ClickHouse/tests/ci/cancel_and_rerun_workflow_lambda/app.py
2022-04-25 14:45:11 +02:00

349 lines
12 KiB
Python

#!/usr/bin/env python3
from collections import namedtuple
from typing import Any, Dict, List
from threading import Thread
from queue import Queue
import json
import time
import jwt
import requests # type: ignore
import boto3 # type: ignore
NEED_RERUN_OR_CANCELL_WORKFLOWS = {
"PullRequestCI",
"DocsCheck",
"DocsRelease",
"BackportPR",
}
# https://docs.github.com/en/rest/reference/actions#cancel-a-workflow-run
#
API_URL = "https://api.github.com/repos/ClickHouse/ClickHouse"
MAX_RETRY = 5
DEBUG_INFO = {} # type: Dict[str, Any]
class Worker(Thread):
def __init__(self, request_queue: Queue, ignore_exception: bool = False):
Thread.__init__(self)
self.queue = request_queue
self.ignore_exception = ignore_exception
self.response = {} # type: Dict
def run(self):
m = self.queue.get()
try:
self.response = _exec_get_with_retry(m)
except Exception as e:
if not self.ignore_exception:
raise
print(f"Exception occured, still continue: {e}")
self.queue.task_done()
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)
response.raise_for_status()
data = response.json()
return data[0]["id"]
def get_access_token(jwt_token, installation_id):
headers = {
"Authorization": f"Bearer {jwt_token}",
"Accept": "application/vnd.github.v3+json",
}
response = requests.post(
f"https://api.github.com/app/installations/{installation_id}/access_tokens",
headers=headers,
)
response.raise_for_status()
data = response.json()
return data["token"]
def get_key_and_app_from_aws():
secret_name = "clickhouse_github_secret_key"
session = boto3.session.Session()
client = session.client(
service_name="secretsmanager",
)
get_secret_value_response = client.get_secret_value(SecretId=secret_name)
data = json.loads(get_secret_value_response["SecretString"])
return data["clickhouse-app-key"], int(data["clickhouse-app-id"])
def get_token_from_aws():
private_key, app_id = get_key_and_app_from_aws()
payload = {
"iat": int(time.time()) - 60,
"exp": int(time.time()) + (10 * 60),
"iss": app_id,
}
encoded_jwt = jwt.encode(payload, private_key, algorithm="RS256")
installation_id = get_installation_id(encoded_jwt)
return get_access_token(encoded_jwt, installation_id)
def _exec_get_with_retry(url):
for i in range(MAX_RETRY):
try:
response = requests.get(url)
response.raise_for_status()
return response.json()
except Exception as ex:
print("Got exception executing request", ex)
time.sleep(i + 1)
raise Exception("Cannot execute GET request with retries")
WorkflowDescription = namedtuple(
"WorkflowDescription",
["run_id", "head_sha", "status", "rerun_url", "cancel_url", "conclusion"],
)
def get_workflows_description_for_pull_request(
pull_request_event,
) -> List[WorkflowDescription]:
head_repo = pull_request_event["head"]["repo"]["full_name"]
head_branch = pull_request_event["head"]["ref"]
print("PR", pull_request_event["number"], "has head ref", head_branch)
workflows_data = []
request_url = f"{API_URL}/actions/runs?per_page=100"
# Get all workflows for the current branch
for i in range(1, 11):
workflows = _exec_get_with_retry(
f"{request_url}&event=pull_request&branch={head_branch}&page={i}"
)
if not workflows["workflow_runs"]:
break
workflows_data += workflows["workflow_runs"]
if i == 10:
print("Too many workflows found")
if not workflows_data:
print("No workflows found by filter")
return []
print(f"Total workflows for the branch {head_branch} found: {len(workflows_data)}")
DEBUG_INFO["workflows"] = []
workflow_descriptions = []
for workflow in workflows_data:
# Some time workflow["head_repository"]["full_name"] is None
if workflow["head_repository"] is None:
continue
DEBUG_INFO["workflows"].append(
{
"full_name": workflow["head_repository"]["full_name"],
"name": workflow["name"],
"branch": workflow["head_branch"],
}
)
# unfortunately we cannot filter workflows from forks in request to API
# so doing it manually
if (
workflow["head_repository"]["full_name"] == head_repo
and workflow["name"] in NEED_RERUN_OR_CANCELL_WORKFLOWS
):
workflow_descriptions.append(
WorkflowDescription(
run_id=workflow["id"],
head_sha=workflow["head_sha"],
status=workflow["status"],
rerun_url=workflow["rerun_url"],
cancel_url=workflow["cancel_url"],
conclusion=workflow["conclusion"],
)
)
return workflow_descriptions
def get_workflow_description_fallback(pull_request_event) -> List[WorkflowDescription]:
head_repo = pull_request_event["head"]["repo"]["full_name"]
head_branch = pull_request_event["head"]["ref"]
print("Get last 500 workflows from API to search related there")
# Fallback for a case of an already deleted branch and no workflows received
request_url = f"{API_URL}/actions/runs?per_page=100"
q = Queue() # type: Queue
workers = []
workflows_data = []
i = 1
for i in range(1, 6):
q.put(f"{request_url}&page={i}")
worker = Worker(q, True)
worker.start()
workers.append(worker)
for worker in workers:
worker.join()
if not worker.response:
# We ignore get errors, so response can be empty
continue
# Prefilter workflows
workflows_data += [
wf
for wf in worker.response["workflow_runs"]
if wf["head_repository"] is not None
and wf["head_repository"]["full_name"] == head_repo
and wf["head_branch"] == head_branch
and wf["name"] in NEED_RERUN_OR_CANCELL_WORKFLOWS
]
print(f"Total workflows in last 500 actions matches: {len(workflows_data)}")
DEBUG_INFO["workflows"] = [
{
"full_name": wf["head_repository"]["full_name"],
"name": wf["name"],
"branch": wf["head_branch"],
}
for wf in workflows_data
]
workflow_descriptions = [
WorkflowDescription(
run_id=wf["id"],
head_sha=wf["head_sha"],
status=wf["status"],
rerun_url=wf["rerun_url"],
cancel_url=wf["cancel_url"],
conclusion=wf["conclusion"],
)
for wf in workflows_data
]
return workflow_descriptions
def get_workflow_description(workflow_id) -> WorkflowDescription:
workflow = _exec_get_with_retry(API_URL + f"/actions/runs/{workflow_id}")
return WorkflowDescription(
run_id=workflow["id"],
head_sha=workflow["head_sha"],
status=workflow["status"],
rerun_url=workflow["rerun_url"],
cancel_url=workflow["cancel_url"],
conclusion=workflow["conclusion"],
)
def _exec_post_with_retry(url, token):
headers = {"Authorization": f"token {token}"}
for i in range(MAX_RETRY):
try:
response = requests.post(url, headers=headers)
response.raise_for_status()
return response.json()
except Exception as ex:
print("Got exception executing request", ex)
time.sleep(i + 1)
raise Exception("Cannot execute POST request with retry")
def exec_workflow_url(urls_to_cancel, token):
for url in urls_to_cancel:
print("Post for workflow workflow using url", url)
_exec_post_with_retry(url, token)
print("Workflow post finished")
def main(event):
token = get_token_from_aws()
DEBUG_INFO["event_body"] = event["body"]
event_data = json.loads(event["body"])
print("Got event for PR", event_data["number"])
action = event_data["action"]
print("Got action", event_data["action"])
pull_request = event_data["pull_request"]
labels = {label["name"] for label in pull_request["labels"]}
print("PR has labels", labels)
if action == "closed" or "do not test" in labels:
print("PR merged/closed or manually labeled 'do not test' will kill workflows")
workflow_descriptions = get_workflows_description_for_pull_request(pull_request)
workflow_descriptions = (
workflow_descriptions or get_workflow_description_fallback(pull_request)
)
urls_to_cancel = []
for workflow_description in workflow_descriptions:
if (
workflow_description.status != "completed"
and workflow_description.conclusion != "cancelled"
):
urls_to_cancel.append(workflow_description.cancel_url)
print(f"Found {len(urls_to_cancel)} workflows to cancel")
exec_workflow_url(urls_to_cancel, token)
elif action == "synchronize":
print("PR is synchronized, going to stop old actions")
workflow_descriptions = get_workflows_description_for_pull_request(pull_request)
workflow_descriptions = (
workflow_descriptions or get_workflow_description_fallback(pull_request)
)
urls_to_cancel = []
for workflow_description in workflow_descriptions:
if (
workflow_description.status != "completed"
and workflow_description.conclusion != "cancelled"
and workflow_description.head_sha != pull_request["head"]["sha"]
):
urls_to_cancel.append(workflow_description.cancel_url)
print(f"Found {len(urls_to_cancel)} workflows to cancel")
exec_workflow_url(urls_to_cancel, token)
elif action == "labeled" and "can be tested" in labels:
print("PR marked with can be tested label, rerun workflow")
workflow_descriptions = get_workflows_description_for_pull_request(pull_request)
workflow_descriptions = (
workflow_descriptions or get_workflow_description_fallback(pull_request)
)
if not workflow_descriptions:
print("Not found any workflows")
return
sorted_workflows = list(sorted(workflow_descriptions, key=lambda x: x.run_id))
most_recent_workflow = sorted_workflows[-1]
print("Latest workflow", most_recent_workflow)
if (
most_recent_workflow.status != "completed"
and most_recent_workflow.conclusion != "cancelled"
):
print("Latest workflow is not completed, cancelling")
exec_workflow_url([most_recent_workflow.cancel_url], token)
print("Cancelled")
for _ in range(30):
latest_workflow_desc = get_workflow_description(most_recent_workflow.run_id)
print("Checking latest workflow", latest_workflow_desc)
if latest_workflow_desc.status in ("completed", "cancelled"):
print("Finally latest workflow done, going to rerun")
exec_workflow_url([most_recent_workflow.rerun_url], token)
print("Rerun finished, exiting")
break
print("Still have strange status")
time.sleep(3)
else:
print("Nothing to do")
def handler(event, _):
try:
main(event)
finally:
for name, value in DEBUG_INFO.items():
print(f"Value of {name}: ", value)