mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-12-12 01:12:12 +00:00
286 lines
11 KiB
Python
286 lines
11 KiB
Python
#!/usr/bin/env python3
|
|
|
|
import csv
|
|
import logging
|
|
import subprocess
|
|
import os
|
|
import json
|
|
import time
|
|
import sys
|
|
|
|
from github import Github
|
|
import requests
|
|
|
|
from report import create_test_html_report
|
|
from s3_helper import S3Helper
|
|
from get_robot_token import get_best_robot_token
|
|
from pr_info import PRInfo
|
|
|
|
DOWNLOAD_RETRIES_COUNT = 5
|
|
|
|
def process_logs(s3_client, additional_logs, s3_path_prefix):
|
|
additional_urls = []
|
|
for log_path in additional_logs:
|
|
if log_path:
|
|
additional_urls.append(
|
|
s3_client.upload_test_report_to_s3(
|
|
log_path,
|
|
s3_path_prefix + "/" + os.path.basename(log_path)))
|
|
|
|
return additional_urls
|
|
|
|
|
|
def upload_results(s3_client, pr_number, commit_sha, test_results, raw_log, additional_files, check_name):
|
|
additional_files = [raw_log] + additional_files
|
|
s3_path_prefix = f"{pr_number}/{commit_sha}/" + check_name.lower().replace(' ', '_').replace('(', '_').replace(')', '_').replace(',', '_')
|
|
additional_urls = process_logs(s3_client, additional_files, s3_path_prefix)
|
|
|
|
branch_url = "https://github.com/ClickHouse/ClickHouse/commits/master"
|
|
branch_name = "master"
|
|
if pr_number != 0:
|
|
branch_name = f"PR #{pr_number}"
|
|
branch_url = f"https://github.com/ClickHouse/ClickHouse/pull/{pr_number}"
|
|
commit_url = f"https://github.com/ClickHouse/ClickHouse/commit/{commit_sha}"
|
|
|
|
task_url = f"https://github.com/ClickHouse/ClickHouse/actions/runs/{os.getenv('GITHUB_RUN_ID')}"
|
|
|
|
raw_log_url = additional_urls[0]
|
|
additional_urls.pop(0)
|
|
|
|
html_report = create_test_html_report(check_name, test_results, raw_log_url, task_url, branch_url, branch_name, commit_url, additional_urls, True)
|
|
with open('report.html', 'w', encoding='utf-8') as f:
|
|
f.write(html_report)
|
|
|
|
url = s3_client.upload_test_report_to_s3('report.html', s3_path_prefix + ".html")
|
|
logging.info("Search result in url %s", url)
|
|
return url
|
|
|
|
def get_commit(gh, commit_sha):
|
|
repo = gh.get_repo(os.getenv("GITHUB_REPOSITORY", "ClickHouse/ClickHouse"))
|
|
commit = repo.get_commit(commit_sha)
|
|
return commit
|
|
|
|
def get_image_name(check_name):
|
|
if 'stateless' in check_name.lower():
|
|
return 'clickhouse/stateless-test'
|
|
if 'stateful' in check_name.lower():
|
|
return 'clickhouse/stateful-test'
|
|
else:
|
|
raise Exception(f"Cannot deduce image name based on check name {check_name}")
|
|
|
|
|
|
def dowload_build_with_progress(url, path):
|
|
logging.info("Downloading from %s to temp path %s", url, path)
|
|
for i in range(DOWNLOAD_RETRIES_COUNT):
|
|
try:
|
|
with open(path, 'wb') as f:
|
|
response = requests.get(url, stream=True)
|
|
response.raise_for_status()
|
|
total_length = response.headers.get('content-length')
|
|
if total_length is None or int(total_length) == 0:
|
|
logging.info("No content-length, will download file without progress")
|
|
f.write(response.content)
|
|
else:
|
|
dl = 0
|
|
total_length = int(total_length)
|
|
logging.info("Content length is %ld bytes", total_length)
|
|
for data in response.iter_content(chunk_size=4096):
|
|
dl += len(data)
|
|
f.write(data)
|
|
if sys.stdout.isatty():
|
|
done = int(50 * dl / total_length)
|
|
percent = int(100 * float(dl) / total_length)
|
|
eq_str = '=' * done
|
|
space_str = ' ' * (50 - done)
|
|
sys.stdout.write(f"\r[{eq_str}{space_str}] {percent}%")
|
|
sys.stdout.flush()
|
|
break
|
|
except Exception as ex:
|
|
sys.stdout.write("\n")
|
|
time.sleep(3)
|
|
logging.info("Exception while downloading %s, retry %s", ex, i + 1)
|
|
if os.path.exists(path):
|
|
os.remove(path)
|
|
else:
|
|
raise Exception(f"Cannot download dataset from {url}, all retries exceeded")
|
|
|
|
sys.stdout.write("\n")
|
|
logging.info("Downloading finished")
