merge_reason` ([Enum8](../../sql-reference/data-types/enum.md)) — The reason for the event with type `MERGE_PARTS`. Can have one of the following values:
- `NotAMerge` — The current event has the type other than `MERGE_PARTS`.

20
docs/en/operations/system-tables/s3_queue_settings.md Normal file
View File

@ -0,0 +1,20 @@
---
slug: /en/operations/system-tables/s3_queue_settings
---
# s3_queue_settings
Contains information about settings of [S3Queue](../../engines/table-engines/integrations/s3queue.md) tables.
Available from `24.10` server version.
Columns:
- `database` ([String](../../sql-reference/data-types/string.md)) — Table name.
- `table` ([String](../../sql-reference/data-types/string.md)) — Database name.
- `name` ([String](../../sql-reference/data-types/string.md)) — Setting name.
- `value` ([String](../../sql-reference/data-types/string.md)) — Setting value.
- `changed` ([UInt8](../../sql-reference/data-types/int-uint.md#uint-ranges)) — Whether the setting was explicitly defined in the config or explicitly changed.
- `description` ([String](../../sql-reference/data-types/string.md)) — Setting description.
- `alterable` ([UInt8](../../sql-reference/data-types/int-uint.md#uint-ranges)) — Shows whether the setting can be changes via `ALTER TABLE ... MODIFY SETTING`.
- `0` — Current user can alter the setting.
- `1` — Current user can’t alter the setting.
- `type` ([String](../../sql-reference/data-types/string.md)) — Setting type (implementation specific string value).

2
docs/en/sql-reference/aggregate-functions/reference/simplelinearregression.md
View File

@ -18,7 +18,7 @@ Parameters:
Returned values:
Constants `(a, b)` of the resulting line `y = a*x + b`.
Constants `(k, b)` of the resulting line `y = k*x + b`.
**Examples**

2
docs/en/sql-reference/functions/other-functions.md
View File

@ -4393,7 +4393,7 @@ Result:
## globalVariable
Takes constant string argument and returns the value of global variable with that name. It is intended for compatibility with MySQL.
Takes a constant string argument and returns the value of the global variable with that name. This function is intended for compatibility with MySQL and not needed or useful for normal operation of ClickHouse. Only few dummy global variables are defined.
**Syntax**

10
docs/en/sql-reference/functions/string-functions.md
View File

@ -1096,7 +1096,7 @@ convertCharset(s, from, to)
## base58Encode
Encodes a String using [Base58](https://datatracker.ietf.org/doc/html/draft-msporny-base58) in the "Bitcoin" alphabet.
Encodes a string using [Base58](https://datatracker.ietf.org/doc/html/draft-msporny-base58) in the "Bitcoin" alphabet.
**Syntax**
@ -1110,7 +1110,7 @@ base58Encode(plaintext)
**Returned value**
- A string containing the encoded value of the argument. [String](../data-types/string.md).
- A string containing the encoded value of the argument. [String](../data-types/string.md) or [FixedString](../data-types/fixedstring.md).
**Example**
@ -1128,7 +1128,7 @@ Result:
## base58Decode
Accepts a String and decodes it using [Base58](https://datatracker.ietf.org/doc/html/draft-msporny-base58) encoding scheme using "Bitcoin" alphabet.
Accepts a string and decodes it using [Base58](https://datatracker.ietf.org/doc/html/draft-msporny-base58) encoding scheme using "Bitcoin" alphabet.
**Syntax**
@ -1138,7 +1138,7 @@ base58Decode(encoded)
**Arguments**
- `encoded` — [String](../data-types/string.md) column or constant. If the string is not a valid Base58-encoded value, an exception is thrown.
- `encoded` — [String](../data-types/string.md) or [FixedString](../data-types/fixedstring.md). If the string is not a valid Base58-encoded value, an exception is thrown.
**Returned value**
@ -1170,7 +1170,7 @@ tryBase58Decode(encoded)
**Parameters**
- `encoded`: [String](../data-types/string.md) column or constant. If the string is not a valid Base58-encoded value, returns an empty string in case of error.
- `encoded`: [String](../data-types/string.md) or [FixedString](../data-types/fixedstring.md). If the string is not a valid Base58-encoded value, returns an empty string in case of error.
**Returned value**

2
docs/ru/sql-reference/aggregate-functions/reference/simplelinearregression.md
View File

@ -18,7 +18,7 @@ simpleLinearRegression(x, y)
Возвращаемые значения:
Константы `(a, b)` результирующей прямой `y = a*x + b`.
Константы `(k, b)` результирующей прямой `y = k*x + b`.
**Примеры**

2
docs/zh/sql-reference/aggregate-functions/reference/simplelinearregression.md
View File

@ -20,7 +20,7 @@ simpleLinearRegression(x, y)
**返回值**
符合`y = a*x + b`的常量 `(a, b)` 。
符合`y = k*x + b`的常量 `(k, b)` 。
**示例**

5
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
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)

195
praktika/_environment.py Normal file
View File

@ -0,0 +1,195 @@
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
PARAMETER: Any = None
REPORT_INFO: List[str] = dataclasses.field(default_factory=list)
LOCAL_RUN_PARAM: str = ""
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 _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

128
praktika/_settings.py Normal file
View File

@ -0,0 +1,128 @@
import dataclasses
from pathlib import Path
from typing import Dict, Iterable, List, Optional
@dataclasses.dataclass
class _Settings:
######################################
# Pipeline generation settings #
######################################
if Path("./ci_v2").is_dir():
# TODO: hack for CH, remove
CI_PATH = "./ci_v2"
else:
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
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
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
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, ""

100
praktika/digest.py Normal file
View File

@ -0,0 +1,100 @@
import dataclasses
import hashlib
from hashlib import md5
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: hash_key [{cache_key}], include [{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):
# Calculate MD5 hash
with open(file_path, "rb") as f:
for chunk in iter(lambda: f.read(4096), b""):
hash_md5.update(chunk)
res = hash_md5.hexdigest()[: Settings.CACHE_DIGEST_LEN]
return res

60
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
praktika/environment.py Normal file
View File

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

0
praktika/execution/__init__.py Normal file
View File

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

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

31
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
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
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
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
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
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)

153
praktika/hook_html.py Normal file
View File

@ -0,0 +1,153 @@
import urllib.parse
from pathlib import Path
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 Utils
class HtmlRunnerHooks:
@classmethod
def configure(cls, _workflow):
# 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):
print(
"ERROR: Failed to set both GH commit status and PR comment with Workflow Status, cannot proceed"
)
raise
@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
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
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
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

727
praktika/json.html Normal file
View File

@ -0,0 +1,727 @@
<!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: monospace, sans-serif;
}
body.night-theme {
--background-color: #1F1F1C;
--text-color: #fff;
--tile-background: black;
}
#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;
text-align: left;
font-size: 18px;
font-weight: bold;
margin: 0; /* Remove margin */
}
#status-container button {
display: block; /* Stack buttons vertically */
width: 100%; /* Full width of container */
padding: 10px;
margin-bottom: 10px; /* Space between buttons */
background-color: #4CAF50; /* Green background color */
color: white;
border: none;
border-radius: 5px;
font-size: 16px;
cursor: pointer;
}
#status-container button:hover {
background-color: #45a049; /* Darker green on hover */
}
#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: #f4f4f4;
}
.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;
}
.json-key {
font-weight: bold;
margin-top: 10px;
}
.json-value {
margin-left: 20px;
}
.json-value a {
color: #007bff;
}
.json-value a:hover {
text-decoration: underline;
}
</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 to format timestamp to "DD-mmm-YYYY HH:MM:SS.MM"
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 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 to format duration from seconds to "HH:MM:SS"
function formatDuration(durationInSeconds) {
// 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);
// Calculate seconds and milliseconds
const seconds = Math.floor(duration); // Whole seconds
const milliseconds = Math.floor((duration % 1) * 1000); // Convert fraction to milliseconds
// Format seconds and milliseconds with leading zeros where needed
const formattedSeconds = String(seconds);
const formattedMilliseconds = String(milliseconds).padStart(3, '0');
// Return the formatted duration as seconds.milliseconds
return `${formattedSeconds}.${formattedMilliseconds}`;
}
function addKeyValueToStatus(key, value) {
const statusContainer = document.getElementById('status-container');
const keyElement = document.createElement('div');
keyElement.className = 'json-key';
keyElement.textContent = key + ':';
const valueElement = document.createElement('div');
valueElement.className = 'json-value';
valueElement.textContent = value;
statusContainer.appendChild(keyElement);
statusContainer.appendChild(valueElement);
}
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 = key + ':';
statusContainer.appendChild(keyElement);
if (Array.isArray(links) && links.length > 0) {
links.forEach(link => {
// const a = document.createElement('a');
// a.href = link;
// a.textContent = link.split('/').pop();
// a.target = '_blank';
// statusContainer.appendChild(a);
const button = document.createElement('button');
button.textContent = link.split('/').pop();
button.addEventListener('click', function () {
window.location.href = link;
});
statusContainer.appendChild(button);
});
}
}
function addStatusToStatus(status, start_time, duration) {
const statusContainer = document.getElementById('status-container');
let keyElement = document.createElement('div');
let valueElement = document.createElement('div');
keyElement.className = 'json-key';
valueElement.className = 'json-value';
keyElement.textContent = 'status:';
valueElement.classList.add('status-value');
valueElement.classList.add(getStatusClass(status));
valueElement.textContent = status;
statusContainer.appendChild(keyElement);
statusContainer.appendChild(valueElement);
keyElement = document.createElement('div');
valueElement = document.createElement('div');
keyElement.className = 'json-key';
valueElement.className = 'json-value';
keyElement.textContent = 'start_time:';
valueElement.textContent = formatTimestamp(start_time);
statusContainer.appendChild(keyElement);
statusContainer.appendChild(valueElement);
keyElement = document.createElement('div');
valueElement = document.createElement('div');
keyElement.className = 'json-key';
valueElement.className = 'json-value';
keyElement.textContent = '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);
}
statusContainer.appendChild(keyElement);
statusContainer.appendChild(valueElement);
}
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: '⚠️'
};
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);
}, 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
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
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
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
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
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="",
).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}.")
# TODO: remove
os.environ["PYTHONPATH"] = os.getcwd()
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.LOCAL_RUN_PARAM = param
env.dump()
print(f"Custom param for local tests [{param}] dumped into Environment")
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}' --volume ./:{Settings.DOCKER_WD} --volume {Settings.TEMP_DIR}:{Settings.TEMP_DIR} --workdir={Settings.DOCKER_WD} {docker} {job.command}"
else:
cmd = job.command
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
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
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
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
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
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=""):
Utils.print_formatted_error(error_message, stdout, stderr)
raise
@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
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
praktika/version.py Normal file
View File

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

68
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

349
praktika/yaml_generator.py Normal file
View File

@ -0,0 +1,349 @@
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: ${{{{ steps.run.outputs.DATA }}}}
steps:
- name: Checkout code
uses: actions/checkout@v4
{JOB_ADDONS}
- name: Prepare env script
run: |
export PYTHONPATH=.:$PYTHONPATH
cat > {ENV_SETUP_SCRIPT} << 'ENV_SETUP_SCRIPT_EOF'
{SETUP_ENVS}
cat > {WORKFLOW_CONFIG_FILE} << 'EOF'
${{{{ needs.{WORKFLOW_CONFIG_JOB_NAME}.outputs.data }}}}
EOF
cat > {WORKFLOW_STATUS_FILE} << 'EOF'
${{{{ toJson(needs) }}}}
EOF
ENV_SETUP_SCRIPT_EOF
rm -rf {INPUT_DIR} {OUTPUT_DIR} {TEMP_DIR}
mkdir -p {TEMP_DIR} {INPUT_DIR} {OUTPUT_DIR}
{DOWNLOADS_GITHUB}
- name: Run
id: run
run: |
set -o pipefail
{PYTHON} -m praktika run --job '''{JOB_NAME}''' --workflow "{WORKFLOW_NAME}" --ci |& tee {RUN_LOG}
{UPLOADS_GITHUB}\
"""
TEMPLATE_SETUP_ENV_SECRETS = """\
export {SECRET_NAME}=$(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)

9
src/AggregateFunctions/IAggregateFunction.cpp
View File

@ -10,6 +10,15 @@ DataTypePtr IAggregateFunction::getStateType() const
return std::make_shared<DataTypeAggregateFunction>(shared_from_this(), argument_types, parameters);
}
DataTypePtr IAggregateFunction::getNormalizedStateType() const
{
DataTypes normalized_argument_types;
normalized_argument_types.reserve(argument_types.size());
for (const auto & arg : argument_types)
normalized_argument_types.emplace_back(arg->getNormalizedType());
return std::make_shared<DataTypeAggregateFunction>(shared_from_this(), normalized_argument_types, parameters);
}
String IAggregateFunction::getDescription() const
{
String description;

2
src/AggregateFunctions/IAggregateFunction.h
View File

@ -73,7 +73,7 @@ public:
virtual DataTypePtr getStateType() const;
/// Same as the above but normalize state types so that variants with the same binary representation will use the same type.
virtual DataTypePtr getNormalizedStateType() const { return getStateType(); }
virtual DataTypePtr getNormalizedStateType() const;
/// Returns true if two aggregate functions have the same state representation in memory and the same serialization,
/// so state of one aggregate function can be safely used with another.

20
src/Analyzer/Passes/LogicalExpressionOptimizerPass.cpp
View File

@ -184,6 +184,26 @@ public:
, join_node(join_node_)
{}
bool needChildVisit(const QueryTreeNodePtr & parent, const QueryTreeNodePtr &)
{
/** Optimization can change the value of some expression from NULL to FALSE.
* For example:
* when `a` is `NULL`, the expression `a = b AND a IS NOT NULL` returns `NULL`
* and it will be optimized to `a = b`, which returns `FALSE`.
* This is valid for JOIN ON condition and for the functions `AND`/`OR` inside it.
* (When we replace `AND`/`OR` operands from `NULL` to `FALSE`, the result value can also change only from `NULL` to `FALSE`)
* However, in the general case, the result can be wrong.
* For example, for NOT: `NOT NULL` is `NULL`, but `NOT FALSE` is `TRUE`.
* Therefore, optimize only top-level expression or expressions inside `AND`/`OR`.
*/
if (const auto * function_node = parent->as<FunctionNode>())
{
const auto & func_name = function_node->getFunctionName();
return func_name == "or" || func_name == "and";
}
return parent->getNodeType() == QueryTreeNodeType::LIST;
}
void enterImpl(QueryTreeNodePtr & node)
{
auto * function_node = node->as<FunctionNode>();

