No readiness health check log entry or event when app is pushed with failing readiness check

[beyhan]

Current behavior

When an application is pushed with a readiness-health-check which passes the check a log entry is printed to the application logs like:

   2024-11-19T13:37:59.95+0100 [CELL/0] OUT Downloaded droplet (47.6M)
   2024-11-19T13:37:59.95+0100 [CELL/0] OUT Starting health monitoring of container
   2024-11-19T13:37:59.95+0100 [HEALTH/0] OUT Starting readiness health monitoring of container
   2024-11-19T13:38:00.03+0100 [APP/PROC/WEB/0] OUT Invoking pre-start scripts.
   2024-11-19T13:38:00.24+0100 [APP/PROC/WEB/0] OUT Invoking start command.
   2024-11-19T13:38:00.63+0100 [APP/PROC/WEB/0] OUT > [email protected] start
   2024-11-19T13:38:00.63+0100 [APP/PROC/WEB/0] OUT > node server.js
   2024-11-19T13:38:02.45+0100 [CELL/0] OUT Container became healthy
   2024-11-19T13:38:02.46+0100 [HEALTH/0] OUT Container passed the readiness health check. Container marked ready and added to route pool.

When the readiness-health-check can't pass the check or is misconfigured there is no log entry after the push. The logs look like:

   2024-11-19T13:39:34.21+0100 [CELL/0] OUT Downloaded droplet (47.6M)
   2024-11-19T13:39:34.21+0100 [HEALTH/0] OUT Starting readiness health monitoring of container
   2024-11-19T13:39:34.21+0100 [CELL/0] OUT Starting health monitoring of container
   2024-11-19T13:39:34.29+0100 [APP/PROC/WEB/0] OUT Invoking pre-start scripts.
   2024-11-19T13:39:34.49+0100 [APP/PROC/WEB/0] OUT Invoking start command.
   2024-11-19T13:39:34.90+0100 [APP/PROC/WEB/0] OUT > [email protected] start
   2024-11-19T13:39:34.90+0100 [APP/PROC/WEB/0] OUT > node server.js
   2024-11-19T13:39:36.00+0100 [CELL/SSHD/0] OUT Exit status 0
   2024-11-19T13:39:36.49+0100 [CELL/0] OUT Container became healthy

The result is that the application route is not registered and you are wondering why. The forked cf-sample-app-nodejs could be used to reproduce this.

Desired behavior

It will be easier to debug the issue in case there is a log entry or event from the readiness-health-check which states that the readiness check didn't pass.

Affected Version

since the readiness check exists

[maxmoehl]

See also cloudfoundry/cli#3350.