Does it make sense to have an option to clean PV before discovering it as a new PV?

[cofyc]

Current lifecycle:

- raw disks -> mount points -> PV (available) -> PV (bound) -> PV (released)
                               ^                                     |   
                               | <---------- clean <-----------------|                                         

New lifecycle (optional):

- raw disks -> mount points -> [clean] -> PV (available) -> PV (bound) -> PV (released)
                                          ^                                     |   
                                          | <---------- clean <-----------------|

[cofyc]

use cases:

• if we clean the volume before provisioning it as a new PV into Kubernetes, it will be safe to delete the PV objects and don't worry about the data will be leaked. (auto remove the pv with hardware problems #126 (comment))

[yuyulei]

another case:

• I use cloud storage as local pv, after pvc&pod deleted, and cleanup pv successfully. But actually, cloud storage not works as expected. Then provisioner creates the pv with same name and path. Later another pod uses the pv, it finds some data not belong to itself.

so I think it really makes sense, at least, check or log if there are some data before pv available.

[msau42]

Yes I think this is a good idea, although this will cause the recycling flow to double for very large disks. We can add it as a new option

/help
/kind feature

[k8s-ci-robot]

@msau42:
This request has been marked as needing help from a contributor.

Please ensure the request meets the requirements listed here.

If this request no longer meets these requirements, the label can be removed
by commenting with the /remove-help command.

In response to this:

Yes I think this is a good idea, although this will cause the recycling flow to double for very large disks. We can add it as a new option

/help
/kind feature

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[cofyc]

/remove-lifecycle stale

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[cofyc]

/remove-lifecycle stale