fix(main): add ci-canary.yml + disable pr-auto-scan schedule on main #1638

Merged
xiaoxia merged 1 commits from infra/fix-main-workflows into main 2026-09-03 10:57:31 +08:00
2 changed files with 52 additions and 1 deletions
+51
View File
@@ -0,0 +1,51 @@
name: CI Canary Check
on:
schedule:
- cron: '*/30 * * * *'
workflow_dispatch:
jobs:
canary:
runs-on: ci-l2
timeout-minutes: 10
steps:
- name: Canary (runner -> docker -> network -> gitea)
run: |
set -e
echo "== runner/container basic =="
date; hostname; whoami
echo "== gitea api reachability =="
code=$(curl -s -o /tmp/v.json -w '%{http_code}' -m 15 "$GITHUB_API_URL/version")
echo "gitea api http_code=$code"
[ "$code" = "200" ] || { echo "::error::Gitea API unreachable, http_code=$code"; exit 1; }
cat /tmp/v.json; echo
echo "== external egress =="
ext=$(curl -s -o /dev/null -w '%{http_code}' -m 15 https://www.baidu.com || echo 000)
echo "external http_code=$ext"
echo "== gitea domain resolves NOT to loopback =="
set -o pipefail
ip=$(getent hosts git.xiaoxiajianji.com | awk '{print $1}' | head -1)
echo "git.xiaoxiajianji.com -> $ip"
if [ -z "$ip" ]; then
echo "::error::DNS resolution failed, git.xiaoxiajianji.com unresolvable"; exit 1
fi
if [ "$ip" = "127.0.0.1" ] || [ "$ip" = "::1" ]; then
echo "::error::Gitea domain resolves to loopback inside job container (hosts/DNS leak)"; exit 1
fi
echo "CANARY OK"
- name: Notify failure
if: failure()
env:
CI_NOTIFY_WEBHOOK: ${{ secrets.CI_NOTIFY_WEBHOOK }}
run: |
set +e
if [ -n "$CI_NOTIFY_WEBHOOK" ]; then
MSG="🚨 CI 金丝雀失败:runner->docker->网络->Gitea 链路异常,时间 $(date '+%Y-%m-%d %H:%M:%S'),请立即检查构建服务器"
python3 - "$CI_NOTIFY_WEBHOOK" "$MSG" <<'PY'
import json,sys,urllib.request
hook,msg=sys.argv[1],sys.argv[2]
data=json.dumps({"msg_type":"text","content":{"text":msg}}).encode()
urllib.request.urlopen(urllib.request.Request(hook,data=data,headers={"Content-Type":"application/json"}),timeout=10)
PY
fi
exit 0
+1 -1
View File
@@ -3,7 +3,7 @@ name: PR Auto Scan
# 作为短作业模式的兜底,防止事件驱动遗漏
on:
schedule:
- cron: "*/15 * * * *" # 每10分钟扫描一次(脚本自带240s墙钟上限,降频减负)
# - cron: "*/15 * * * *" # DISABLED: was causing failure spam on main (2026-09-03) # 每10分钟扫描一次(脚本自带240s墙钟上限,降频减负)
workflow_dispatch:
permissions: