Fix raw block volume resizing for EBS

[gnufied]

We need to fix raw block volume resizing (NodeExpandVolume) to not resize the volume if given volume_path in resize request is a block device and not a mounted file system.

[gnufied]

@wongma7 will you have time to fix this?

[wongma7]

/assign
Yes, will have a look

[leakingtapan]

It's more explicit to return NodeExpansionRequired=false during ControllerExpandVolume call when the controller block volume expansion is requested. But I also realized that volume capability is not passed into ControllerExpandVolume which makes controller harder to know if is this request for block or filesystem.

[wongma7]

Yeah, @leakingtapan let's help review container-storage-interface/spec#381 & container-storage-interface/spec#380 to get volume capability passed to ControllerExpandVolume

[leakingtapan]

We need to fix raw block volume resizing (NodeExpandVolume) to not resize the volume if given volume_path in resize request is a block device and not a mounted file system.

I can understand not do node resize for block volume. But what does not a mounted file system mean?

@gnufied

[gnufied]

the volume_path specified in NodeExpandVolumeRequest field could be a mount point (or a directory with file system) or it could be a block device (such as /dev/xvdba). We should be able to differentiate between two and only resize the volume if volume_path is a mounted file system and not a raw block device. I think skipping fs resize if mount_path is a raw block device should do the job.

[leakingtapan]

/assign @sreis

[k8s-ci-robot]

@leakingtapan: GitHub didn't allow me to assign the following users: sreis.

Note that only kubernetes-sigs members, repo collaborators and people who have commented on this issue/PR can be assigned. Additionally, issues/PRs can only have 10 assignees at the same time.
For more information please see the contributor guide

In response to this:

/assign @sreis

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.

[zhucan]

@gnufied what's version with the kubernetes support NodeExpandVolume for raw block volume? v1.14&v1.15 doesn't support it?

[leakingtapan]

@zhucan There is not feature flag for raw block volume expansion, it fall under ExpandCSIVolumes flag, which is beta since 1.16

[zhucan]

@leakingtapan Yes, In my kubernetes cluster(v1.15.0), The expansion(NodeExpandVolume) for raw block not be called.

[leakingtapan]

/kind bug

[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

[frittentheke]

/remove-lifecycle stale

[frittentheke]

@leakingtapan I took the liberty to "unstale" this ... should bugs, especially those you already triaged and marked as such, ever go stale?

[leakingtapan]

Thx for doing this. It looks fejta-bot isn’t smart enough to figure out if the issue is triaged or not... and I don’t have more time to work on the fix

[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

[frittentheke]

/remove-lifecycle stale

[ayberk]

/lifecycle frozen

[AndyXiangLi]

/assign
looking into this