96
src/Backups/BackupIO_S3.cpp
View File

@ -10,6 +10,7 @@
#include <IO/WriteBufferFromS3.h>
#include <IO/HTTPHeaderEntries.h>
#include <IO/S3/copyS3File.h>
#include <IO/S3/deleteFileFromS3.h>
#include <IO/S3/Client.h>
#include <IO/S3/Credentials.h>
#include <Disks/IDisk.h>
@ -117,12 +118,6 @@ namespace
throw S3Exception(outcome.GetError().GetMessage(), outcome.GetError().GetErrorType());
return outcome.GetResult().GetContents();
}
bool isNotFoundError(Aws::S3::S3Errors error)
{
return error == Aws::S3::S3Errors::RESOURCE_NOT_FOUND
|| error == Aws::S3::S3Errors::NO_SUCH_KEY;
}
}
@ -236,6 +231,7 @@ BackupWriterS3::BackupWriterS3(
: BackupWriterDefault(read_settings_, write_settings_, getLogger("BackupWriterS3"))
, s3_uri(s3_uri_)
, data_source_description{DataSourceType::ObjectStorage, ObjectStorageType::S3, MetadataStorageType::None, s3_uri.endpoint, false, false}
, s3_capabilities(getCapabilitiesFromConfig(context_->getConfigRef(), "s3"))
{
s3_settings.loadFromConfig(context_->getConfigRef(), "s3", context_->getSettingsRef());
@ -358,92 +354,22 @@ std::unique_ptr<WriteBuffer> BackupWriterS3::writeFile(const String & file_name)
void BackupWriterS3::removeFile(const String & file_name)
{
S3::DeleteObjectRequest request;
request.SetBucket(s3_uri.bucket);
auto key = fs::path(s3_uri.key) / file_name;
request.SetKey(key);
auto outcome = client->DeleteObject(request);
if (blob_storage_log)
{
blob_storage_log->addEvent(
BlobStorageLogElement::EventType::Delete,
s3_uri.bucket, key, /* local_path */ "", /* data_size */ 0,
outcome.IsSuccess() ? nullptr : &outcome.GetError());
}
if (!outcome.IsSuccess() && !isNotFoundError(outcome.GetError().GetErrorType()))
throw S3Exception(outcome.GetError().GetMessage(), outcome.GetError().GetErrorType());
deleteFileFromS3(client, s3_uri.bucket, fs::path(s3_uri.key) / file_name, /* if_exists = */ false,
blob_storage_log);
}
void BackupWriterS3::removeFiles(const Strings & file_names)
{
try
{
if (!supports_batch_delete.has_value() || supports_batch_delete.value() == true)
{
removeFilesBatch(file_names);
supports_batch_delete = true;
}
else
{
for (const auto & file_name : file_names)
removeFile(file_name);
}
}
catch (const Exception &)
{
if (!supports_batch_delete.has_value())
{
supports_batch_delete = false;
LOG_TRACE(log, "DeleteObjects is not supported. Retrying with plain DeleteObject.");
Strings keys;
keys.reserve(file_names.size());
for (const String & file_name : file_names)
keys.push_back(fs::path(s3_uri.key) / file_name);
for (const auto & file_name : file_names)
removeFile(file_name);
}
else
throw;
}
}
void BackupWriterS3::removeFilesBatch(const Strings & file_names)
{
/// One call of DeleteObjects() cannot remove more than 1000 keys.
size_t chunk_size_limit = 1000;
size_t batch_size = 1000;
size_t current_position = 0;
while (current_position < file_names.size())
{
std::vector<Aws::S3::Model::ObjectIdentifier> current_chunk;
for (; current_position < file_names.size() && current_chunk.size() < chunk_size_limit; ++current_position)
{
Aws::S3::Model::ObjectIdentifier obj;
obj.SetKey(fs::path(s3_uri.key) / file_names[current_position]);
current_chunk.push_back(obj);
}
Aws::S3::Model::Delete delkeys;
delkeys.SetObjects(current_chunk);
S3::DeleteObjectsRequest request;
request.SetBucket(s3_uri.bucket);
request.SetDelete(delkeys);
auto outcome = client->DeleteObjects(request);
if (blob_storage_log)
{
const auto * outcome_error = outcome.IsSuccess() ? nullptr : &outcome.GetError();
auto time_now = std::chrono::system_clock::now();
for (const auto & obj : current_chunk)
blob_storage_log->addEvent(BlobStorageLogElement::EventType::Delete, s3_uri.bucket, obj.GetKey(),
/* local_path */ "", /* data_size */ 0, outcome_error, time_now);
}
if (!outcome.IsSuccess() && !isNotFoundError(outcome.GetError().GetErrorType()))
throw S3Exception(outcome.GetError().GetMessage(), outcome.GetError().GetErrorType());
}
deleteFilesFromS3(client, s3_uri.bucket, keys, /* if_exists = */ false,
s3_capabilities, batch_size, blob_storage_log);
}
}

5
src/Backups/BackupIO_S3.h
View File

@ -10,6 +10,7 @@
#include <IO/S3Settings.h>
#include <Interpreters/Context_fwd.h>
#include <IO/S3/BlobStorageLogWriter.h>
#include <IO/S3/S3Capabilities.h>
namespace DB
{
@ -76,14 +77,12 @@ public:
private:
std::unique_ptr<ReadBuffer> readFile(const String & file_name, size_t expected_file_size) override;
void removeFilesBatch(const Strings & file_names);
const S3::URI s3_uri;
const DataSourceDescription data_source_description;
S3Settings s3_settings;
std::shared_ptr<S3::Client> client;
std::optional<bool> supports_batch_delete;
S3Capabilities s3_capabilities;
BlobStorageLogWriterPtr blob_storage_log;
};

21
src/Common/Config/ConfigProcessor.cpp
View File

@ -671,9 +671,11 @@ XMLDocumentPtr ConfigProcessor::parseConfig(const std::string & config_path)
XMLDocumentPtr ConfigProcessor::processConfig(
bool * has_zk_includes,
zkutil::ZooKeeperNodeCache * zk_node_cache,
const zkutil::EventPtr & zk_changed_event)
const zkutil::EventPtr & zk_changed_event,
bool is_config_changed)
{
LOG_DEBUG(log, "Processing configuration file '{}'.", path);
if (is_config_changed)
LOG_DEBUG(log, "Processing configuration file '{}'.", path);
XMLDocumentPtr config;
@ -686,7 +688,8 @@ XMLDocumentPtr ConfigProcessor::processConfig(
/// When we can use a config embedded in the binary.
if (auto it = embedded_configs.find(path); it != embedded_configs.end())
{
LOG_DEBUG(log, "There is no file '{}', will use embedded config.", path);
if (is_config_changed)
LOG_DEBUG(log, "There is no file '{}', will use embedded config.", path);
config = dom_parser.parseMemory(it->second.data(), it->second.size());
}
else
@ -700,7 +703,8 @@ XMLDocumentPtr ConfigProcessor::processConfig(
{
try
{
LOG_DEBUG(log, "Merging configuration file '{}'.", merge_file);
if (is_config_changed)
LOG_DEBUG(log, "Merging configuration file '{}'.", merge_file);
XMLDocumentPtr with;
with = parseConfig(merge_file);
@ -790,10 +794,10 @@ XMLDocumentPtr ConfigProcessor::processConfig(
return config;
}
ConfigProcessor::LoadedConfig ConfigProcessor::loadConfig(bool allow_zk_includes)
ConfigProcessor::LoadedConfig ConfigProcessor::loadConfig(bool allow_zk_includes, bool is_config_changed)
{
bool has_zk_includes;
XMLDocumentPtr config_xml = processConfig(&has_zk_includes);
XMLDocumentPtr config_xml = processConfig(&has_zk_includes, nullptr, nullptr, is_config_changed);
if (has_zk_includes && !allow_zk_includes)
throw Poco::Exception("Error while loading config '" + path + "': from_zk includes are not allowed!");
@ -806,14 +810,15 @@ ConfigProcessor::LoadedConfig ConfigProcessor::loadConfig(bool allow_zk_includes
ConfigProcessor::LoadedConfig ConfigProcessor::loadConfigWithZooKeeperIncludes(
zkutil::ZooKeeperNodeCache & zk_node_cache,
const zkutil::EventPtr & zk_changed_event,
bool fallback_to_preprocessed)
bool fallback_to_preprocessed,
bool is_config_changed)
{
XMLDocumentPtr config_xml;
bool has_zk_includes;
bool processed_successfully = false;
try
{
config_xml = processConfig(&has_zk_includes, &zk_node_cache, zk_changed_event);
config_xml = processConfig(&has_zk_includes, &zk_node_cache, zk_changed_event, is_config_changed);
processed_successfully = true;
}
catch (const Poco::Exception & ex)

8
src/Common/Config/ConfigProcessor.h
View File

@ -63,7 +63,8 @@ public:
XMLDocumentPtr processConfig(
bool * has_zk_includes = nullptr,
zkutil::ZooKeeperNodeCache * zk_node_cache = nullptr,
const zkutil::EventPtr & zk_changed_event = nullptr);
const zkutil::EventPtr & zk_changed_event = nullptr,
bool is_config_changed = true);
XMLDocumentPtr parseConfig(const std::string & config_path);
@ -88,14 +89,15 @@ public:
/// If allow_zk_includes is true, expect that the configuration XML can contain from_zk nodes.
/// If it is the case, set has_zk_includes to true and don't write config-preprocessed.xml,
/// expecting that config would be reloaded with zookeeper later.
LoadedConfig loadConfig(bool allow_zk_includes = false);
LoadedConfig loadConfig(bool allow_zk_includes = false, bool is_config_changed = true);
/// If fallback_to_preprocessed is true, then if KeeperException is thrown during config
/// processing, load the configuration from the preprocessed file.
LoadedConfig loadConfigWithZooKeeperIncludes(
zkutil::ZooKeeperNodeCache & zk_node_cache,
const zkutil::EventPtr & zk_changed_event,
bool fallback_to_preprocessed = false);
bool fallback_to_preprocessed = false,
bool is_config_changed = true);
/// Save preprocessed config to specified directory.
/// If preprocessed_dir is empty - calculate from loaded_config.path + /preprocessed_configs/

7
src/Common/Config/ConfigReloader.cpp
View File

@ -111,7 +111,8 @@ std::optional<ConfigProcessor::LoadedConfig> ConfigReloader::reloadIfNewer(bool
std::lock_guard lock(reload_mutex);
FilesChangesTracker new_files = getNewFileList();
if (force || need_reload_from_zk || new_files.isDifferOrNewerThan(files))
const bool is_config_changed = new_files.isDifferOrNewerThan(files);
if (force || need_reload_from_zk || is_config_changed)
{
ConfigProcessor config_processor(config_path);
ConfigProcessor::LoadedConfig loaded_config;
@ -120,10 +121,10 @@ std::optional<ConfigProcessor::LoadedConfig> ConfigReloader::reloadIfNewer(bool
try
{
loaded_config = config_processor.loadConfig(/* allow_zk_includes = */ true);
loaded_config = config_processor.loadConfig(/* allow_zk_includes = */ true, is_config_changed);
if (loaded_config.has_zk_includes)
loaded_config = config_processor.loadConfigWithZooKeeperIncludes(
zk_node_cache, zk_changed_event, fallback_to_preprocessed);
zk_node_cache, zk_changed_event, fallback_to_preprocessed, is_config_changed);
}
catch (const Coordination::Exception & e)
{

39
src/Common/HashTable/HashTable.h
View File

@ -67,19 +67,6 @@ struct HashTableNoState
};
/// These functions can be overloaded for custom types.
namespace ZeroTraits
{
template <typename T>
bool check(const T x) { return x == T{}; }
template <typename T>
void set(T & x) { x = T{}; }
}
/** Numbers are compared bitwise.
* Complex types are compared by operator== as usual (this is important if there are gaps).
*
@ -87,18 +74,32 @@ void set(T & x) { x = T{}; }
* Otherwise the invariants in hash table probing do not met when NaNs are present.
*/
template <typename T>
inline bool bitEquals(T && a, T && b)
inline bool bitEquals(T a, T b)
{
using RealT = std::decay_t<T>;
if constexpr (std::is_floating_point_v<RealT>)
/// Note that memcmp with constant size is compiler builtin.
return 0 == memcmp(&a, &b, sizeof(RealT)); /// NOLINT
if constexpr (std::is_floating_point_v<T>)
/// Note that memcmp with constant size is a compiler builtin.
return 0 == memcmp(&a, &b, sizeof(T)); /// NOLINT
else
return a == b;
}
/// These functions can be overloaded for custom types.
namespace ZeroTraits
{
template <typename T>
bool check(const T x)
{
return bitEquals(x, T{});
}
template <typename T>
void set(T & x) { x = T{}; }
}
/**
* getKey/Mapped -- methods to get key/"mapped" values from the LookupResult returned by find() and
* emplace() methods of HashTable. Must not be called for a null LookupResult.

