Debug logs, common CI/CD failures, strict shell mode, retries, and faster pipeline diagnosis.
Use platform features and shell diagnostics to isolate broken steps.
Turn on runner debug logs for detailed troubleshooting.
Set repository or organization secret:
ACTIONS_STEP_DEBUG=trueGitHub Actions can emit extra debug output when debug logging is enabled.
Review the full job log and artifacts from failed jobs.
Open the failed job in GitLab.
Expand collapsed sections.
Download artifacts and reports if available.Keep scripts verbose enough to explain what is happening, but avoid leaking secrets into logs.
Retry a failed Jenkins pipeline with small script adjustments.
Open the failed pipeline run.
Use Replay if your Jenkins setup permits it.
Test the fix before committing it back to Jenkinsfile.Replay can shorten iteration during pipeline development, but committed fixes should remain the source of truth.
Patterns that commonly break builds or deploys.
Fail early when credentials or roles are unavailable.
test -n "$API_TOKEN" || { echo "API_TOKEN is required"; exit 1; }Early validation produces clearer failures than letting a deploy break halfway through.
set -euo pipefailStrict mode is one of the simplest ways to eliminate silent shell-script failures in CI.
Wrap downloads or package installs with bounded retries.
for i in 1 2 3; do
npm ci && break
sleep 5
doneTransient network failures are common in CI. Retry carefully, but do not hide deterministic failures.
Keep only the latest run active for a branch or PR.
concurrency:
group: ci-${{ github.ref }}
cancel-in-progress: trueCanceling superseded runs keeps queues healthy and reduces waste.