Too many healthcheck notifications in the journal

[QuLogic]

While looking at the image info for valkey, I see that it supports systemd startup notification natively:

https://hub.docker.com/r/valkey/valkey#systemd-startup-notification-with-podman-quadlet

Would this be better than the health commands you have manually set up?

[QuLogic]

Every 10 seconds, podman logs the health check status; this is filling up my logs for no reason (I should only care if health changes status.)

[jbtrystram]

Neat. Altough the documentation makes it sounds it's only about startup, and not health so we can't remove the healthcheck.
You can add a drop-in limiting the size of the healthcheck log : https://docs.podman.io/en/latest/markdown/podman-systemd.unit.5.html#healthmaxlogsize

[QuLogic]

Neat. Altough the documentation makes it sounds it's only about startup, and not health so we can't remove the healthcheck.

With the native stuff, I get

Status: "Ready to accept connections"

in systemctl status since valkey reports its status during startup.

What benefits do the health checks have? They seem to be purely informational in podman only? I can just run podman healthcheck to manually test it, then.

You can add a drop-in limiting the size of the healthcheck log : https://docs.podman.io/en/latest/markdown/podman-systemd.unit.5.html#healthmaxlogsize

That's not the log I mean; every health check prints this out, which goes to the system journal:

Jul 28 15:59:38 hostname podman[2663660]: 2025-07-28 15:59:38.593685783 -0400 EDT m=+0.085368390 container health_status 28ba2a9772bf8f4df46e3830f98c720866ac516392a1162892dfdcf017b5326c (image=ghcr.io/immich-app/immich-server:v1.135.3, name=immich_server, pod_id=24ecdab65bb1f13d2f4ea0e974dc951b9e521a8fea6e8e29f10c9e6e76116028, health_status=healthy, health_failing_streak=0, health_log=, org.opencontainers.image.created=2025-06-20T20:19:57.294Z, org.opencontainers.image.version=v1.135.3, org.opencontainers.image.revision=6563fa608a5a4414f40264b847e60928916fb986, PODMAN_SYSTEMD_UNIT=immich-server.service, org.opencontainers.image.description=High performance self-hosted photo and video management solution., org.opencontainers.image.licenses=AGPL-3.0, org.opencontainers.image.source=https://github.com/immich-app/immich, org.opencontainers.image.title=immich, org.opencontainers.image.url=https://github.com/immich-app/immich)

and there are 4 of those every 10s (plus 1 if using the hardware-specific ML containers.)

[QuLogic]

That's not the log I mean; every health check prints this out, which goes to the system journal:

Well, I've gone and added a drop-in to set HealthInterval=disable; Notify=false. Hopefully containers/podman#24003 will be fixed soon.