109
src/Coordination/CoordinationSettings.cpp
View File

@ -1,5 +1,6 @@
#include <Coordination/CoordinationSettings.h>
#include <Core/BaseSettings.h>
#include <Core/BaseSettingsFwdMacrosImpl.h>
#include <IO/WriteHelpers.h>
#include <IO/WriteIntText.h>
#include <Common/ZooKeeper/ZooKeeperConstants.h>
@ -16,62 +17,61 @@ namespace ErrorCodes
/** These settings represent fine tunes for internal details of Coordination storages
* and should not be changed by the user without a reason.
*/
#define LIST_OF_COORDINATION_SETTINGS(M, ALIAS) \
M(Milliseconds, min_session_timeout_ms, Coordination::DEFAULT_MIN_SESSION_TIMEOUT_MS, "Min client session timeout", 0) \
M(Milliseconds, session_timeout_ms, Coordination::DEFAULT_MAX_SESSION_TIMEOUT_MS, "Max client session timeout", 0) \
M(Milliseconds, operation_timeout_ms, Coordination::DEFAULT_OPERATION_TIMEOUT_MS, "Default client operation timeout", 0) \
M(Milliseconds, dead_session_check_period_ms, 500, "How often leader will check sessions to consider them dead and remove", 0) \
M(Milliseconds, heart_beat_interval_ms, 500, "Heartbeat interval between quorum nodes", 0) \
M(Milliseconds, election_timeout_lower_bound_ms, 1000, "Lower bound of election timer (avoid too often leader elections)", 0) \
M(Milliseconds, election_timeout_upper_bound_ms, 2000, "Upper bound of election timer (avoid too often leader elections)", 0) \
M(Milliseconds, leadership_expiry_ms, 0, "Duration after which a leader will expire if it fails to receive responses from peers. Set it lower or equal to election_timeout_lower_bound_ms to avoid multiple leaders.", 0) \
M(UInt64, reserved_log_items, 100000, "How many log items to store (don't remove during compaction)", 0) \
M(UInt64, snapshot_distance, 100000, "How many log items we have to collect to write new snapshot", 0) \
M(Bool, auto_forwarding, true, "Allow to forward write requests from followers to leader", 0) \
M(Milliseconds, shutdown_timeout, 5000, "How much time we will wait until RAFT shutdown", 0) \
M(Milliseconds, session_shutdown_timeout, 10000, "How much time we will wait until sessions are closed during shutdown", 0) \
M(Milliseconds, startup_timeout, 180000, "How much time we will wait until RAFT to start.", 0) \
M(Milliseconds, sleep_before_leader_change_ms, 8000, "How much time we will wait before removing leader (so as leader could commit accepted but non-committed commands and they won't be lost -- leader removal is not synchronized with committing)", 0) \
M(LogsLevel, raft_logs_level, LogsLevel::information, "Log internal RAFT logs into main server log level. Valid values: 'trace', 'debug', 'information', 'warning', 'error', 'fatal', 'none'", 0) \
M(UInt64, rotate_log_storage_interval, 100000, "How many records will be stored in one log storage file", 0) \
M(UInt64, snapshots_to_keep, 3, "How many compressed snapshots to keep on disk", 0) \
M(UInt64, stale_log_gap, 10000, "When node became stale and should receive snapshots from leader", 0) \
M(UInt64, fresh_log_gap, 200, "When node became fresh", 0) \
M(UInt64, max_request_queue_size, 100000, "Maximum number of request that can be in queue for processing", 0) \
M(UInt64, max_requests_batch_size, 100, "Max size of batch of requests that can be sent to RAFT", 0) \
M(UInt64, max_requests_batch_bytes_size, 100*1024, "Max size in bytes of batch of requests that can be sent to RAFT", 0) \
M(UInt64, max_requests_append_size, 100, "Max size of batch of requests that can be sent to replica in append request", 0) \
M(UInt64, max_flush_batch_size, 1000, "Max size of batch of requests that can be flushed together", 0) \
M(UInt64, max_requests_quick_batch_size, 100, "Max size of batch of requests to try to get before proceeding with RAFT. Keeper will not wait for requests but take only requests that are already in queue" , 0) \
M(Bool, quorum_reads, false, "Execute read requests as writes through whole RAFT consesus with similar speed", 0) \
M(Bool, force_sync, true, "Call fsync on each change in RAFT changelog", 0) \
M(Bool, compress_logs, false, "Write compressed coordination logs in ZSTD format", 0) \
M(Bool, compress_snapshots_with_zstd_format, true, "Write compressed snapshots in ZSTD format (instead of custom LZ4)", 0) \
M(UInt64, configuration_change_tries_count, 20, "How many times we will try to apply configuration change (add/remove server) to the cluster", 0) \
M(UInt64, max_log_file_size, 50 * 1024 * 1024, "Max size of the Raft log file. If possible, each created log file will preallocate this amount of bytes on disk. Set to 0 to disable the limit", 0) \
M(UInt64, log_file_overallocate_size, 50 * 1024 * 1024, "If max_log_file_size is not set to 0, this value will be added to it for preallocating bytes on disk. If a log record is larger than this value, it could lead to uncaught out-of-space issues so a larger value is preferred", 0) \
M(UInt64, min_request_size_for_cache, 50 * 1024, "Minimal size of the request to cache the deserialization result. Caching can have negative effect on latency for smaller requests, set to 0 to disable", 0) \
M(UInt64, raft_limits_reconnect_limit, 50, "If connection to a peer is silent longer than this limit * (multiplied by heartbeat interval), we re-establish the connection.", 0) \
M(UInt64, raft_limits_response_limit, 20, "Total wait time for a response is calculated by multiplying response_limit with heart_beat_interval_ms", 0) \
M(Bool, async_replication, false, "Enable async replication. All write and read guarantees are preserved while better performance is achieved. Settings is disabled by default to not break backwards compatibility.", 0) \
M(Bool, experimental_use_rocksdb, false, "Use rocksdb as backend storage", 0) \
M(UInt64, latest_logs_cache_size_threshold, 1 * 1024 * 1024 * 1024, "Maximum total size of in-memory cache of latest log entries.", 0) \
M(UInt64, commit_logs_cache_size_threshold, 500 * 1024 * 1024, "Maximum total size of in-memory cache of log entries needed next for commit.", 0) \
M(UInt64, disk_move_retries_wait_ms, 1000, "How long to wait between retries after a failure which happened while a file was being moved between disks.", 0) \
M(UInt64, disk_move_retries_during_init, 100, "The amount of retries after a failure which happened while a file was being moved between disks during initialization.", 0) \
M(UInt64, log_slow_total_threshold_ms, 5000, "Requests for which the total latency is larger than this settings will be logged", 0) \
M(UInt64, log_slow_cpu_threshold_ms, 100, "Requests for which the CPU (preprocessing and processing) latency is larger than this settings will be logged", 0) \
M(UInt64, log_slow_connection_operation_threshold_ms, 1000, "Log message if a certain operation took too long inside a single connection", 0)
#define LIST_OF_COORDINATION_SETTINGS(DECLARE, ALIAS) \
DECLARE(Milliseconds, min_session_timeout_ms, Coordination::DEFAULT_MIN_SESSION_TIMEOUT_MS, "Min client session timeout", 0) \
DECLARE(Milliseconds, session_timeout_ms, Coordination::DEFAULT_MAX_SESSION_TIMEOUT_MS, "Max client session timeout", 0) \
DECLARE(Milliseconds, operation_timeout_ms, Coordination::DEFAULT_OPERATION_TIMEOUT_MS, "Default client operation timeout", 0) \
DECLARE(Milliseconds, dead_session_check_period_ms, 500, "How often leader will check sessions to consider them dead and remove", 0) \
DECLARE(Milliseconds, heart_beat_interval_ms, 500, "Heartbeat interval between quorum nodes", 0) \
DECLARE(Milliseconds, election_timeout_lower_bound_ms, 1000, "Lower bound of election timer (avoid too often leader elections)", 0) \
DECLARE(Milliseconds, election_timeout_upper_bound_ms, 2000, "Upper bound of election timer (avoid too often leader elections)", 0) \
DECLARE(Milliseconds, leadership_expiry_ms, 0, "Duration after which a leader will expire if it fails to receive responses from peers. Set it lower or equal to election_timeout_lower_bound_ms to avoid multiple leaders.", 0) \
DECLARE(UInt64, reserved_log_items, 100000, "How many log items to store (don't remove during compaction)", 0) \
DECLARE(UInt64, snapshot_distance, 100000, "How many log items we have to collect to write new snapshot", 0) \
DECLARE(Bool, auto_forwarding, true, "Allow to forward write requests from followers to leader", 0) \
DECLARE(Milliseconds, shutdown_timeout, 5000, "How much time we will wait until RAFT shutdown", 0) \
DECLARE(Milliseconds, session_shutdown_timeout, 10000, "How much time we will wait until sessions are closed during shutdown", 0) \
DECLARE(Milliseconds, startup_timeout, 180000, "How much time we will wait until RAFT to start.", 0) \
DECLARE(Milliseconds, sleep_before_leader_change_ms, 8000, "How much time we will wait before removing leader (so as leader could commit accepted but non-committed commands and they won't be lost -- leader removal is not synchronized with committing)", 0) \
DECLARE(LogsLevel, raft_logs_level, LogsLevel::information, "Log internal RAFT logs into main server log level. Valid values: 'trace', 'debug', 'information', 'warning', 'error', 'fatal', 'none'", 0) \
DECLARE(UInt64, rotate_log_storage_interval, 100000, "How many records will be stored in one log storage file", 0) \
DECLARE(UInt64, snapshots_to_keep, 3, "How many compressed snapshots to keep on disk", 0) \
DECLARE(UInt64, stale_log_gap, 10000, "When node became stale and should receive snapshots from leader", 0) \
DECLARE(UInt64, fresh_log_gap, 200, "When node became fresh", 0) \
DECLARE(UInt64, max_request_queue_size, 100000, "Maximum number of request that can be in queue for processing", 0) \
DECLARE(UInt64, max_requests_batch_size, 100, "Max size of batch of requests that can be sent to RAFT", 0) \
DECLARE(UInt64, max_requests_batch_bytes_size, 100*1024, "Max size in bytes of batch of requests that can be sent to RAFT", 0) \
DECLARE(UInt64, max_requests_append_size, 100, "Max size of batch of requests that can be sent to replica in append request", 0) \
DECLARE(UInt64, max_flush_batch_size, 1000, "Max size of batch of requests that can be flushed together", 0) \
DECLARE(UInt64, max_requests_quick_batch_size, 100, "Max size of batch of requests to try to get before proceeding with RAFT. Keeper will not wait for requests but take only requests that are already in queue" , 0) \
DECLARE(Bool, quorum_reads, false, "Execute read requests as writes through whole RAFT consesus with similar speed", 0) \
DECLARE(Bool, force_sync, true, "Call fsync on each change in RAFT changelog", 0) \
DECLARE(Bool, compress_logs, false, "Write compressed coordination logs in ZSTD format", 0) \
DECLARE(Bool, compress_snapshots_with_zstd_format, true, "Write compressed snapshots in ZSTD format (instead of custom LZ4)", 0) \
DECLARE(UInt64, configuration_change_tries_count, 20, "How many times we will try to apply configuration change (add/remove server) to the cluster", 0) \
DECLARE(UInt64, max_log_file_size, 50 * 1024 * 1024, "Max size of the Raft log file. If possible, each created log file will preallocate this amount of bytes on disk. Set to 0 to disable the limit", 0) \
DECLARE(UInt64, log_file_overallocate_size, 50 * 1024 * 1024, "If max_log_file_size is not set to 0, this value will be added to it for preallocating bytes on disk. If a log record is larger than this value, it could lead to uncaught out-of-space issues so a larger value is preferred", 0) \
DECLARE(UInt64, min_request_size_for_cache, 50 * 1024, "Minimal size of the request to cache the deserialization result. Caching can have negative effect on latency for smaller requests, set to 0 to disable", 0) \
DECLARE(UInt64, raft_limits_reconnect_limit, 50, "If connection to a peer is silent longer than this limit * (multiplied by heartbeat interval), we re-establish the connection.", 0) \
DECLARE(UInt64, raft_limits_response_limit, 20, "Total wait time for a response is calculated by multiplying response_limit with heart_beat_interval_ms", 0) \
DECLARE(Bool, async_replication, false, "Enable async replication. All write and read guarantees are preserved while better performance is achieved. Settings is disabled by default to not break backwards compatibility.", 0) \
DECLARE(Bool, experimental_use_rocksdb, false, "Use rocksdb as backend storage", 0) \
DECLARE(UInt64, latest_logs_cache_size_threshold, 1 * 1024 * 1024 * 1024, "Maximum total size of in-memory cache of latest log entries.", 0) \
DECLARE(UInt64, commit_logs_cache_size_threshold, 500 * 1024 * 1024, "Maximum total size of in-memory cache of log entries needed next for commit.", 0) \
DECLARE(UInt64, disk_move_retries_wait_ms, 1000, "How long to wait between retries after a failure which happened while a file was being moved between disks.", 0) \
DECLARE(UInt64, disk_move_retries_during_init, 100, "The amount of retries after a failure which happened while a file was being moved between disks during initialization.", 0) \
DECLARE(UInt64, log_slow_total_threshold_ms, 5000, "Requests for which the total latency is larger than this settings will be logged", 0) \
DECLARE(UInt64, log_slow_cpu_threshold_ms, 100, "Requests for which the CPU (preprocessing and processing) latency is larger than this settings will be logged", 0) \
DECLARE(UInt64, log_slow_connection_operation_threshold_ms, 1000, "Log message if a certain operation took too long inside a single connection", 0)
DECLARE_SETTINGS_TRAITS(CoordinationSettingsTraits, LIST_OF_COORDINATION_SETTINGS)
IMPLEMENT_SETTINGS_TRAITS(CoordinationSettingsTraits, LIST_OF_COORDINATION_SETTINGS)
struct CoordinationSettingsImpl : public BaseSettings<CoordinationSettingsTraits>
{
void loadFromConfig(const String & config_elem, const Poco::Util::AbstractConfiguration & config);
};
IMPLEMENT_SETTINGS_TRAITS(CoordinationSettingsTraits, LIST_OF_COORDINATION_SETTINGS)
void CoordinationSettingsImpl::loadFromConfig(const String & config_elem, const Poco::Util::AbstractConfiguration & config)
{
if (!config.has(config_elem))
@ -99,7 +99,7 @@ void CoordinationSettingsImpl::loadFromConfig(const String & config_elem, const
}
#define INITIALIZE_SETTING_EXTERN(TYPE, NAME, DEFAULT, DESCRIPTION, FLAGS) \
CoordinationSettings##TYPE NAME = &CoordinationSettings##Impl ::NAME;
CoordinationSettings##TYPE NAME = &CoordinationSettingsImpl ::NAME;
namespace CoordinationSetting
{
@ -119,18 +119,7 @@ CoordinationSettings::CoordinationSettings(const CoordinationSettings & settings
CoordinationSettings::~CoordinationSettings() = default;
#define IMPLEMENT_SETTING_SUBSCRIPT_OPERATOR(CLASS_NAME, TYPE) \
const SettingField##TYPE & CoordinationSettings::operator[](CLASS_NAME##TYPE t) const \
{ \
return impl.get()->*t; \
} \
SettingField##TYPE & CoordinationSettings::operator[](CLASS_NAME##TYPE t) \
{ \
return impl.get()->*t; \
}
COORDINATION_SETTINGS_SUPPORTED_TYPES(CoordinationSettings, IMPLEMENT_SETTING_SUBSCRIPT_OPERATOR)
#undef IMPLEMENT_SETTING_SUBSCRIPT_OPERATOR
void CoordinationSettings::loadFromConfig(const String & config_elem, const Poco::Util::AbstractConfiguration & config)
{

2
src/Coordination/LoggerWrapper.h
View File

@ -1,8 +1,8 @@
#pragma once
#include <Core/LogsLevel.h>
#include <libnuraft/nuraft.hxx>
#include <Common/logger_useful.h>
#include <Core/SettingsEnums.h>
namespace DB
{

49
src/Core/BaseSettings.h
View File

@ -22,22 +22,57 @@ class ReadBuffer;
class WriteBuffer;
/** Template class to define collections of settings.
* If you create a new setting, please also add it to ./utils/check-style/check-settings-style
* for validation
*
* Example of usage:
*
* mysettings.h:
* #define APPLY_FOR_MYSETTINGS(M) \
* M(UInt64, a, 100, "Description of a", 0) \
* M(Float, f, 3.11, "Description of f", IMPORTANT) // IMPORTANT - means the setting can't be ignored by older versions) \
* M(String, s, "default", "Description of s", 0)
* #include <Core/BaseSettingsFwdMacros.h>
* #include <Core/SettingsFields.h>
*
* DECLARE_SETTINGS_TRAITS(MySettingsTraits, APPLY_FOR_MYSETTINGS)
* struct MySettings : public BaseSettings<MySettingsTraits>
* #define MY_SETTINGS_SUPPORTED_TYPES(CLASS_NAME, M) \
* M(CLASS_NAME, Float) \
* M(CLASS_NAME, String) \
* M(CLASS_NAME, UInt64)
*
* MY_SETTINGS_SUPPORTED_TYPES(MySettings, DECLARE_SETTING_TRAIT)
*
* struct MySettings
* {
* MySettings();
* ~MySettings();
*
* MY_SETTINGS_SUPPORTED_TYPES(MySettings, DECLARE_SETTING_SUBSCRIPT_OPERATOR)
* private:
* std::unique_ptr<MySettingsImpl> impl;
* };
*
* mysettings.cpp:
* #include <Core/BaseSettings.h>
* #include <Core/BaseSettingsFwdMacrosImpl.h>
*
* #define APPLY_FOR_MYSETTINGS(DECLARE, ALIAS) \
* DECLARE(UInt64, a, 100, "Description of a", 0) \
* DECLARE(Float, f, 3.11, "Description of f", IMPORTANT) // IMPORTANT - means the setting can't be ignored by older versions) \
* DECLARE(String, s, "default", "Description of s", 0)
*
* DECLARE_SETTINGS_TRAITS(MySettingsTraits, APPLY_FOR_MYSETTINGS)
* IMPLEMENT_SETTINGS_TRAITS(MySettingsTraits, APPLY_FOR_MYSETTINGS)
*
* struct MySettingsImpl : public BaseSettings<MySettingsTraits>
* {
* };
*
* #define INITIALIZE_SETTING_EXTERN(TYPE, NAME, DEFAULT, DESCRIPTION, FLAGS) MySettings##TYPE NAME = &MySettings##Impl ::NAME;
*
* namespace MySetting
* {
* APPLY_FOR_MYSETTINGS(INITIALIZE_SETTING_EXTERN, SKIP_ALIAS)
* }
* #undef INITIALIZE_SETTING_EXTERN
*
* MY_SETTINGS_SUPPORTED_TYPES(MySettings, IMPLEMENT_SETTING_SUBSCRIPT_OPERATOR)
*/
template <class TTraits>
class BaseSettings : public TTraits::Data

11
src/Core/BaseSettingsFwdMacrosImpl.h Normal file
View File

@ -0,0 +1,11 @@
#pragma once
#define IMPLEMENT_SETTING_SUBSCRIPT_OPERATOR(CLASS_NAME, TYPE) \
const SettingField##TYPE & CLASS_NAME::operator[](CLASS_NAME##TYPE t) const \
{ \
return impl.get()->*t; \
} \
SettingField##TYPE & CLASS_NAME::operator[](CLASS_NAME##TYPE t) \
{ \
return impl.get()->*t; \
}

6
src/Core/FormatFactorySettings.cpp
View File

@ -1,5 +1,5 @@
#include <Core/BaseSettings.h>
#include <Core/FormatFactorySettingsDeclaration.h>
#include <Core/FormatFactorySettings.h>
#include <Core/SettingsEnums.h>
namespace DB
@ -8,11 +8,9 @@ namespace DB
* User-specified file format settings for File and URL engines.
*/
DECLARE_SETTINGS_TRAITS(FormatFactorySettingsTraits, LIST_OF_ALL_FORMAT_SETTINGS)
IMPLEMENT_SETTINGS_TRAITS(FormatFactorySettingsTraits, LIST_OF_ALL_FORMAT_SETTINGS)
struct FormatFactorySettingsImpl : public BaseSettings<FormatFactorySettingsTraits>
{
};
IMPLEMENT_SETTINGS_TRAITS(FormatFactorySettingsTraits, LIST_OF_ALL_FORMAT_SETTINGS)
}

1299
src/Core/FormatFactorySettings.h
View File

File diff suppressed because it is too large Load Diff

1259
src/Core/FormatFactorySettingsDeclaration.h
View File

File diff suppressed because it is too large Load Diff

329
src/Core/ServerSettings.cpp
View File

@ -1,5 +1,6 @@
#include <Core/BackgroundSchedulePool.h>
#include <Core/BaseSettings.h>
#include <Core/BaseSettingsFwdMacrosImpl.h>
#include <Core/ServerSettings.h>
#include <IO/MMappedFileCache.h>
#include <IO/UncompressedCache.h>
@ -25,182 +26,181 @@ extern const Metric BackgroundMessageBrokerSchedulePoolSize;
namespace DB
{
#define LIST_OF_SERVER_SETTINGS(M, ALIAS) \
M(Bool, show_addresses_in_stack_traces, true, "If it is set true will show addresses in stack traces", 0) \
M(Bool, shutdown_wait_unfinished_queries, false, "If set true ClickHouse will wait for running queries finish before shutdown.", 0) \
M(UInt64, shutdown_wait_unfinished, 5, "Delay in seconds to wait for unfinished queries", 0) \
M(UInt64, max_thread_pool_size, 10000, "The maximum number of threads that could be allocated from the OS and used for query execution and background operations.", 0) \
M(UInt64, max_thread_pool_free_size, 1000, "The maximum number of threads that will always stay in a global thread pool once allocated and remain idle in case of insufficient number of tasks.", 0) \
M(UInt64, thread_pool_queue_size, 10000, "The maximum number of tasks that will be placed in a queue and wait for execution.", 0) \
M(UInt64, max_io_thread_pool_size, 100, "The maximum number of threads that would be used for IO operations", 0) \
M(UInt64, max_io_thread_pool_free_size, 0, "Max free size for IO thread pool.", 0) \
M(UInt64, io_thread_pool_queue_size, 10000, "Queue size for IO thread pool.", 0) \
M(UInt64, max_active_parts_loading_thread_pool_size, 64, "The number of threads to load active set of data parts (Active ones) at startup.", 0) \
M(UInt64, max_outdated_parts_loading_thread_pool_size, 32, "The number of threads to load inactive set of data parts (Outdated ones) at startup.", 0) \
M(UInt64, max_unexpected_parts_loading_thread_pool_size, 8, "The number of threads to load inactive set of data parts (Unexpected ones) at startup.", 0) \
M(UInt64, max_parts_cleaning_thread_pool_size, 128, "The number of threads for concurrent removal of inactive data parts.", 0) \
M(UInt64, max_mutations_bandwidth_for_server, 0, "The maximum read speed of all mutations on server in bytes per second. Zero means unlimited.", 0) \
M(UInt64, max_merges_bandwidth_for_server, 0, "The maximum read speed of all merges on server in bytes per second. Zero means unlimited.", 0) \
M(UInt64, max_replicated_fetches_network_bandwidth_for_server, 0, "The maximum speed of data exchange over the network in bytes per second for replicated fetches. Zero means unlimited.", 0) \
M(UInt64, max_replicated_sends_network_bandwidth_for_server, 0, "The maximum speed of data exchange over the network in bytes per second for replicated sends. Zero means unlimited.", 0) \
M(UInt64, max_remote_read_network_bandwidth_for_server, 0, "The maximum speed of data exchange over the network in bytes per second for read. Zero means unlimited.", 0) \
M(UInt64, max_remote_write_network_bandwidth_for_server, 0, "The maximum speed of data exchange over the network in bytes per second for write. Zero means unlimited.", 0) \
M(UInt64, max_local_read_bandwidth_for_server, 0, "The maximum speed of local reads in bytes per second. Zero means unlimited.", 0) \
M(UInt64, max_local_write_bandwidth_for_server, 0, "The maximum speed of local writes in bytes per second. Zero means unlimited.", 0) \
M(UInt64, max_backups_io_thread_pool_size, 1000, "The maximum number of threads that would be used for IO operations for BACKUP queries", 0) \
M(UInt64, max_backups_io_thread_pool_free_size, 0, "Max free size for backups IO thread pool.", 0) \
M(UInt64, backups_io_thread_pool_queue_size, 0, "Queue size for backups IO thread pool.", 0) \
M(UInt64, backup_threads, 16, "The maximum number of threads to execute BACKUP requests.", 0) \
M(UInt64, max_backup_bandwidth_for_server, 0, "The maximum read speed in bytes per second for all backups on server. Zero means unlimited.", 0) \
M(UInt64, restore_threads, 16, "The maximum number of threads to execute RESTORE requests.", 0) \
M(Bool, shutdown_wait_backups_and_restores, true, "If set to true ClickHouse will wait for running backups and restores to finish before shutdown.", 0) \
M(Double, cannot_allocate_thread_fault_injection_probability, 0, "For testing purposes.", 0) \
M(Int32, max_connections, 1024, "Max server connections.", 0) \
M(UInt32, asynchronous_metrics_update_period_s, 1, "Period in seconds for updating asynchronous metrics.", 0) \
M(UInt32, asynchronous_heavy_metrics_update_period_s, 120, "Period in seconds for updating heavy asynchronous metrics.", 0) \
M(String, default_database, "default", "Default database name.", 0) \
M(String, tmp_policy, "", "Policy for storage with temporary data.", 0) \
M(UInt64, max_temporary_data_on_disk_size, 0, "The maximum amount of storage that could be used for external aggregation, joins or sorting.", 0) \
M(String, temporary_data_in_cache, "", "Cache disk name for temporary data.", 0) \
M(UInt64, aggregate_function_group_array_max_element_size, 0xFFFFFF, "Max array element size in bytes for groupArray function. This limit is checked at serialization and help to avoid large state size.", 0) \
M(GroupArrayActionWhenLimitReached, aggregate_function_group_array_action_when_limit_is_reached, GroupArrayActionWhenLimitReached::THROW, "Action to execute when max array element size is exceeded in groupArray: `throw` exception, or `discard` extra values", 0) \
M(UInt64, max_server_memory_usage, 0, "Maximum total memory usage of the server in bytes. Zero means unlimited.", 0) \
M(Double, max_server_memory_usage_to_ram_ratio, 0.9, "Same as max_server_memory_usage but in to RAM ratio. Allows to lower max memory on low-memory systems.", 0) \
M(UInt64, merges_mutations_memory_usage_soft_limit, 0, "Maximum total memory usage for merges and mutations in bytes. Zero means unlimited.", 0) \
M(Double, merges_mutations_memory_usage_to_ram_ratio, 0.5, "Same as merges_mutations_memory_usage_soft_limit but in to RAM ratio. Allows to lower memory limit on low-memory systems.", 0) \
M(Bool, allow_use_jemalloc_memory, true, "Allows to use jemalloc memory.", 0) \
M(UInt64, cgroups_memory_usage_observer_wait_time, 15, "Polling interval in seconds to read the current memory usage from cgroups. Zero means disabled.", 0) \
M(Double, cgroup_memory_watcher_hard_limit_ratio, 0.95, "Hard memory limit ratio for cgroup memory usage observer", 0) \
M(Double, cgroup_memory_watcher_soft_limit_ratio, 0.9, "Soft memory limit ratio limit for cgroup memory usage observer", 0) \
M(UInt64, async_insert_threads, 16, "Maximum number of threads to actually parse and insert data in background. Zero means asynchronous mode is disabled", 0) \
M(Bool, async_insert_queue_flush_on_shutdown, true, "If true queue of asynchronous inserts is flushed on graceful shutdown", 0) \
M(Bool, ignore_empty_sql_security_in_create_view_query, true, "If true, ClickHouse doesn't write defaults for empty SQL security statement in CREATE VIEW queries. This setting is only necessary for the migration period and will become obsolete in 24.4", 0) \
M(UInt64, max_build_vector_similarity_index_thread_pool_size, 16, "The maximum number of threads to use to build vector similarity indexes. 0 means all cores.", 0) \
#define LIST_OF_SERVER_SETTINGS(DECLARE, ALIAS) \
DECLARE(Bool, show_addresses_in_stack_traces, true, "If it is set true will show addresses in stack traces", 0) \
DECLARE(Bool, shutdown_wait_unfinished_queries, false, "If set true ClickHouse will wait for running queries finish before shutdown.", 0) \
DECLARE(UInt64, shutdown_wait_unfinished, 5, "Delay in seconds to wait for unfinished queries", 0) \
DECLARE(UInt64, max_thread_pool_size, 10000, "The maximum number of threads that could be allocated from the OS and used for query execution and background operations.", 0) \
DECLARE(UInt64, max_thread_pool_free_size, 1000, "The maximum number of threads that will always stay in a global thread pool once allocated and remain idle in case of insufficient number of tasks.", 0) \
DECLARE(UInt64, thread_pool_queue_size, 10000, "The maximum number of tasks that will be placed in a queue and wait for execution.", 0) \
DECLARE(UInt64, max_io_thread_pool_size, 100, "The maximum number of threads that would be used for IO operations", 0) \
DECLARE(UInt64, max_io_thread_pool_free_size, 0, "Max free size for IO thread pool.", 0) \
DECLARE(UInt64, io_thread_pool_queue_size, 10000, "Queue size for IO thread pool.", 0) \
DECLARE(UInt64, max_active_parts_loading_thread_pool_size, 64, "The number of threads to load active set of data parts (Active ones) at startup.", 0) \
DECLARE(UInt64, max_outdated_parts_loading_thread_pool_size, 32, "The number of threads to load inactive set of data parts (Outdated ones) at startup.", 0) \
DECLARE(UInt64, max_unexpected_parts_loading_thread_pool_size, 8, "The number of threads to load inactive set of data parts (Unexpected ones) at startup.", 0) \
DECLARE(UInt64, max_parts_cleaning_thread_pool_size, 128, "The number of threads for concurrent removal of inactive data parts.", 0) \
DECLARE(UInt64, max_mutations_bandwidth_for_server, 0, "The maximum read speed of all mutations on server in bytes per second. Zero means unlimited.", 0) \
DECLARE(UInt64, max_merges_bandwidth_for_server, 0, "The maximum read speed of all merges on server in bytes per second. Zero means unlimited.", 0) \
DECLARE(UInt64, max_replicated_fetches_network_bandwidth_for_server, 0, "The maximum speed of data exchange over the network in bytes per second for replicated fetches. Zero means unlimited.", 0) \
DECLARE(UInt64, max_replicated_sends_network_bandwidth_for_server, 0, "The maximum speed of data exchange over the network in bytes per second for replicated sends. Zero means unlimited.", 0) \
DECLARE(UInt64, max_remote_read_network_bandwidth_for_server, 0, "The maximum speed of data exchange over the network in bytes per second for read. Zero means unlimited.", 0) \
DECLARE(UInt64, max_remote_write_network_bandwidth_for_server, 0, "The maximum speed of data exchange over the network in bytes per second for write. Zero means unlimited.", 0) \
DECLARE(UInt64, max_local_read_bandwidth_for_server, 0, "The maximum speed of local reads in bytes per second. Zero means unlimited.", 0) \
DECLARE(UInt64, max_local_write_bandwidth_for_server, 0, "The maximum speed of local writes in bytes per second. Zero means unlimited.", 0) \
DECLARE(UInt64, max_backups_io_thread_pool_size, 1000, "The maximum number of threads that would be used for IO operations for BACKUP queries", 0) \
DECLARE(UInt64, max_backups_io_thread_pool_free_size, 0, "Max free size for backups IO thread pool.", 0) \
DECLARE(UInt64, backups_io_thread_pool_queue_size, 0, "Queue size for backups IO thread pool.", 0) \
DECLARE(UInt64, backup_threads, 16, "The maximum number of threads to execute BACKUP requests.", 0) \
DECLARE(UInt64, max_backup_bandwidth_for_server, 0, "The maximum read speed in bytes per second for all backups on server. Zero means unlimited.", 0) \
DECLARE(UInt64, restore_threads, 16, "The maximum number of threads to execute RESTORE requests.", 0) \
DECLARE(Bool, shutdown_wait_backups_and_restores, true, "If set to true ClickHouse will wait for running backups and restores to finish before shutdown.", 0) \
DECLARE(Double, cannot_allocate_thread_fault_injection_probability, 0, "For testing purposes.", 0) \
DECLARE(Int32, max_connections, 1024, "Max server connections.", 0) \
DECLARE(UInt32, asynchronous_metrics_update_period_s, 1, "Period in seconds for updating asynchronous metrics.", 0) \
DECLARE(UInt32, asynchronous_heavy_metrics_update_period_s, 120, "Period in seconds for updating heavy asynchronous metrics.", 0) \
DECLARE(String, default_database, "default", "Default database name.", 0) \
DECLARE(String, tmp_policy, "", "Policy for storage with temporary data.", 0) \
DECLARE(UInt64, max_temporary_data_on_disk_size, 0, "The maximum amount of storage that could be used for external aggregation, joins or sorting.", 0) \
DECLARE(String, temporary_data_in_cache, "", "Cache disk name for temporary data.", 0) \
DECLARE(UInt64, aggregate_function_group_array_max_element_size, 0xFFFFFF, "Max array element size in bytes for groupArray function. This limit is checked at serialization and help to avoid large state size.", 0) \
DECLARE(GroupArrayActionWhenLimitReached, aggregate_function_group_array_action_when_limit_is_reached, GroupArrayActionWhenLimitReached::THROW, "Action to execute when max array element size is exceeded in groupArray: `throw` exception, or `discard` extra values", 0) \
DECLARE(UInt64, max_server_memory_usage, 0, "Maximum total memory usage of the server in bytes. Zero means unlimited.", 0) \
DECLARE(Double, max_server_memory_usage_to_ram_ratio, 0.9, "Same as max_server_memory_usage but in to RAM ratio. Allows to lower max memory on low-memory systems.", 0) \
DECLARE(UInt64, merges_mutations_memory_usage_soft_limit, 0, "Maximum total memory usage for merges and mutations in bytes. Zero means unlimited.", 0) \
DECLARE(Double, merges_mutations_memory_usage_to_ram_ratio, 0.5, "Same as merges_mutations_memory_usage_soft_limit but in to RAM ratio. Allows to lower memory limit on low-memory systems.", 0) \
DECLARE(Bool, allow_use_jemalloc_memory, true, "Allows to use jemalloc memory.", 0) \
DECLARE(UInt64, cgroups_memory_usage_observer_wait_time, 15, "Polling interval in seconds to read the current memory usage from cgroups. Zero means disabled.", 0) \
DECLARE(Double, cgroup_memory_watcher_hard_limit_ratio, 0.95, "Hard memory limit ratio for cgroup memory usage observer", 0) \
DECLARE(Double, cgroup_memory_watcher_soft_limit_ratio, 0.9, "Soft memory limit ratio limit for cgroup memory usage observer", 0) \
DECLARE(UInt64, async_insert_threads, 16, "Maximum number of threads to actually parse and insert data in background. Zero means asynchronous mode is disabled", 0) \
DECLARE(Bool, async_insert_queue_flush_on_shutdown, true, "If true queue of asynchronous inserts is flushed on graceful shutdown", 0) \
DECLARE(Bool, ignore_empty_sql_security_in_create_view_query, true, "If true, ClickHouse doesn't write defaults for empty SQL security statement in CREATE VIEW queries. This setting is only necessary for the migration period and will become obsolete in 24.4", 0) \
DECLARE(UInt64, max_build_vector_similarity_index_thread_pool_size, 16, "The maximum number of threads to use to build vector similarity indexes. 0 means all cores.", 0) \
\
/* Database Catalog */ \
M(UInt64, database_atomic_delay_before_drop_table_sec, 8 * 60, "The delay during which a dropped table can be restored using the UNDROP statement. If DROP TABLE ran with a SYNC modifier, the setting is ignored.", 0) \
M(UInt64, database_catalog_unused_dir_hide_timeout_sec, 60 * 60, "Parameter of a task that cleans up garbage from store/ directory. If some subdirectory is not used by clickhouse-server and this directory was not modified for last database_catalog_unused_dir_hide_timeout_sec seconds, the task will 'hide' this directory by removing all access rights. It also works for directories that clickhouse-server does not expect to see inside store/. Zero means 'immediately'.", 0) \
M(UInt64, database_catalog_unused_dir_rm_timeout_sec, 30 * 24 * 60 * 60, "Parameter of a task that cleans up garbage from store/ directory. If some subdirectory is not used by clickhouse-server and it was previously 'hidden' (see database_catalog_unused_dir_hide_timeout_sec) and this directory was not modified for last database_catalog_unused_dir_rm_timeout_sec seconds, the task will remove this directory. It also works for directories that clickhouse-server does not expect to see inside store/. Zero means 'never'.", 0) \
M(UInt64, database_catalog_unused_dir_cleanup_period_sec, 24 * 60 * 60, "Parameter of a task that cleans up garbage from store/ directory. Sets scheduling period of the task. Zero means 'never'.", 0) \
M(UInt64, database_catalog_drop_error_cooldown_sec, 5, "In case if drop table failed, ClickHouse will wait for this timeout before retrying the operation.", 0) \
M(UInt64, database_catalog_drop_table_concurrency, 16, "The size of the threadpool used for dropping tables.", 0) \
DECLARE(UInt64, database_atomic_delay_before_drop_table_sec, 8 * 60, "The delay during which a dropped table can be restored using the UNDROP statement. If DROP TABLE ran with a SYNC modifier, the setting is ignored.", 0) \
DECLARE(UInt64, database_catalog_unused_dir_hide_timeout_sec, 60 * 60, "Parameter of a task that cleans up garbage from store/ directory. If some subdirectory is not used by clickhouse-server and this directory was not modified for last database_catalog_unused_dir_hide_timeout_sec seconds, the task will 'hide' this directory by removing all access rights. It also works for directories that clickhouse-server does not expect to see inside store/. Zero means 'immediately'.", 0) \
DECLARE(UInt64, database_catalog_unused_dir_rm_timeout_sec, 30 * 24 * 60 * 60, "Parameter of a task that cleans up garbage from store/ directory. If some subdirectory is not used by clickhouse-server and it was previously 'hidden' (see database_catalog_unused_dir_hide_timeout_sec) and this directory was not modified for last database_catalog_unused_dir_rm_timeout_sec seconds, the task will remove this directory. It also works for directories that clickhouse-server does not expect to see inside store/. Zero means 'never'.", 0) \
DECLARE(UInt64, database_catalog_unused_dir_cleanup_period_sec, 24 * 60 * 60, "Parameter of a task that cleans up garbage from store/ directory. Sets scheduling period of the task. Zero means 'never'.", 0) \
DECLARE(UInt64, database_catalog_drop_error_cooldown_sec, 5, "In case if drop table failed, ClickHouse will wait for this timeout before retrying the operation.", 0) \
