podman restart container - healthcheck transient timer

[fornitlakon]

On 'podman restart container' the healthcheck transient timer is not cleared resulting with an ERROR
Failed to start transient timer unit: Unit a454fc67b6569accee4d2e65b9f9e93c34284caa5e092dce05206267a674af81.timer already exists
i.e
[root@localhost system]# podman restart redis ERRO[0000] Failed to start transient timer unit: Unit a454fc67b6569accee4d2e65b9f9e93c34284caa5e092dce05206267a674af81.timer already exists.

systemd files
/run/systemd/system/a454fc67b6569accee4d2e65b9f9e93c34284caa5e092dce05206267a674af81.service.d
/run/systemd/system/a454fc67b6569accee4d2e65b9f9e93c34284caa5e092dce05206267a674af81.timer.d

Are not cleaned on restart resulting with this ERROR

podman restart should recreate the systemd healthcheck service/timer to prevent this ERROR

Restart behavior is Ok, Container is restarted correctly
But how can the container be restarted without this ERROR ?

[mheon]

@baude You were looking at this one earlier this week, I recall?

[baude]

You want -> #12697

[fornitlakon]

Look like #12697 is a different case.
Here the healthcheck does not fail, it is successful and on container restart it is not cleaned

Maybe this is relevant here :
https://serverfault.com/questions/965724/make-systemd-run-fail-gracefully-if-the-unit-name-already-exists

And adding --collect to createTimer() in
./libpod/healthcheck_linux.go

[fornitlakon]

@baude @mheon Can you take a second look ?
Look like #12697 is a different case.
Here the healthcheck does not fail, it is successful and on container restart it is not cleaned

Maybe this is relevant here :
https://serverfault.com/questions/965724/make-systemd-run-fail-gracefully-if-the-unit-name-already-exists

And adding --collect to createTimer() in
./libpod/healthcheck_linux.go