def send_report_to_s3(head_report: Version2Report) -> None:
job = os.getenv('JOB_BASE_NAME', os.environ.get('CIRCLE_JOB'))
sha1 = os.environ.get('SHA1', os.environ.get('CIRCLE_SHA1', ''))
branch = os.environ.get('BRANCH', os.environ.get('CIRCLE_BRANCH', ''))
now = datetime.datetime.utcnow().isoformat()
# SHARD_NUMBER and TEST_CONFIG are specific to GHA, as these details would be included in CIRCLE_JOB already
shard = os.environ.get('SHARD_NUMBER', '')
test_config = os.environ.get('TEST_CONFIG')
job_report_dirname = f'{job}{f"-{test_config}" if test_config is not None else ""}{shard}'
if branch not in ['master', 'nightly'
] and not branch.startswith("release/"):
pr = os.environ.get('PR_NUMBER',
os.environ.get('CIRCLE_PR_NUMBER', 'unknown'))
key = f'pr_test_time/{pr}/{sha1}/{job_report_dirname}/{now}Z.json.bz2' # Z meaning UTC
else:
key = f'test_time/{sha1}/{job_report_dirname}/{now}Z.json.bz2' # Z meaning UTC
obj = get_S3_object_from_bucket('ossci-metrics', key)
# use bz2 because the results are smaller than gzip, and the
# compression time penalty we pay is only about half a second for
# input files of a few megabytes in size like these JSON files, and
# because for some reason zlib doesn't seem to play nice with the
# gunzip command whereas Python's bz2 does work with bzip2
obj.put(Body=bz2.compress(json.dumps(head_report).encode()))
def assemble_flaky_test_stats(
duplicated_tests_by_file: Dict[str, DuplicatedDict]) -> Any:
flaky_tests = []
workflow_id = os.environ.get("GITHUB_RUN_ID",
os.environ.get("CIRCLE_WORKFLOW_ID", None))
for file_name, suite_to_dict in duplicated_tests_by_file.items():
for suite_name, testcase_to_runs in suite_to_dict.items():
for testcase_name, list_of_runs in testcase_to_runs.items():
num_green, num_red = process_intentional_test_runs(
list_of_runs)
if num_green > 0: # Otherwise, it's likely just a failing test
flaky_tests.append({
"name": testcase_name,
"suite": suite_name,
"file": file_name,
"num_green": num_green,
"num_red": num_red,
})
if len(flaky_tests) > 0:
# write to RDS
register_rds_schema("flaky_tests", schema_from_sample(flaky_tests[0]))
rds_write("flaky_tests", flaky_tests, only_on_master=False)
# write to S3 to go to Rockset as well
import uuid
for flaky_test in flaky_tests:
flaky_test["workflow_id"] = workflow_id
key = f"flaky_tests/{workflow_id}/{uuid.uuid4()}.json"
obj = get_S3_object_from_bucket("ossci-raw-job-status", key)
obj.put(Body=json.dumps(flaky_test),
ContentType="application/json")
def write_flaky_test_stats_to_rockset(
duplicated_tests_by_file: Dict[str, DuplicatedDict]) -> Any:
flaky_tests = []
workflow_id = os.environ.get("GITHUB_RUN_ID",
os.environ.get("CIRCLE_WORKFLOW_ID", None))
for file_name, suite_to_dict in duplicated_tests_by_file.items():
for suite_name, testcase_to_runs in suite_to_dict.items():
for testcase_name, list_of_runs in testcase_to_runs.items():
num_green, num_red = process_intentional_test_runs(
list_of_runs)
if (
num_green > 0 and num_red > 0
): # Flaky tests show different results in consecutive reruns
flaky_tests.append({
"name": testcase_name,
"suite": suite_name,
"file": file_name,
"num_green": num_green,
"num_red": num_red,
})
if len(flaky_tests) > 0:
import uuid
for flaky_test in flaky_tests:
flaky_test["job_id"] = os.environ["GHA_WORKFLOW_JOB_ID"]
flaky_test["workflow_id"] = workflow_id
key = f"flaky_tests/{workflow_id}/{uuid.uuid4()}.json"
obj = get_S3_object_from_bucket("ossci-raw-job-status", key)
obj.put(Body=json.dumps(flaky_test),
ContentType="application/json")
def send_report_to_s3(head_report: Version2Report) -> None:
job = os.environ.get('JOB_BASE_NAME')
sha1 = os.environ.get('CIRCLE_SHA1')
branch = os.environ.get('CIRCLE_BRANCH', '')
now = datetime.datetime.utcnow().isoformat()
if branch not in ['master', 'nightly'] and not branch.startswith("release/"):
pr = os.environ.get('CIRCLE_PR_NUMBER', 'unknown')
key = f'pr_test_time/{pr}/{sha1}/{job}/{now}Z.json.bz2' # Z meaning UTC
else:
key = f'test_time/{sha1}/{job}/{now}Z.json.bz2' # Z meaning UTC
obj = get_S3_object_from_bucket('ossci-metrics', key)
# use bz2 because the results are smaller than gzip, and the
# compression time penalty we pay is only about half a second for
# input files of a few megabytes in size like these JSON files, and
# because for some reason zlib doesn't seem to play nice with the
# gunzip command whereas Python's bz2 does work with bzip2
obj.put(Body=bz2.compress(json.dumps(head_report).encode()))
def send_report_to_s3(head_report: Version2Report) -> None:
job = os.getenv("JOB_BASE_NAME", os.environ.get("CIRCLE_JOB"))
sha1 = os.environ.get("SHA1", os.environ.get("CIRCLE_SHA1", ""))
now = datetime.datetime.utcnow().isoformat()
# SHARD_NUMBER and TEST_CONFIG are specific to GHA, as these details would be included in CIRCLE_JOB already
shard = os.environ.get("SHARD_NUMBER", "")
test_config = os.environ.get("TEST_CONFIG")
job_report_dirname = (
f'{job}{f"-{test_config}" if test_config is not None else ""}{shard}')
key = f"test_time/{sha1}/{job_report_dirname}/{now}Z.json.bz2" # Z meaning UTC
obj = get_S3_object_from_bucket("ossci-metrics", key)
# use bz2 because the results are smaller than gzip, and the
# compression time penalty we pay is only about half a second for
# input files of a few megabytes in size like these JSON files, and
# because for some reason zlib doesn't seem to play nice with the
# gunzip command whereas Python's bz2 does work with bzip2
obj.put(Body=bz2.compress(json.dumps(head_report).encode()))