DECLARE(UInt64, database_catalog_drop_table_concurrency, 16, "The size of the threadpool used for dropping tables.", 0) \
\
\
M(UInt64, max_concurrent_queries, 0, "Maximum number of concurrently executed queries. Zero means unlimited.", 0) \
M(UInt64, max_concurrent_insert_queries, 0, "Maximum number of concurrently INSERT queries. Zero means unlimited.", 0) \
M(UInt64, max_concurrent_select_queries, 0, "Maximum number of concurrently SELECT queries. Zero means unlimited.", 0) \
M(UInt64, max_waiting_queries, 0, "Maximum number of concurrently waiting queries blocked due to `async_load_databases`. Note that waiting queries are not considered by `max_concurrent_*queries*` limits. Zero means unlimited.", 0) \
DECLARE(UInt64, max_concurrent_queries, 0, "Maximum number of concurrently executed queries. Zero means unlimited.", 0) \
DECLARE(UInt64, max_concurrent_insert_queries, 0, "Maximum number of concurrently INSERT queries. Zero means unlimited.", 0) \
DECLARE(UInt64, max_concurrent_select_queries, 0, "Maximum number of concurrently SELECT queries. Zero means unlimited.", 0) \
DECLARE(UInt64, max_waiting_queries, 0, "Maximum number of concurrently waiting queries blocked due to `async_load_databases`. Note that waiting queries are not considered by `max_concurrent_*queries*` limits. Zero means unlimited.", 0) \
\
M(Double, cache_size_to_ram_max_ratio, 0.5, "Set cache size to RAM max ratio. Allows to lower cache size on low-memory systems.", 0) \
M(String, uncompressed_cache_policy, DEFAULT_UNCOMPRESSED_CACHE_POLICY, "Uncompressed cache policy name.", 0) \
M(UInt64, uncompressed_cache_size, DEFAULT_UNCOMPRESSED_CACHE_MAX_SIZE, "Size of cache for uncompressed blocks. Zero means disabled.", 0) \
M(Double, uncompressed_cache_size_ratio, DEFAULT_UNCOMPRESSED_CACHE_SIZE_RATIO, "The size of the protected queue in the uncompressed cache relative to the cache's total size.", 0) \
M(String, mark_cache_policy, DEFAULT_MARK_CACHE_POLICY, "Mark cache policy name.", 0) \
M(UInt64, mark_cache_size, DEFAULT_MARK_CACHE_MAX_SIZE, "Size of cache for marks (index of MergeTree family of tables).", 0) \
M(Double, mark_cache_size_ratio, DEFAULT_MARK_CACHE_SIZE_RATIO, "The size of the protected queue in the mark cache relative to the cache's total size.", 0) \
M(String, index_uncompressed_cache_policy, DEFAULT_INDEX_UNCOMPRESSED_CACHE_POLICY, "Secondary index uncompressed cache policy name.", 0) \
M(UInt64, index_uncompressed_cache_size, DEFAULT_INDEX_UNCOMPRESSED_CACHE_MAX_SIZE, "Size of cache for uncompressed blocks of secondary indices. Zero means disabled.", 0) \
M(Double, index_uncompressed_cache_size_ratio, DEFAULT_INDEX_UNCOMPRESSED_CACHE_SIZE_RATIO, "The size of the protected queue in the secondary index uncompressed cache relative to the cache's total size.", 0) \
M(String, index_mark_cache_policy, DEFAULT_INDEX_MARK_CACHE_POLICY, "Secondary index mark cache policy name.", 0) \
M(UInt64, index_mark_cache_size, DEFAULT_INDEX_MARK_CACHE_MAX_SIZE, "Size of cache for secondary index marks. Zero means disabled.", 0) \
M(Double, index_mark_cache_size_ratio, DEFAULT_INDEX_MARK_CACHE_SIZE_RATIO, "The size of the protected queue in the secondary index mark cache relative to the cache's total size.", 0) \
M(UInt64, page_cache_chunk_size, 2 << 20, "Bytes per chunk in userspace page cache. Rounded up to a multiple of page size (typically 4 KiB) or huge page size (typically 2 MiB, only if page_cache_use_thp is enabled).", 0) \
M(UInt64, page_cache_mmap_size, 1 << 30, "Bytes per memory mapping in userspace page cache. Not important.", 0) \
M(UInt64, page_cache_size, 0, "Amount of virtual memory to map for userspace page cache. If page_cache_use_madv_free is enabled, it's recommended to set this higher than the machine's RAM size. Use 0 to disable userspace page cache.", 0) \
M(Bool, page_cache_use_madv_free, DBMS_DEFAULT_PAGE_CACHE_USE_MADV_FREE, "If true, the userspace page cache will allow the OS to automatically reclaim memory from the cache on memory pressure (using MADV_FREE).", 0) \
M(Bool, page_cache_use_transparent_huge_pages, true, "Userspace will attempt to use transparent huge pages on Linux. This is best-effort.", 0) \
M(UInt64, mmap_cache_size, DEFAULT_MMAP_CACHE_MAX_SIZE, "A cache for mmapped files.", 0) \
M(UInt64, compiled_expression_cache_size, DEFAULT_COMPILED_EXPRESSION_CACHE_MAX_SIZE, "Byte size of compiled expressions cache.", 0) \
M(UInt64, compiled_expression_cache_elements_size, DEFAULT_COMPILED_EXPRESSION_CACHE_MAX_ENTRIES, "Maximum entries in compiled expressions cache.", 0) \
DECLARE(Double, cache_size_to_ram_max_ratio, 0.5, "Set cache size to RAM max ratio. Allows to lower cache size on low-memory systems.", 0) \
DECLARE(String, uncompressed_cache_policy, DEFAULT_UNCOMPRESSED_CACHE_POLICY, "Uncompressed cache policy name.", 0) \
DECLARE(UInt64, uncompressed_cache_size, DEFAULT_UNCOMPRESSED_CACHE_MAX_SIZE, "Size of cache for uncompressed blocks. Zero means disabled.", 0) \
DECLARE(Double, uncompressed_cache_size_ratio, DEFAULT_UNCOMPRESSED_CACHE_SIZE_RATIO, "The size of the protected queue in the uncompressed cache relative to the cache's total size.", 0) \
DECLARE(String, mark_cache_policy, DEFAULT_MARK_CACHE_POLICY, "Mark cache policy name.", 0) \
DECLARE(UInt64, mark_cache_size, DEFAULT_MARK_CACHE_MAX_SIZE, "Size of cache for marks (index of MergeTree family of tables).", 0) \
DECLARE(Double, mark_cache_size_ratio, DEFAULT_MARK_CACHE_SIZE_RATIO, "The size of the protected queue in the mark cache relative to the cache's total size.", 0) \
DECLARE(String, index_uncompressed_cache_policy, DEFAULT_INDEX_UNCOMPRESSED_CACHE_POLICY, "Secondary index uncompressed cache policy name.", 0) \
DECLARE(UInt64, index_uncompressed_cache_size, DEFAULT_INDEX_UNCOMPRESSED_CACHE_MAX_SIZE, "Size of cache for uncompressed blocks of secondary indices. Zero means disabled.", 0) \
DECLARE(Double, index_uncompressed_cache_size_ratio, DEFAULT_INDEX_UNCOMPRESSED_CACHE_SIZE_RATIO, "The size of the protected queue in the secondary index uncompressed cache relative to the cache's total size.", 0) \
DECLARE(String, index_mark_cache_policy, DEFAULT_INDEX_MARK_CACHE_POLICY, "Secondary index mark cache policy name.", 0) \
DECLARE(UInt64, index_mark_cache_size, DEFAULT_INDEX_MARK_CACHE_MAX_SIZE, "Size of cache for secondary index marks. Zero means disabled.", 0) \
DECLARE(Double, index_mark_cache_size_ratio, DEFAULT_INDEX_MARK_CACHE_SIZE_RATIO, "The size of the protected queue in the secondary index mark cache relative to the cache's total size.", 0) \
DECLARE(UInt64, page_cache_chunk_size, 2 << 20, "Bytes per chunk in userspace page cache. Rounded up to a multiple of page size (typically 4 KiB) or huge page size (typically 2 MiB, only if page_cache_use_thp is enabled).", 0) \
DECLARE(UInt64, page_cache_mmap_size, 1 << 30, "Bytes per memory mapping in userspace page cache. Not important.", 0) \
DECLARE(UInt64, page_cache_size, 0, "Amount of virtual memory to map for userspace page cache. If page_cache_use_madv_free is enabled, it's recommended to set this higher than the machine's RAM size. Use 0 to disable userspace page cache.", 0) \
DECLARE(Bool, page_cache_use_madv_free, DBMS_DEFAULT_PAGE_CACHE_USE_MADV_FREE, "If true, the userspace page cache will allow the OS to automatically reclaim memory from the cache on memory pressure (using MADV_FREE).", 0) \
DECLARE(Bool, page_cache_use_transparent_huge_pages, true, "Userspace will attempt to use transparent huge pages on Linux. This is best-effort.", 0) \
DECLARE(UInt64, mmap_cache_size, DEFAULT_MMAP_CACHE_MAX_SIZE, "A cache for mmapped files.", 0) \
DECLARE(UInt64, compiled_expression_cache_size, DEFAULT_COMPILED_EXPRESSION_CACHE_MAX_SIZE, "Byte size of compiled expressions cache.", 0) \
DECLARE(UInt64, compiled_expression_cache_elements_size, DEFAULT_COMPILED_EXPRESSION_CACHE_MAX_ENTRIES, "Maximum entries in compiled expressions cache.", 0) \
\
M(Bool, disable_internal_dns_cache, false, "Disable internal DNS caching at all.", 0) \
M(UInt64, dns_cache_max_entries, 10000, "Internal DNS cache max entries.", 0) \
M(Int32, dns_cache_update_period, 15, "Internal DNS cache update period in seconds.", 0) \
M(UInt32, dns_max_consecutive_failures, 10, "Max DNS resolve failures of a hostname before dropping the hostname from ClickHouse DNS cache.", 0) \
M(Bool, dns_allow_resolve_names_to_ipv4, true, "Allows resolve names to ipv4 addresses.", 0) \
M(Bool, dns_allow_resolve_names_to_ipv6, true, "Allows resolve names to ipv6 addresses.", 0) \
DECLARE(Bool, disable_internal_dns_cache, false, "Disable internal DNS caching at all.", 0) \
DECLARE(UInt64, dns_cache_max_entries, 10000, "Internal DNS cache max entries.", 0) \
DECLARE(Int32, dns_cache_update_period, 15, "Internal DNS cache update period in seconds.", 0) \
DECLARE(UInt32, dns_max_consecutive_failures, 10, "Max DNS resolve failures of a hostname before dropping the hostname from ClickHouse DNS cache.", 0) \
DECLARE(Bool, dns_allow_resolve_names_to_ipv4, true, "Allows resolve names to ipv4 addresses.", 0) \
DECLARE(Bool, dns_allow_resolve_names_to_ipv6, true, "Allows resolve names to ipv6 addresses.", 0) \
\
M(UInt64, max_table_size_to_drop, 50000000000lu, "If size of a table is greater than this value (in bytes) than table could not be dropped with any DROP query.", 0) \
M(UInt64, max_partition_size_to_drop, 50000000000lu, "Same as max_table_size_to_drop, but for the partitions.", 0) \
M(UInt64, max_table_num_to_warn, 5000lu, "If the number of tables is greater than this value, the server will create a warning that will displayed to user.", 0) \
M(UInt64, max_view_num_to_warn, 10000lu, "If the number of views is greater than this value, the server will create a warning that will displayed to user.", 0) \
M(UInt64, max_dictionary_num_to_warn, 1000lu, "If the number of dictionaries is greater than this value, the server will create a warning that will displayed to user.", 0) \
M(UInt64, max_database_num_to_warn, 1000lu, "If the number of databases is greater than this value, the server will create a warning that will displayed to user.", 0) \
M(UInt64, max_part_num_to_warn, 100000lu, "If the number of parts is greater than this value, the server will create a warning that will displayed to user.", 0) \
M(UInt64, max_table_num_to_throw, 0lu, "If number of tables is greater than this value, server will throw an exception. 0 means no limitation. View, remote tables, dictionary, system tables are not counted. Only count table in Atomic/Ordinary/Replicated/Lazy database engine.", 0) \
M(UInt64, max_database_num_to_throw, 0lu, "If number of databases is greater than this value, server will throw an exception. 0 means no limitation.", 0) \
M(UInt64, max_authentication_methods_per_user, 100, "The maximum number of authentication methods a user can be created with or altered. Changing this setting does not affect existing users. Zero means unlimited", 0) \
M(UInt64, concurrent_threads_soft_limit_num, 0, "Sets how many concurrent thread can be allocated before applying CPU pressure. Zero means unlimited.", 0) \
M(UInt64, concurrent_threads_soft_limit_ratio_to_cores, 0, "Same as concurrent_threads_soft_limit_num, but with ratio to cores.", 0) \
DECLARE(UInt64, max_table_size_to_drop, 50000000000lu, "If size of a table is greater than this value (in bytes) than table could not be dropped with any DROP query.", 0) \
DECLARE(UInt64, max_partition_size_to_drop, 50000000000lu, "Same as max_table_size_to_drop, but for the partitions.", 0) \
DECLARE(UInt64, max_table_num_to_warn, 5000lu, "If the number of tables is greater than this value, the server will create a warning that will displayed to user.", 0) \
DECLARE(UInt64, max_view_num_to_warn, 10000lu, "If the number of views is greater than this value, the server will create a warning that will displayed to user.", 0) \
DECLARE(UInt64, max_dictionary_num_to_warn, 1000lu, "If the number of dictionaries is greater than this value, the server will create a warning that will displayed to user.", 0) \
DECLARE(UInt64, max_database_num_to_warn, 1000lu, "If the number of databases is greater than this value, the server will create a warning that will displayed to user.", 0) \
DECLARE(UInt64, max_part_num_to_warn, 100000lu, "If the number of parts is greater than this value, the server will create a warning that will displayed to user.", 0) \
DECLARE(UInt64, max_table_num_to_throw, 0lu, "If number of tables is greater than this value, server will throw an exception. 0 means no limitation. View, remote tables, dictionary, system tables are not counted. Only count table in Atomic/Ordinary/Replicated/Lazy database engine.", 0) \
DECLARE(UInt64, max_database_num_to_throw, 0lu, "If number of databases is greater than this value, server will throw an exception. 0 means no limitation.", 0) \
DECLARE(UInt64, max_authentication_methods_per_user, 100, "The maximum number of authentication methods a user can be created with or altered. Changing this setting does not affect existing users. Zero means unlimited", 0) \
DECLARE(UInt64, concurrent_threads_soft_limit_num, 0, "Sets how many concurrent thread can be allocated before applying CPU pressure. Zero means unlimited.", 0) \
DECLARE(UInt64, concurrent_threads_soft_limit_ratio_to_cores, 0, "Same as concurrent_threads_soft_limit_num, but with ratio to cores.", 0) \
\
M(UInt64, background_pool_size, 16, "The maximum number of threads what will be used for merging or mutating data parts for *MergeTree-engine tables in a background.", 0) \
M(Float, background_merges_mutations_concurrency_ratio, 2, "The number of part mutation tasks that can be executed concurrently by each thread in background pool.", 0) \
M(String, background_merges_mutations_scheduling_policy, "round_robin", "The policy on how to perform a scheduling for background merges and mutations. Possible values are: `round_robin` and `shortest_task_first`. ", 0) \
M(UInt64, background_move_pool_size, 8, "The maximum number of threads that will be used for moving data parts to another disk or volume for *MergeTree-engine tables in a background.", 0) \
M(UInt64, background_fetches_pool_size, 16, "The maximum number of threads that will be used for fetching data parts from another replica for *MergeTree-engine tables in a background.", 0) \
M(UInt64, background_common_pool_size, 8, "The maximum number of threads that will be used for performing a variety of operations (mostly garbage collection) for *MergeTree-engine tables in a background.", 0) \
M(UInt64, background_buffer_flush_schedule_pool_size, 16, "The maximum number of threads that will be used for performing flush operations for Buffer-engine tables in a background.", 0) \
M(UInt64, background_schedule_pool_size, 512, "The maximum number of threads that will be used for constantly executing some lightweight periodic operations.", 0) \
M(UInt64, background_message_broker_schedule_pool_size, 16, "The maximum number of threads that will be used for executing background operations for message streaming.", 0) \
M(UInt64, background_distributed_schedule_pool_size, 16, "The maximum number of threads that will be used for executing distributed sends.", 0) \
M(UInt64, tables_loader_foreground_pool_size, 0, "The maximum number of threads that will be used for foreground (that is being waited for by a query) loading of tables. Also used for synchronous loading of tables before the server start. Zero means use all CPUs.", 0) \
M(UInt64, tables_loader_background_pool_size, 0, "The maximum number of threads that will be used for background async loading of tables. Zero means use all CPUs.", 0) \
M(Bool, async_load_databases, false, "Enable asynchronous loading of databases and tables to speedup server startup. Queries to not yet loaded entity will be blocked until load is finished.", 0) \
M(Bool, display_secrets_in_show_and_select, false, "Allow showing secrets in SHOW and SELECT queries via a format setting and a grant", 0) \
M(Seconds, keep_alive_timeout, DEFAULT_HTTP_KEEP_ALIVE_TIMEOUT, "The number of seconds that ClickHouse waits for incoming requests before closing the connection.", 0) \
M(UInt64, max_keep_alive_requests, 10000, "The maximum number of requests handled via a single http keepalive connection before the server closes this connection.", 0) \
M(Seconds, replicated_fetches_http_connection_timeout, 0, "HTTP connection timeout for part fetch requests. Inherited from default profile `http_connection_timeout` if not set explicitly.", 0) \
M(Seconds, replicated_fetches_http_send_timeout, 0, "HTTP send timeout for part fetch requests. Inherited from default profile `http_send_timeout` if not set explicitly.", 0) \
M(Seconds, replicated_fetches_http_receive_timeout, 0, "HTTP receive timeout for fetch part requests. Inherited from default profile `http_receive_timeout` if not set explicitly.", 0) \
M(UInt64, total_memory_profiler_step, 0, "Whenever server memory usage becomes larger than every next step in number of bytes the memory profiler will collect the allocating stack trace. Zero means disabled memory profiler. Values lower than a few megabytes will slow down server.", 0) \
M(Double, total_memory_tracker_sample_probability, 0, "Collect random allocations and deallocations and write them into system.trace_log with 'MemorySample' trace_type. The probability is for every alloc/free regardless to the size of the allocation (can be changed with `memory_profiler_sample_min_allocation_size` and `memory_profiler_sample_max_allocation_size`). Note that sampling happens only when the amount of untracked memory exceeds 'max_untracked_memory'. You may want to set 'max_untracked_memory' to 0 for extra fine grained sampling.", 0) \
M(UInt64, total_memory_profiler_sample_min_allocation_size, 0, "Collect random allocations of size greater or equal than specified value with probability equal to `total_memory_profiler_sample_probability`. 0 means disabled. You may want to set 'max_untracked_memory' to 0 to make this threshold to work as expected.", 0) \
M(UInt64, total_memory_profiler_sample_max_allocation_size, 0, "Collect random allocations of size less or equal than specified value with probability equal to `total_memory_profiler_sample_probability`. 0 means disabled. You may want to set 'max_untracked_memory' to 0 to make this threshold to work as expected.", 0) \
M(Bool, validate_tcp_client_information, false, "Validate client_information in the query packet over the native TCP protocol.", 0) \
M(Bool, storage_metadata_write_full_object_key, false, "Write disk metadata files with VERSION_FULL_OBJECT_KEY format", 0) \
M(UInt64, max_materialized_views_count_for_table, 0, "A limit on the number of materialized views attached to a table.", 0) \
M(UInt32, max_database_replicated_create_table_thread_pool_size, 1, "The number of threads to create tables during replica recovery in DatabaseReplicated. Zero means number of threads equal number of cores.", 0) \
M(Bool, database_replicated_allow_detach_permanently, true, "Allow detaching tables permanently in Replicated databases", 0) \
M(Bool, format_alter_operations_with_parentheses, false, "If enabled, each operation in alter queries will be surrounded with parentheses in formatted queries to make them less ambiguous.", 0) \
M(String, default_replica_path, "/clickhouse/tables/{uuid}/{shard}", "The path to the table in ZooKeeper", 0) \
M(String, default_replica_name, "{replica}", "The replica name in ZooKeeper", 0) \
M(UInt64, disk_connections_soft_limit, 5000, "Connections above this limit have significantly shorter time to live. The limit applies to the disks connections.", 0) \
M(UInt64, disk_connections_warn_limit, 10000, "Warning massages are written to the logs if number of in-use connections are higher than this limit. The limit applies to the disks connections.", 0) \
M(UInt64, disk_connections_store_limit, 30000, "Connections above this limit reset after use. Set to 0 to turn connection cache off. The limit applies to the disks connections.", 0) \
M(UInt64, storage_connections_soft_limit, 100, "Connections above this limit have significantly shorter time to live. The limit applies to the storages connections.", 0) \
M(UInt64, storage_connections_warn_limit, 1000, "Warning massages are written to the logs if number of in-use connections are higher than this limit. The limit applies to the storages connections.", 0) \
M(UInt64, storage_connections_store_limit, 5000, "Connections above this limit reset after use. Set to 0 to turn connection cache off. The limit applies to the storages connections.", 0) \
M(UInt64, http_connections_soft_limit, 100, "Connections above this limit have significantly shorter time to live. The limit applies to the http connections which do not belong to any disk or storage.", 0) \
M(UInt64, http_connections_warn_limit, 1000, "Warning massages are written to the logs if number of in-use connections are higher than this limit. The limit applies to the http connections which do not belong to any disk or storage.", 0) \
M(UInt64, http_connections_store_limit, 5000, "Connections above this limit reset after use. Set to 0 to turn connection cache off. The limit applies to the http connections which do not belong to any disk or storage.", 0) \
M(UInt64, global_profiler_real_time_period_ns, 0, "Period for real clock timer of global profiler (in nanoseconds). Set 0 value to turn off the real clock global profiler. Recommended value is at least 10000000 (100 times a second) for single queries or 1000000000 (once a second) for cluster-wide profiling.", 0) \
M(UInt64, global_profiler_cpu_time_period_ns, 0, "Period for CPU clock timer of global profiler (in nanoseconds). Set 0 value to turn off the CPU clock global profiler. Recommended value is at least 10000000 (100 times a second) for single queries or 1000000000 (once a second) for cluster-wide profiling.", 0) \
M(Bool, enable_azure_sdk_logging, false, "Enables logging from Azure sdk", 0) \
M(UInt64, max_entries_for_hash_table_stats, 10'000, "How many entries hash table statistics collected during aggregation is allowed to have", 0) \
M(String, merge_workload, "default", "Name of workload to be used to access resources for all merges (may be overridden by a merge tree setting)", 0) \
M(String, mutation_workload, "default", "Name of workload to be used to access resources for all mutations (may be overridden by a merge tree setting)", 0) \
M(Bool, prepare_system_log_tables_on_startup, false, "If true, ClickHouse creates all configured `system.*_log` tables before the startup. It can be helpful if some startup scripts depend on these tables.", 0) \
M(Double, gwp_asan_force_sample_probability, 0.0003, "Probability that an allocation from specific places will be sampled by GWP Asan (i.e. PODArray allocations)", 0) \
M(UInt64, config_reload_interval_ms, 2000, "How often clickhouse will reload config and check for new changes", 0) \
M(UInt64, memory_worker_period_ms, 0, "Tick period of background memory worker which corrects memory tracker memory usages and cleans up unused pages during higher memory usage. If set to 0, default value will be used depending on the memory usage source", 0) \
M(Bool, disable_insertion_and_mutation, false, "Disable all insert/alter/delete queries. This setting will be enabled if someone needs read-only nodes to prevent insertion and mutation affect reading performance.", 0) \
M(UInt64, parts_kill_delay_period, 30, "Period to completely remove parts for SharedMergeTree. Only available in ClickHouse Cloud", 0) \
M(UInt64, parts_kill_delay_period_random_add, 10, "Add uniformly distributed value from 0 to x seconds to kill_delay_period to avoid thundering herd effect and subsequent DoS of ZooKeeper in case of very large number of tables. Only available in ClickHouse Cloud", 0) \
M(UInt64, parts_killer_pool_size, 128, "Threads for cleanup of shared merge tree outdated threads. Only available in ClickHouse Cloud", 0) \
M(UInt64, keeper_multiread_batch_size, 10'000, "Maximum size of batch for MultiRead request to [Zoo]Keeper that support batching. If set to 0, batching is disabled. Available only in ClickHouse Cloud.", 0) \
M(Bool, use_legacy_mongodb_integration, true, "Use the legacy MongoDB integration implementation. Note: it's highly recommended to set this option to false, since legacy implementation will be removed in the future. Please submit any issues you encounter with the new implementation.", 0) \
DECLARE(UInt64, background_pool_size, 16, "The maximum number of threads what will be used for merging or mutating data parts for *MergeTree-engine tables in a background.", 0) \
DECLARE(Float, background_merges_mutations_concurrency_ratio, 2, "The number of part mutation tasks that can be executed concurrently by each thread in background pool.", 0) \
DECLARE(String, background_merges_mutations_scheduling_policy, "round_robin", "The policy on how to perform a scheduling for background merges and mutations. Possible values are: `round_robin` and `shortest_task_first`. ", 0) \
DECLARE(UInt64, background_move_pool_size, 8, "The maximum number of threads that will be used for moving data parts to another disk or volume for *MergeTree-engine tables in a background.", 0) \
DECLARE(UInt64, background_fetches_pool_size, 16, "The maximum number of threads that will be used for fetching data parts from another replica for *MergeTree-engine tables in a background.", 0) \
DECLARE(UInt64, background_common_pool_size, 8, "The maximum number of threads that will be used for performing a variety of operations (mostly garbage collection) for *MergeTree-engine tables in a background.", 0) \
DECLARE(UInt64, background_buffer_flush_schedule_pool_size, 16, "The maximum number of threads that will be used for performing flush operations for Buffer-engine tables in a background.", 0) \
DECLARE(UInt64, background_schedule_pool_size, 512, "The maximum number of threads that will be used for constantly executing some lightweight periodic operations.", 0) \
DECLARE(UInt64, background_message_broker_schedule_pool_size, 16, "The maximum number of threads that will be used for executing background operations for message streaming.", 0) \
DECLARE(UInt64, background_distributed_schedule_pool_size, 16, "The maximum number of threads that will be used for executing distributed sends.", 0) \
DECLARE(UInt64, tables_loader_foreground_pool_size, 0, "The maximum number of threads that will be used for foreground (that is being waited for by a query) loading of tables. Also used for synchronous loading of tables before the server start. Zero means use all CPUs.", 0) \
DECLARE(UInt64, tables_loader_background_pool_size, 0, "The maximum number of threads that will be used for background async loading of tables. Zero means use all CPUs.", 0) \
DECLARE(Bool, async_load_databases, false, "Enable asynchronous loading of databases and tables to speedup server startup. Queries to not yet loaded entity will be blocked until load is finished.", 0) \
DECLARE(Bool, display_secrets_in_show_and_select, false, "Allow showing secrets in SHOW and SELECT queries via a format setting and a grant", 0) \
DECLARE(Seconds, keep_alive_timeout, DEFAULT_HTTP_KEEP_ALIVE_TIMEOUT, "The number of seconds that ClickHouse waits for incoming requests before closing the connection.", 0) \
DECLARE(UInt64, max_keep_alive_requests, 10000, "The maximum number of requests handled via a single http keepalive connection before the server closes this connection.", 0) \
DECLARE(Seconds, replicated_fetches_http_connection_timeout, 0, "HTTP connection timeout for part fetch requests. Inherited from default profile `http_connection_timeout` if not set explicitly.", 0) \
DECLARE(Seconds, replicated_fetches_http_send_timeout, 0, "HTTP send timeout for part fetch requests. Inherited from default profile `http_send_timeout` if not set explicitly.", 0) \
DECLARE(Seconds, replicated_fetches_http_receive_timeout, 0, "HTTP receive timeout for fetch part requests. Inherited from default profile `http_receive_timeout` if not set explicitly.", 0) \
DECLARE(UInt64, total_memory_profiler_step, 0, "Whenever server memory usage becomes larger than every next step in number of bytes the memory profiler will collect the allocating stack trace. Zero means disabled memory profiler. Values lower than a few megabytes will slow down server.", 0) \
DECLARE(Double, total_memory_tracker_sample_probability, 0, "Collect random allocations and deallocations and write them into system.trace_log with 'MemorySample' trace_type. The probability is for every alloc/free regardless to the size of the allocation (can be changed with `memory_profiler_sample_min_allocation_size` and `memory_profiler_sample_max_allocation_size`). Note that sampling happens only when the amount of untracked memory exceeds 'max_untracked_memory'. You may want to set 'max_untracked_memory' to 0 for extra fine grained sampling.", 0) \
DECLARE(UInt64, total_memory_profiler_sample_min_allocation_size, 0, "Collect random allocations of size greater or equal than specified value with probability equal to `total_memory_profiler_sample_probability`. 0 means disabled. You may want to set 'max_untracked_memory' to 0 to make this threshold to work as expected.", 0) \
DECLARE(UInt64, total_memory_profiler_sample_max_allocation_size, 0, "Collect random allocations of size less or equal than specified value with probability equal to `total_memory_profiler_sample_probability`. 0 means disabled. You may want to set 'max_untracked_memory' to 0 to make this threshold to work as expected.", 0) \
DECLARE(Bool, validate_tcp_client_information, false, "Validate client_information in the query packet over the native TCP protocol.", 0) \
DECLARE(Bool, storage_metadata_write_full_object_key, false, "Write disk metadata files with VERSION_FULL_OBJECT_KEY format", 0) \
DECLARE(UInt64, max_materialized_views_count_for_table, 0, "A limit on the number of materialized views attached to a table.", 0) \
DECLARE(UInt32, max_database_replicated_create_table_thread_pool_size, 1, "The number of threads to create tables during replica recovery in DatabaseReplicated. Zero means number of threads equal number of cores.", 0) \
DECLARE(Bool, database_replicated_allow_detach_permanently, true, "Allow detaching tables permanently in Replicated databases", 0) \
DECLARE(Bool, format_alter_operations_with_parentheses, false, "If enabled, each operation in alter queries will be surrounded with parentheses in formatted queries to make them less ambiguous.", 0) \
DECLARE(String, default_replica_path, "/clickhouse/tables/{uuid}/{shard}", "The path to the table in ZooKeeper", 0) \
DECLARE(String, default_replica_name, "{replica}", "The replica name in ZooKeeper", 0) \
DECLARE(UInt64, disk_connections_soft_limit, 5000, "Connections above this limit have significantly shorter time to live. The limit applies to the disks connections.", 0) \
DECLARE(UInt64, disk_connections_warn_limit, 10000, "Warning massages are written to the logs if number of in-use connections are higher than this limit. The limit applies to the disks connections.", 0) \
DECLARE(UInt64, disk_connections_store_limit, 30000, "Connections above this limit reset after use. Set to 0 to turn connection cache off. The limit applies to the disks connections.", 0) \
DECLARE(UInt64, storage_connections_soft_limit, 100, "Connections above this limit have significantly shorter time to live. The limit applies to the storages connections.", 0) \
DECLARE(UInt64, storage_connections_warn_limit, 1000, "Warning massages are written to the logs if number of in-use connections are higher than this limit. The limit applies to the storages connections.", 0) \
DECLARE(UInt64, storage_connections_store_limit, 5000, "Connections above this limit reset after use. Set to 0 to turn connection cache off. The limit applies to the storages connections.", 0) \
DECLARE(UInt64, http_connections_soft_limit, 100, "Connections above this limit have significantly shorter time to live. The limit applies to the http connections which do not belong to any disk or storage.", 0) \
DECLARE(UInt64, http_connections_warn_limit, 1000, "Warning massages are written to the logs if number of in-use connections are higher than this limit. The limit applies to the http connections which do not belong to any disk or storage.", 0) \
DECLARE(UInt64, http_connections_store_limit, 5000, "Connections above this limit reset after use. Set to 0 to turn connection cache off. The limit applies to the http connections which do not belong to any disk or storage.", 0) \
DECLARE(UInt64, global_profiler_real_time_period_ns, 0, "Period for real clock timer of global profiler (in nanoseconds). Set 0 value to turn off the real clock global profiler. Recommended value is at least 10000000 (100 times a second) for single queries or 1000000000 (once a second) for cluster-wide profiling.", 0) \
DECLARE(UInt64, global_profiler_cpu_time_period_ns, 0, "Period for CPU clock timer of global profiler (in nanoseconds). Set 0 value to turn off the CPU clock global profiler. Recommended value is at least 10000000 (100 times a second) for single queries or 1000000000 (once a second) for cluster-wide profiling.", 0) \
DECLARE(Bool, enable_azure_sdk_logging, false, "Enables logging from Azure sdk", 0) \
DECLARE(UInt64, max_entries_for_hash_table_stats, 10'000, "How many entries hash table statistics collected during aggregation is allowed to have", 0) \
DECLARE(String, merge_workload, "default", "Name of workload to be used to access resources for all merges (may be overridden by a merge tree setting)", 0) \
DECLARE(String, mutation_workload, "default", "Name of workload to be used to access resources for all mutations (may be overridden by a merge tree setting)", 0) \
DECLARE(Bool, prepare_system_log_tables_on_startup, false, "If true, ClickHouse creates all configured `system.*_log` tables before the startup. It can be helpful if some startup scripts depend on these tables.", 0) \
DECLARE(Double, gwp_asan_force_sample_probability, 0.0003, "Probability that an allocation from specific places will be sampled by GWP Asan (i.e. PODArray allocations)", 0) \
DECLARE(UInt64, config_reload_interval_ms, 2000, "How often clickhouse will reload config and check for new changes", 0) \
DECLARE(UInt64, memory_worker_period_ms, 0, "Tick period of background memory worker which corrects memory tracker memory usages and cleans up unused pages during higher memory usage. If set to 0, default value will be used depending on the memory usage source", 0) \
DECLARE(Bool, disable_insertion_and_mutation, false, "Disable all insert/alter/delete queries. This setting will be enabled if someone needs read-only nodes to prevent insertion and mutation affect reading performance.", 0) \
DECLARE(UInt64, parts_kill_delay_period, 30, "Period to completely remove parts for SharedMergeTree. Only available in ClickHouse Cloud", 0) \
DECLARE(UInt64, parts_kill_delay_period_random_add, 10, "Add uniformly distributed value from 0 to x seconds to kill_delay_period to avoid thundering herd effect and subsequent DoS of ZooKeeper in case of very large number of tables. Only available in ClickHouse Cloud", 0) \
DECLARE(UInt64, parts_killer_pool_size, 128, "Threads for cleanup of shared merge tree outdated threads. Only available in ClickHouse Cloud", 0) \
DECLARE(UInt64, keeper_multiread_batch_size, 10'000, "Maximum size of batch for MultiRead request to [Zoo]Keeper that support batching. If set to 0, batching is disabled. Available only in ClickHouse Cloud.", 0) \
DECLARE(Bool, use_legacy_mongodb_integration, true, "Use the legacy MongoDB integration implementation. Note: it's highly recommended to set this option to false, since legacy implementation will be removed in the future. Please submit any issues you encounter with the new implementation.", 0) \
/// If you add a setting which can be updated at runtime, please update 'changeable_settings' map in dumpToSystemServerSettingsColumns below
DECLARE_SETTINGS_TRAITS(ServerSettingsTraits, LIST_OF_SERVER_SETTINGS)
IMPLEMENT_SETTINGS_TRAITS(ServerSettingsTraits, LIST_OF_SERVER_SETTINGS)
struct ServerSettingsImpl : public BaseSettings<ServerSettingsTraits>
{
void loadSettingsFromConfig(const Poco::Util::AbstractConfiguration & config);
};
IMPLEMENT_SETTINGS_TRAITS(ServerSettingsTraits, LIST_OF_SERVER_SETTINGS)
void ServerSettingsImpl::loadSettingsFromConfig(const Poco::Util::AbstractConfiguration & config)
{
// settings which can be loaded from the the default profile, see also MAKE_DEPRECATED_BY_SERVER_CONFIG in src/Core/Settings.h
@ -231,7 +231,7 @@ void ServerSettingsImpl::loadSettingsFromConfig(const Poco::Util::AbstractConfig
}
#define INITIALIZE_SETTING_EXTERN(TYPE, NAME, DEFAULT, DESCRIPTION, FLAGS) ServerSettings##TYPE NAME = &ServerSettings##Impl ::NAME;
#define INITIALIZE_SETTING_EXTERN(TYPE, NAME, DEFAULT, DESCRIPTION, FLAGS) ServerSettings##TYPE NAME = &ServerSettingsImpl ::NAME;
namespace ServerSetting
{
@ -250,18 +250,7 @@ ServerSettings::ServerSettings(const ServerSettings & settings) : impl(std::make
ServerSettings::~ServerSettings() = default;
#define IMPLEMENT_SETTING_SUBSCRIPT_OPERATOR(CLASS_NAME, TYPE) \
const SettingField##TYPE & ServerSettings::operator[](CLASS_NAME##TYPE t) const \
{ \
return impl.get()->*t; \
} \
SettingField##TYPE & ServerSettings::operator[](CLASS_NAME##TYPE t) \
{ \
return impl.get()->*t; \
}
SERVER_SETTINGS_SUPPORTED_TYPES(ServerSettings, IMPLEMENT_SETTING_SUBSCRIPT_OPERATOR)
#undef IMPLEMENT_SETTING_SUBSCRIPT_OPERATOR
void ServerSettings::set(std::string_view name, const Field & value)
{

