Skip to main content

Health and monitoring

Jiandu provides health endpoints for orchestration and monitoring. Use them as layers:

  • a liveness check answers whether the process can respond;
  • a readiness check answers whether it is ready to receive normal traffic;
  • a real browser or API probe confirms the public HTTPS route and an important user journey.

Use the exact paths described by the release's API reference. Avoid putting credentials or document content in health-probe logs.

What to alert on​

  • repeated readiness failure;
  • restart loops or an installation lock error;
  • increasing database or storage errors;
  • failed capture/email imports;
  • low filesystem, bucket, or database capacity;
  • backup failures or missed restore drills;
  • certificate expiry or public-origin failures.

What to test periodically​

A green process check does not prove documents are usable. From a browser through the same HTTPS route as members, sign in with a test account, search for a known sample, and open its original. Keep this sample non-sensitive and stable.

When an error appears, preserve the exact message and the surrounding timestamp. Search it in the error message guide before changing the installation.