vendor: bump containers/storage to v1.24.0

[kolyshkin]

What type of PR is this?

/kind bug

What this PR does / why we need it:

This is mostly to have containers/storage#757
which should fix occasional CI failures (and probably real failures, too).

Which issue(s) this PR fixes:

time="2020-11-02T19:00:03Z" level=fatal msg="creating container: rpc error: code = Unknown desc = error creating an ID-mapped copy of layer "01d6f430ddfdf1203b505fbb03c6479802a3366609d5e0231a30f85f412f873d": exit status 1: error during chown: storage-chown-by-maps: chown("lib/apk/db/scripts.tar"): lchown lib/apk/db/scripts.tar: interrupted system call"

periodically caught by CI

Special notes for your reviewer:

This is currently blocked by #4371

Backports:

• 1.20: no need
• 1.19: [1.19] bump containers/storage to 1.20.5 to fix EINTR #4340
• 1.18: [1.18] vendor: bump containers/storage to 1.20.5 to fix EINTR #4341

Does this PR introduce a user-facing change?

Fix occasional "chown: interrupted system call" error on container creation.

[haircommander]

/approve

[kolyshkin]

Backports:

• 1.20: no need
• 1.19: [1.19] vendor: bump containers/storage to v1.20.5 #4337 (need to fix containers/storage version to 1.12.x once released)

[codecov]

Codecov Report

Merging #4334 (359c60f) into master (a004cba) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #4334   +/-   ##
=======================================
  Coverage   39.29%   39.29%           
=======================================
  Files         112      112           
  Lines        8835     8835           
=======================================
  Hits         3472     3472           
  Misses       4977     4977           
  Partials      386      386           

[mrunalp]

/lgtm

[mrunalp]

/test e2e-aws

[kolyshkin]

Failure in critest related to AppArmor:

Summarizing 3 Failures:

[Fail] [k8s.io] AppArmor runtime should support apparmor [It] should enforce a permissive profile 
/home/sascha/go/src/github.com/kubernetes-sigs/cri-tools/pkg/validate/apparmor_linux.go:143

[Fail] [k8s.io] AppArmor runtime should support apparmor [It] should enforce a permissive profile 
/home/sascha/go/src/github.com/kubernetes-sigs/cri-tools/pkg/validate/apparmor_linux.go:143

[Fail] [k8s.io] AppArmor runtime should support apparmor [It] should enforce a permissive profile 
/home/sascha/go/src/github.com/kubernetes-sigs/cri-tools/pkg/validate/apparmor_linux.go:143

I have audited the logs (https://circleci.com/api/v1.1/project/github/cri-o/cri-o/105877/output/107/0?file=true&allocation-id=5fa1278c1be5695c9ebd99d7-0-build%2F7DFD59DA) but it's not clear what happened (other than the touch foo returned 1 while it is expected to return 0).

[kolyshkin]

/retest

[mrunalp]

/test e2e-aws

[haircommander]

/retest

[openshift-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: haircommander, kolyshkin, mrunalp

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 [haircommander,kolyshkin,mrunalp]

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

[kolyshkin]

rebased on top of current master; maybe this will help ci :-\

[kolyshkin]

Same apparmor failure as earlier. Most probably caused by containers/storage#743 as we touch the file in the root directory. The fix (in critest) is easy but I'm not sure the new default won't bring any other breakage.

[saschagrunert]

Same apparmor failure as earlier. Most probably caused by containers/storage#743 as we touch the file in the root directory. The fix (in critest) is easy but I'm not sure the new default won't bring any other breakage.

I have the same issue in this PR: #4358

[kolyshkin]

closed by mistake

[kolyshkin]

I have the same issue in this PR: #4358

@saschagrunert this is probably because your PR also bumps containers/storage to 1.23.10-dev-something, which includes containers/storage#743. Let's see if kubernetes-sigs/cri-tools#682 helps (testing via #4366)

[kolyshkin]

OK, due to containers/storage#743 this needs kubernetes-sigs/cri-tools#682 and a new critools release.

[saschagrunert]

OK, due to containers/storage#743 this needs kubernetes-sigs/cri-tools#682 and a new critools release.

Hm, but isn't this a regression in containers/storage? My plan was to cut a new cri-tools release with k8s 1.20.0 but now this seems kind of urgent. :-/

[kolyshkin]

Hm, but isn't this a regression in containers/storage?

It's hard to say whether it's a regression or not. Looks more like a fortifying measure.

My plan was to cut a new cri-tools release with k8s 1.20.0 but now this seems kind of urgent. :-/

Alternatively, we can use critools master as I did in #4366. One side benefit is we'll be able to catch bugs like kubernetes-sigs/cri-tools#683 earlier.

@mrunalp @haircommander @saschagrunert WDYT?

[kolyshkin]

Alternatively, we can use critools master as I did in #4366. One side benefit is we'll be able to catch bugs like kubernetes-sigs/cri-tools#683 earlier.

See #4371

[saschagrunert]

Alternatively, we can use critools master as I did in #4366. One side benefit is we'll be able to catch bugs like kubernetes-sigs/cri-tools#683 earlier.

I'm in favor of switching to the master for now. I will cut a new cri-tools release together with Kubernetes v1.20.0 in any case.

[kolyshkin]

Rebased on top of recently merged #4371 to fix critest woes.

Assuming the CI is green (except e2e-aws which is broken globally atm), this one should be good to go.

[openshift-merge-robot]

@kolyshkin: The following test failed, say /retest to rerun all failed tests:

Test name	Commit	Details	Rerun command
ci/prow/e2e-aws	359c60f	link	/test e2e-aws

Full PR test history. Your PR dashboard.

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. I understand the commands that are listed here.

[kolyshkin]

failure in e2e_rhel ([k8s.io] [sig-node] Pods Extended [k8s.io] Pod Container Status should never report success for a pending container) is known (kubernetes/kubernetes#96565)

[kolyshkin]

/test e2e_rhel

[fidencio]

/lgtm