|
|
|
|
|
|
def download_builds(result_path, build_urls):
|
|
for url in build_urls:
|
|
if url.endswith('.deb'):
|
|
fname = os.path.basename(url)
|
|
logging.info("Will download %s to %s", fname, result_path)
|
|
dowload_build_with_progress(url, os.path.join(result_path, fname))
|
|
|
|
|
|
def get_build_config(build_number, repo_path):
|
|
ci_config_path = os.path.join(repo_path, "tests/ci/ci_config.json")
|
|
with open(ci_config_path, 'r', encoding='utf-8') as ci_config:
|
|
config_dict = json.load(ci_config)
|
|
return config_dict['build_config'][build_number]
|
|
|
|
def get_build_urls(build_config_str, reports_path):
|
|
for root, _, files in os.walk(reports_path):
|
|
for f in files:
|
|
if build_config_str in f :
|
|
logging.info("Found build report json %s", f)
|
|
with open(os.path.join(root, f), 'r', encoding='utf-8') as file_handler:
|
|
build_report = json.load(file_handler)
|
|
return build_report['build_urls']
|
|
return []
|
|
|
|
def build_config_to_string(build_config):
|
|
if build_config["package-type"] == "performance":
|
|
return "performance"
|
|
|
|
return "_".join([
|
|
build_config['compiler'],
|
|
build_config['build-type'] if build_config['build-type'] else "relwithdebuginfo",
|
|
build_config['sanitizer'] if build_config['sanitizer'] else "none",
|
|
build_config['bundled'],
|
|
build_config['splitted'],
|
|
"tidy" if build_config['tidy'] == "enable" else "notidy",
|
|
"with_coverage" if build_config['with_coverage'] else "without_coverage",
|
|
build_config['package-type'],
|
|
])
|
|
|
|
def get_run_command(builds_path, result_path, server_log_path, kill_timeout, additional_envs, image):
|
|
additional_options = ['--hung-check']
|
|
additional_options.append('--print-time')
|
|
additional_options_str = '-e ADDITIONAL_OPTIONS="' + ' '.join(additional_options) + '"'
|
|
|
|
envs = [f'-e MAX_RUN_TIME={int(0.9 * kill_timeout)}', '-e S3_URL="https://clickhouse-datasets.s3.amazonaws.com"']
|
|
envs += [f'-e {e}' for e in additional_envs]
|
|
env_str = ' '.join(envs)
|
|
|
|
return f"docker run --volume={builds_path}:/package_folder " \
|
|
f"--volume={result_path}:/test_output --volume={server_log_path}:/var/log/clickhouse-server " \
|
|
f"--cap-add=SYS_PTRACE {env_str} {additional_options_str} {image}"
|
|
|
|
|
|
def process_results(result_folder, server_log_path):
|
|
test_results = []
|
|
additional_files = []
|
|
# Just upload all files from result_folder.
|
|
# If task provides processed results, then it's responsible for content of result_folder.
|
|
if os.path.exists(result_folder):
|
|
test_files = [f for f in os.listdir(result_folder) if os.path.isfile(os.path.join(result_folder, f))]
|
|
additional_files = [os.path.join(result_folder, f) for f in test_files]
|
|
|
|
if os.path.exists(server_log_path):
|
|
server_log_files = [f for f in os.listdir(server_log_path) if os.path.isfile(os.path.join(server_log_path, f))]
|
|
additional_files = additional_files + [os.path.join(server_log_path, f) for f in server_log_files]
|
|
|
|
status_path = os.path.join(result_folder, "check_status.tsv")
|
|
logging.info("Found test_results.tsv")
|
|
with open(status_path, 'r', encoding='utf-8') as status_file:
|
|
status = list(csv.reader(status_file, delimiter='\t'))
|
|
|
|
if len(status) != 1 or len(status[0]) != 2:
|
|
return "error", "Invalid check_status.tsv", test_results, additional_files
|
|
state, description = status[0][0], status[0][1]
|
|
|
|
results_path = os.path.join(result_folder, "test_results.tsv")
|
|
with open(results_path, 'r', encoding='utf-8') as results_file:
|
|
test_results = list(csv.reader(results_file, delimiter='\t'))
|
|
if len(test_results) == 0:
|
|
raise Exception("Empty results")
|
|
|
|
return state, description, test_results, additional_files
|
|
|
|
|
|
if __name__ == "__main__":
|
|
logging.basicConfig(level=logging.INFO)
|
|
temp_path = os.getenv("TEMP_PATH", os.path.abspath("."))
|
|
repo_path = os.getenv("REPO_COPY", os.path.abspath("../../"))
|
|
reports_path = os.getenv("REPORTS_PATH", "./reports")
|
|
|
|
check_name = sys.argv[1]
|
|
build_number = int(sys.argv[2])
|
|
kill_timeout = int(sys.argv[3])
|
|
|
|
if not os.path.exists(temp_path):
|
|
os.makedirs(temp_path)
|
|
|
|
with open(os.getenv('GITHUB_EVENT_PATH'), 'r', encoding='utf-8') as event_file:
|
|
event = json.load(event_file)
|
|
|
|
pr_info = PRInfo(event)
|
|
|
|
gh = Github(get_best_robot_token())
|
|
|
|
|
|
for root, _, files in os.walk(reports_path):
|
|
for f in files:
|
|
if f == 'changed_images.json':
|
|
images_path = os.path.join(root, 'changed_images.json')
|
|
break
|
|
|
|
image_name = get_image_name(check_name)
|
|
|
|
docker_image = image_name
|
|
if images_path and os.path.exists(images_path):
|
|
logging.info("Images file exists")
|
|
with open(images_path, 'r', encoding='utf-8') as images_fd:
|
|
images = json.load(images_fd)
|
|
logging.info("Got images %s", images)
|
|
if image_name in images:
|
|
docker_image += ':' + images[image_name]
|
|
else:
|
|
logging.info("Images file not found")
|
|
|
|
for i in range(10):
|
|
try:
|
|
logging.info("Pulling image %s", docker_image)
|
|
subprocess.check_output(f"docker pull {docker_image}", stderr=subprocess.STDOUT, shell=True)
|
|
break
|
|
except Exception as ex:
|
|
time.sleep(i * 3)
|
|
logging.info("Got execption pulling docker %s", ex)
|
|
else:
|
|
raise Exception(f"Cannot pull dockerhub for image docker pull {docker_image}")
|
|
|
|
build_config = get_build_config(build_number, repo_path)
|
|
build_config_str = build_config_to_string(build_config)
|
|
urls = get_build_urls(build_config_str, reports_path)
|
|
if not urls:
|
|
raise Exception("No build URLs found")
|
|
|
|
packages_path = os.path.join(temp_path, "packages")
|
|
if not os.path.exists(packages_path):
|
|
os.makedirs(packages_path)
|
|
|
|
server_log_path = os.path.join(temp_path, "server_log")
|
|
if not os.path.exists(server_log_path):
|
|
os.makedirs(server_log_path)
|
|
|
|
result_path = os.path.join(temp_path, "result_path")
|
|
if not os.path.exists(result_path):
|
|
os.makedirs(result_path)
|
|
|
|
run_log_path = os.path.join(result_path, "runlog.log")
|
|
|
|
download_builds(packages_path, urls)
|
|
run_command = get_run_command(packages_path, result_path, server_log_path, kill_timeout, [], docker_image)
|
|
logging.info("Going to run func tests: %s", run_command)
|
|
|
|
with open(run_log_path, 'w', encoding='utf-8') as log:
|
|
with subprocess.Popen(run_command, shell=True, stderr=log, stdout=log) as process:
|
|
retcode = process.wait()
|
|
if retcode == 0:
|
|
logging.info("Run successfully")
|
|
else:
|
|
logging.info("Run failed")
|
|
|
|
subprocess.check_call(f"sudo chown -R ubuntu:ubuntu {temp_path}", shell=True)
|
|
|
|
s3_helper = S3Helper('https://s3.amazonaws.com')
|
|
state, description, test_results, additional_logs = process_results(result_path, server_log_path)
|
|
report_url = upload_results(s3_helper, pr_info.number, pr_info.sha, test_results, run_log_path, additional_logs, check_name)
|
|
print(f"::notice ::Report url: {report_url}")
|
|
commit = get_commit(gh, pr_info.sha)
|
|
commit.create_status(context=check_name, description=description, state=state, target_url=report_url)
|