libpod: cleanup default cache on system reset

[giuseppe]

Closes: #22825

None

[Luap99]

Do you know if there a reason that the blob cache is not located on the actual graphroot but rather hard coded to ~/.local/share/containers/cache? I am asking because this means that running e2e system reset tests (which should run isolated) will now nuke my local blob info cache.
I guess it will not be big deal in practice as it is just a cache but it still seems rather unexpected to me.

[giuseppe]

Do you know if there a reason that the blob cache is not located on the actual graphroot but rather hard coded to ~/.local/share/containers/cache? I am asking because this means that running e2e system reset tests (which should run isolated) will now nuke my local blob info cache.
I guess it will not be big deal in practice as it is just a cache but it still seems rather unexpected to me.

no idea why we picked that directory.

@mtrmac do you think we should change it to be under storage?

[mtrmac]

The reason is that it has no direct relationship with the c/storage graph root. It is used, and beneficial, e.g. also for skopeo copy docker://… docker://…

[Luap99]

The reason is that it has no direct relationship with the c/storage graph root. It is used, and beneficial, e.g. also for skopeo copy docker://… docker://…

Should podman system reset then delete it or not in your opinion?

[mtrmac]

I don’t have a strong opinion.

You’re right to point out that the relationship is not that tight.

OTOH, from time to time, the blob info cache is key in behavior decisions, and users who are unaware of it find it very hard to reproduce the unwanted behavior they are trying to report. In that sense, having podman system reset reset the cache’s state might help those users.

[Luap99]

Ack thanks, as it is just a cache and it is not like many user running podman system reset with non default location anyway I think it is fine then

[Luap99]

LGTM

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: giuseppe, Luap99

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [Luap99,giuseppe]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[mtrmac]

I do agree it’s not ideal if Podman tests affect things outside of the test perimeter. I’m tempted to say that should be changed in the tests, maybe by changing XDG_DATA_HOME. But that feels like a big and risky hammer.

[mheon]

/lgtm

[TomSweeneyRedHat]

Addresses: https://issues.redhat.com/browse/RHEL-40811 and https://issues.redhat.com/browse/RHEL-40814