Add email health command

sonikagutharequested to merge
phoenix/gsoc/feature/email-health-check into phoenix/gsoc/feature/email-rate-limiting

Summary

  • Add bin/cake email_health with six checks: Redis, worker heartbeat, queue backlog, Postfix reachability, recent failure rate, mail config
  • Hard failures exit 1; warnings exit 0
  • Add unit tests using the email-worker TestableCommand + fake Redis pattern

Check behavior

Check PASS FAIL
Redis connectivity PING/PONG OK and cdli:queue:email accessible Redis unreachable or queue key inaccessible
Worker heartbeat email:worker:heartbeat age ≤ 2× loop interval Missing, invalid, or stale heartbeat
Queue backlog pending+processing ≤ threshold DB query error
Postfix reachability TCP to host:port succeeds Connection refused / timeout / DNS failure
Recent failure rate failed jobs in last hour ≤ threshold DB query error
Mail config validity CDLI_POSTFIX_SMTP_SERVER / USERNAME / PASSWORD set; port valid if present Missing/empty vars or invalid port

Notes

  • Env knobs: EMAIL_HEALTH_BACKLOG_THRESHOLD, EMAIL_HEALTH_FAILURE_THRESHOLD, EMAIL_WORKER_LOOP_INTERVAL (default 6), EMAIL_HEALTH_POSTFIX_HOST/PORT (default postfix:25)

Test plan

  • docker exec -w /srv/app/cake -e CDLI_POSTFIX_SMTP_SERVER=... -e CDLI_POSTFIX_SMTP_USERNAME=... -e CDLI_POSTFIX_SMTP_PASSWORD=... -e CDLI_POSTFIX_SMTP_PORT=587 dev_cake_composer bin/cake email_health
  • Confirm Redis / backlog / Postfix / failure-rate / mail-config PASS
  • Confirm heartbeat FAIL until Redis key is set (or worker heartbeat is added)
  • docker exec -w /srv/app/cake dev_cake_composer vendor/bin/phpunit tests/TestCase/Command/EmailHealthCommandTest.php

Merge request reports