1698
src/Core/Settings.cpp
View File

File diff suppressed because it is too large Load Diff

5
src/Core/SettingsChangesHistory.cpp
View File

@ -77,13 +77,16 @@ static std::initializer_list<std::pair<ClickHouseVersion, SettingsChangesHistory
{"restore_replace_external_dictionary_source_to_null", false, false, "New setting."},
{"show_create_query_identifier_quoting_rule", "when_necessary", "when_necessary", "New setting."},
{"show_create_query_identifier_quoting_style", "Backticks", "Backticks", "New setting."},
{"merge_tree_min_read_task_size", 8, 8, "New setting"},
{"merge_tree_min_rows_for_concurrent_read_for_remote_filesystem", (20 * 8192), 0, "Setting is deprecated"},
{"merge_tree_min_bytes_for_concurrent_read_for_remote_filesystem", (24 * 10 * 1024 * 1024), 0, "Setting is deprecated"},
{"implicit_select", false, false, "A new setting."},
{"output_format_native_write_json_as_string", false, false, "Add new setting to allow write JSON column as single String column in Native format"},
{"output_format_binary_write_json_as_string", false, false, "Add new setting to write values of JSON type as JSON string in RowBinary output format"},
{"input_format_binary_read_json_as_string", false, false, "Add new setting to read values of JSON type as JSON string in RowBinary input format"},
{"min_free_disk_bytes_to_perform_insert", 0, 0, "New setting."},
{"min_free_disk_ratio_to_perform_insert", 0.0, 0.0, "New setting."},
{"enable_named_columns_in_function_tuple", false, false, "Force disable the setting since it breaks queries"},
{"enable_named_columns_in_function_tuple", false, true, "Re-enable the setting since all known bugs are fixed"},
{"cloud_mode_database_engine", 1, 1, "A setting for ClickHouse Cloud"},
{"allow_experimental_shared_set_join", 1, 1, "A setting for ClickHouse Cloud"},
{"read_through_distributed_cache", 0, 0, "A setting for ClickHouse Cloud"},

15
src/Core/SettingsEnums.h
View File

@ -1,20 +1,21 @@
#pragma once
#include <Access/Common/SQLSecurityDefs.h>
#include <Core/Joins.h>
#include <Core/LoadBalancing.h>
#include <Core/LogsLevel.h>
#include <Core/MergeSelectorAlgorithm.h>
#include <Core/ParallelReplicasMode.h>
#include <Core/QueryLogElementType.h>
#include <Core/SchemaInferenceMode.h>
#include <Core/SettingsFields.h>
#include <Core/ShortCircuitFunctionEvaluation.h>
#include <Core/ParallelReplicasMode.h>
#include <Core/StreamingHandleErrorMode.h>
#include <Formats/FormatSettings.h>
#include <IO/ReadSettings.h>
#include <Access/Common/SQLSecurityDefs.h>
#include <Parsers/IdentifierQuotingStyle.h>
#include <QueryPipeline/SizeLimits.h>
#include <Common/ShellCommandSettings.h>
#include <Core/MergeSelectorAlgorithm.h>
namespace DB
@ -263,14 +264,6 @@ enum class DistributedDDLOutputMode : uint8_t
DECLARE_SETTING_ENUM(DistributedDDLOutputMode)
enum class StreamingHandleErrorMode : uint8_t
{
DEFAULT = 0, // Ignore errors with threshold.
STREAM, // Put errors to stream in the virtual column named ``_error.
/*FIXED_SYSTEM_TABLE, Put errors to in a fixed system table likely system.kafka_errors. This is not implemented now. */
/*CUSTOM_SYSTEM_TABLE, Put errors to in a custom system table. This is not implemented now. */
};
DECLARE_SETTING_ENUM(StreamingHandleErrorMode)
DECLARE_SETTING_ENUM(ShortCircuitFunctionEvaluation)

16
src/Core/StreamingHandleErrorMode.h Normal file
View File

@ -0,0 +1,16 @@
#pragma once
#include <cstdint>
namespace DB
{
enum class StreamingHandleErrorMode : uint8_t
{
DEFAULT = 0, // Ignore errors with threshold.
STREAM, // Put errors to stream in the virtual column named ``_error.
/*FIXED_SYSTEM_TABLE, Put errors to in a fixed system table likely system.kafka_errors. This is not implemented now. */
/*CUSTOM_SYSTEM_TABLE, Put errors to in a custom system table. This is not implemented now. */
};
}

2
src/DataTypes/DataTypeArray.h
View File

@ -47,8 +47,8 @@ public:
Field getDefault() const override;
DataTypePtr getNormalizedType() const override { return std::make_shared<DataTypeArray>(nested->getNormalizedType()); }
bool equals(const IDataType & rhs) const override;
bool isParametric() const override { return true; }
bool haveSubtypes() const override { return true; }
bool cannotBeStoredInTables() const override { return nested->cannotBeStoredInTables(); }

5
src/DataTypes/DataTypeMap.h
View File

@ -43,7 +43,10 @@ public:
bool isParametric() const override { return true; }
bool haveSubtypes() const override { return true; }
bool hasDynamicSubcolumnsDeprecated() const override { return nested->hasDynamicSubcolumnsDeprecated(); }
DataTypePtr getNormalizedType() const override
{
return std::make_shared<DataTypeMap>(key_type->getNormalizedType(), value_type->getNormalizedType());
}
const DataTypePtr & getKeyType() const { return key_type; }
const DataTypePtr & getValueType() const { return value_type; }
DataTypes getKeyValueTypes() const { return {key_type, value_type}; }

8
src/DataTypes/DataTypeTuple.cpp
View File

@ -133,6 +133,14 @@ std::string DataTypeTuple::doGetPrettyName(size_t indent) const
return s.str();
}
DataTypePtr DataTypeTuple::getNormalizedType() const
{
DataTypes normalized_elems;
normalized_elems.reserve(elems.size());
for (const auto & elem : elems)
normalized_elems.emplace_back(elem->getNormalizedType());
return std::make_shared<DataTypeTuple>(normalized_elems);
}
static inline IColumn & extractElementColumn(IColumn & column, size_t idx)
{

1
src/DataTypes/DataTypeTuple.h
View File

@ -61,6 +61,7 @@ public:
MutableSerializationInfoPtr createSerializationInfo(const SerializationInfoSettings & settings) const override;
SerializationInfoPtr getSerializationInfo(const IColumn & column) const override;
DataTypePtr getNormalizedType() const override;
const DataTypePtr & getElement(size_t i) const { return elems[i]; }
const DataTypes & getElements() const { return elems; }
const Strings & getElementNames() const { return names; }

9
src/DataTypes/IDataType.h
View File

@ -88,6 +88,15 @@ public:
DataTypePtr getPtr() const { return shared_from_this(); }
/// Returns the normalized form of the current type, currently handling the
/// conversion of named tuples to unnamed tuples.
///
/// This is useful for converting aggregate states into a normalized form with
/// normalized argument types. E.g, `AggregateFunction(uniq, Tuple(a int, b int))`
/// should be convertible to `AggregateFunction(uniq, Tuple(int, int))`, as both
/// have same memory layouts for state representation and the same serialization.
virtual DataTypePtr getNormalizedType() const { return shared_from_this(); }
/// Name of data type family (example: FixedString, Array).
virtual const char * getFamilyName() const = 0;

12
src/Databases/DatabaseReplicated.cpp
View File

@ -64,6 +64,12 @@ namespace ServerSetting
extern const ServerSettingsUInt32 max_database_replicated_create_table_thread_pool_size;
}
namespace DatabaseReplicatedSetting
{
extern const DatabaseReplicatedSettingsString collection_name;
extern const DatabaseReplicatedSettingsFloat max_broken_tables_ratio;
}
namespace ErrorCodes
{
extern const int NO_ZOOKEEPER;
@ -141,8 +147,8 @@ DatabaseReplicated::DatabaseReplicated(
if (zookeeper_path.front() != '/')
zookeeper_path = "/" + zookeeper_path;
if (!db_settings.collection_name.value.empty())
fillClusterAuthInfo(db_settings.collection_name.value, context_->getConfigRef());
if (!db_settings[DatabaseReplicatedSetting::collection_name].value.empty())
fillClusterAuthInfo(db_settings[DatabaseReplicatedSetting::collection_name].value, context_->getConfigRef());
replica_group_name = context_->getConfigRef().getString("replica_group_name", "");
@ -1220,7 +1226,7 @@ void DatabaseReplicated::recoverLostReplica(const ZooKeeperPtr & current_zookeep
String db_name = getDatabaseName();
String to_db_name = getDatabaseName() + BROKEN_TABLES_SUFFIX;
String to_db_name_replicated = getDatabaseName() + BROKEN_REPLICATED_TABLES_SUFFIX;
if (total_tables * db_settings.max_broken_tables_ratio < tables_to_detach.size())
if (total_tables * db_settings[DatabaseReplicatedSetting::max_broken_tables_ratio] < tables_to_detach.size())
throw Exception(ErrorCodes::DATABASE_REPLICATION_FAILED, "Too many tables to recreate: {} of {}", tables_to_detach.size(), total_tables);
if (!tables_to_detach.empty())
{

47
src/Databases/DatabaseReplicatedSettings.cpp
View File

@ -1,17 +1,60 @@
#include <Core/BaseSettings.h>
#include <Core/BaseSettingsFwdMacrosImpl.h>
#include <Databases/DatabaseReplicatedSettings.h>
#include <Parsers/ASTFunction.h>
#include <Parsers/ASTCreateQuery.h>
#include <Parsers/ASTFunction.h>
namespace DB
{
#define LIST_OF_DATABASE_REPLICATED_SETTINGS(DECLARE, ALIAS) \
DECLARE(Float, max_broken_tables_ratio, 1, "Do not recover replica automatically if the ratio of staled tables to all tables is greater", 0) \
DECLARE(UInt64, max_replication_lag_to_enqueue, 50, "Replica will throw exception on attempt to execute query if its replication lag greater", 0) \
DECLARE(UInt64, wait_entry_commited_timeout_sec, 3600, "Replicas will try to cancel query if timeout exceed, but initiator host has not executed it yet", 0) \
DECLARE(String, collection_name, "", "A name of a collection defined in server's config where all info for cluster authentication is defined", 0) \
DECLARE(Bool, check_consistency, true, "Check consistency of local metadata and metadata in Keeper, do replica recovery on inconsistency", 0) \
DECLARE(UInt64, max_retries_before_automatic_recovery, 100, "Max number of attempts to execute a queue entry before marking replica as lost recovering it from snapshot (0 means infinite)", 0) \
DECLARE_SETTINGS_TRAITS(DatabaseReplicatedSettingsTraits, LIST_OF_DATABASE_REPLICATED_SETTINGS)
IMPLEMENT_SETTINGS_TRAITS(DatabaseReplicatedSettingsTraits, LIST_OF_DATABASE_REPLICATED_SETTINGS)
struct DatabaseReplicatedSettingsImpl : public BaseSettings<DatabaseReplicatedSettingsTraits>
{
};
#define INITIALIZE_SETTING_EXTERN(TYPE, NAME, DEFAULT, DESCRIPTION, FLAGS) \
DatabaseReplicatedSettings##TYPE NAME = &DatabaseReplicatedSettingsImpl ::NAME;
namespace DatabaseReplicatedSetting
{
LIST_OF_DATABASE_REPLICATED_SETTINGS(INITIALIZE_SETTING_EXTERN, SKIP_ALIAS)
}
#undef INITIALIZE_SETTING_EXTERN
DatabaseReplicatedSettings::DatabaseReplicatedSettings() : impl(std::make_unique<DatabaseReplicatedSettingsImpl>())
{
}
DatabaseReplicatedSettings::DatabaseReplicatedSettings(const DatabaseReplicatedSettings & settings)
: impl(std::make_unique<DatabaseReplicatedSettingsImpl>(*settings.impl))
{
}
DatabaseReplicatedSettings::DatabaseReplicatedSettings(DatabaseReplicatedSettings && settings) noexcept
: impl(std::make_unique<DatabaseReplicatedSettingsImpl>(std::move(*settings.impl)))
{
}
DatabaseReplicatedSettings::~DatabaseReplicatedSettings() = default;
DATABASE_REPLICATED_SETTINGS_SUPPORTED_TYPES(DatabaseReplicatedSettings, IMPLEMENT_SETTING_SUBSCRIPT_OPERATOR)
void DatabaseReplicatedSettings::loadFromQuery(ASTStorage & storage_def)
{
if (storage_def.settings)
{
applyChanges(storage_def.settings->changes);
impl->applyChanges(storage_def.settings->changes);
return;
}

36
src/Databases/DatabaseReplicatedSettings.h
View File

@ -1,27 +1,35 @@
#pragma once
#include <Core/Defines.h>
#include <Core/BaseSettings.h>
#include <Core/BaseSettingsFwdMacros.h>
#include <Core/SettingsFields.h>
namespace DB
{
class ASTStorage;
struct DatabaseReplicatedSettingsImpl;
#define LIST_OF_DATABASE_REPLICATED_SETTINGS(M, ALIAS) \
M(Float, max_broken_tables_ratio, 1, "Do not recover replica automatically if the ratio of staled tables to all tables is greater", 0) \
M(UInt64, max_replication_lag_to_enqueue, 50, "Replica will throw exception on attempt to execute query if its replication lag greater", 0) \
M(UInt64, wait_entry_commited_timeout_sec, 3600, "Replicas will try to cancel query if timeout exceed, but initiator host has not executed it yet", 0) \
M(String, collection_name, "", "A name of a collection defined in server's config where all info for cluster authentication is defined", 0) \
M(Bool, check_consistency, true, "Check consistency of local metadata and metadata in Keeper, do replica recovery on inconsistency", 0) \
M(UInt64, max_retries_before_automatic_recovery, 100, "Max number of attempts to execute a queue entry before marking replica as lost recovering it from snapshot (0 means infinite)", 0) \
/// List of available types supported in ReplicatedSettings object
#define DATABASE_REPLICATED_SETTINGS_SUPPORTED_TYPES(CLASS_NAME, M) \
M(CLASS_NAME, Bool) \
M(CLASS_NAME, Float) \
M(CLASS_NAME, String) \
M(CLASS_NAME, UInt64)
DATABASE_REPLICATED_SETTINGS_SUPPORTED_TYPES(DatabaseReplicatedSettings, DECLARE_SETTING_TRAIT)
DECLARE_SETTINGS_TRAITS(DatabaseReplicatedSettingsTraits, LIST_OF_DATABASE_REPLICATED_SETTINGS)
struct DatabaseReplicatedSettings : public BaseSettings<DatabaseReplicatedSettingsTraits>
struct DatabaseReplicatedSettings
{
DatabaseReplicatedSettings();
DatabaseReplicatedSettings(const DatabaseReplicatedSettings & settings);
DatabaseReplicatedSettings(DatabaseReplicatedSettings && settings) noexcept;
~DatabaseReplicatedSettings();
DATABASE_REPLICATED_SETTINGS_SUPPORTED_TYPES(DatabaseReplicatedSettings, DECLARE_SETTING_SUBSCRIPT_OPERATOR)
void loadFromQuery(ASTStorage & storage_def);
private:
std::unique_ptr<DatabaseReplicatedSettingsImpl> impl;
};
}

18
src/Databases/DatabaseReplicatedWorker.cpp
View File

@ -18,6 +18,14 @@ namespace Setting
extern const SettingsUInt64 database_replicated_initial_query_timeout_sec;
}
namespace DatabaseReplicatedSetting
{
extern const DatabaseReplicatedSettingsBool check_consistency;
extern const DatabaseReplicatedSettingsUInt64 max_replication_lag_to_enqueue;
extern const DatabaseReplicatedSettingsUInt64 max_retries_before_automatic_recovery;
extern const DatabaseReplicatedSettingsUInt64 wait_entry_commited_timeout_sec;
}
namespace ErrorCodes
{
extern const int LOGICAL_ERROR;
@ -63,8 +71,8 @@ bool DatabaseReplicatedDDLWorker::initializeMainThread()
break;
}
if (database->db_settings.max_retries_before_automatic_recovery &&
database->db_settings.max_retries_before_automatic_recovery <= subsequent_errors_count)
if (database->db_settings[DatabaseReplicatedSetting::max_retries_before_automatic_recovery]
&& database->db_settings[DatabaseReplicatedSetting::max_retries_before_automatic_recovery] <= subsequent_errors_count)
{
String current_task_name;
{
@ -155,7 +163,7 @@ void DatabaseReplicatedDDLWorker::initializeReplication()
bool is_new_replica = our_log_ptr == 0;
bool lost_according_to_log_ptr = our_log_ptr + logs_to_keep < max_log_ptr;
bool lost_according_to_digest = database->db_settings.check_consistency && local_digest != digest;
bool lost_according_to_digest = database->db_settings[DatabaseReplicatedSetting::check_consistency] && local_digest != digest;
if (is_new_replica || lost_according_to_log_ptr || lost_according_to_digest)
{
@ -306,7 +314,7 @@ String DatabaseReplicatedDDLWorker::tryEnqueueAndExecuteEntry(DDLLogEntry & entr
UInt32 our_log_ptr = getLogPointer();
UInt32 max_log_ptr = parse<UInt32>(zookeeper->get(database->zookeeper_path + "/max_log_ptr"));
if (our_log_ptr + database->db_settings.max_replication_lag_to_enqueue < max_log_ptr)
if (our_log_ptr + database->db_settings[DatabaseReplicatedSetting::max_replication_lag_to_enqueue] < max_log_ptr)
throw Exception(ErrorCodes::NOT_A_LEADER, "Cannot enqueue query on this replica, "
"because it has replication lag of {} queries. Try other replica.", max_log_ptr - our_log_ptr);
@ -399,7 +407,7 @@ DDLTaskPtr DatabaseReplicatedDDLWorker::initAndCheckTask(const String & entry_na
/// Query is not committed yet. We cannot just skip it and execute next one, because reordering may break replication.
LOG_TRACE(log, "Waiting for initiator {} to commit or rollback entry {}", initiator_name, entry_path);
constexpr size_t wait_time_ms = 1000;
size_t max_iterations = database->db_settings.wait_entry_commited_timeout_sec;
size_t max_iterations = database->db_settings[DatabaseReplicatedSetting::wait_entry_commited_timeout_sec];
size_t iteration = 0;
while (!wait_committed_or_failed->tryWait(wait_time_ms))

23
src/Databases/MySQL/DatabaseMaterializedMySQL.cpp
View File

@ -4,6 +4,7 @@
# include <Core/Settings.h>
# include <Databases/MySQL/DatabaseMaterializedMySQL.h>
# include <Databases/MySQL/MaterializedMySQLSettings.h>
# include <Common/parseAddress.h>
# include <Common/parseRemoteDescription.h>
@ -18,6 +19,7 @@
# include <Storages/StorageMySQL.h>
# include <Storages/StorageMaterializedMySQL.h>
# include <Storages/NamedCollectionsHelpers.h>
# include <Storages/MySQL/MySQLSettings.h>
# include <Common/setThreadName.h>
# include <Common/PoolId.h>
# include <filesystem>
@ -31,6 +33,15 @@ namespace Setting
extern const SettingsUInt64 glob_expansion_max_elements;
}
namespace MaterializedMySQLSetting
{
extern const MaterializedMySQLSettingsBool allows_query_when_mysql_lost;
extern const MaterializedMySQLSettingsBool allow_startup_database_without_connection_to_mysql;
extern const MaterializedMySQLSettingsUInt64 max_bytes_in_binlog_dispatcher_buffer;
extern const MaterializedMySQLSettingsUInt64 max_flush_milliseconds_in_binlog_dispatcher;
extern const MaterializedMySQLSettingsBool use_binlog_client;
}
namespace ErrorCodes
{
extern const int NOT_IMPLEMENTED;
@ -53,11 +64,13 @@ DatabaseMaterializedMySQL::DatabaseMaterializedMySQL(
{
}
DatabaseMaterializedMySQL::~DatabaseMaterializedMySQL() = default;
void DatabaseMaterializedMySQL::rethrowExceptionIfNeeded() const
{
std::lock_guard lock(mutex);
if (!settings->allows_query_when_mysql_lost && exception)
if (!(*settings)[MaterializedMySQLSetting::allows_query_when_mysql_lost] && exception)
{
try
{
@ -89,7 +102,7 @@ LoadTaskPtr DatabaseMaterializedMySQL::startupDatabaseAsync(AsyncLoader & async_
[this, mode] (AsyncLoader &, const LoadJobPtr &)
{
LOG_TRACE(log, "Starting MaterializeMySQL database");
if (!settings->allow_startup_database_without_connection_to_mysql
if (!(*settings)[MaterializedMySQLSetting::allow_startup_database_without_connection_to_mysql]
&& mode < LoadingStrictnessLevel::FORCE_ATTACH)
materialize_thread.assertMySQLAvailable();
@ -266,11 +279,11 @@ void registerDatabaseMaterializedMySQL(DatabaseFactory & factory)
if (engine_define->settings)
materialize_mode_settings->loadFromQuery(*engine_define);
if (materialize_mode_settings->use_binlog_client)
if ((*materialize_mode_settings)[MaterializedMySQLSetting::use_binlog_client])
binlog_client = DB::MySQLReplication::BinlogClientFactory::instance().getClient(
configuration.host, configuration.port, configuration.username, configuration.password,
materialize_mode_settings->max_bytes_in_binlog_dispatcher_buffer,
materialize_mode_settings->max_flush_milliseconds_in_binlog_dispatcher);
(*materialize_mode_settings)[MaterializedMySQLSetting::max_bytes_in_binlog_dispatcher_buffer],
(*materialize_mode_settings)[MaterializedMySQLSetting::max_flush_milliseconds_in_binlog_dispatcher]);
if (args.uuid == UUIDHelpers::Nil)
{

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