Skip to content

Workaround criu re-linking output in system test#13958

Merged
openshift-merge-robot merged 1 commit intocontainers:mainfrom
cevich:fix_system_criu_relink
Apr 22, 2022
Merged

Workaround criu re-linking output in system test#13958
openshift-merge-robot merged 1 commit intocontainers:mainfrom
cevich:fix_system_criu_relink

Conversation

@cevich
Copy link
Copy Markdown
Member

@cevich cevich commented Apr 21, 2022

When run on an F36 host using netavark/aardvark-dns, for whatever
underlying reason most checkpoint/restore tests are emitting an error
similar to:

criu: Symbol __rseq_offset' has different size in shared object,
consider re-linking`

This extraneous output is causing the basic checkpoint system test to
fail. Since, all other testing of checkpoint/restore feature is
passing (also with the extraneous message) loosen the system test
sensitivity to match.

Signed-off-by: Chris Evich cevich@redhat.com

@edsantiago
Copy link
Copy Markdown
Member

Ref: checkpoint-restore/criu#1706

@edsantiago
Copy link
Copy Markdown
Member

This is nauseating and offensive (not your fault - I mean, the need for this workaround). I'm really sorry, but could I ask you to resubmit with a loud FIXME comment explaining the reason for the .*, linking to the criu bug, and requesting future maintainers to remove the .* as soon as a fixed criu makes its way into production?

@cevich
Copy link
Copy Markdown
Member Author

cevich commented Apr 21, 2022

Oh thanks for looking up the underlying issue. I can add a FIXME but they tend to be ignored more than dust-mite vacuum-protest rallies. Especially when the problem fixes itself when new VM images come along. I wish we could make it revert automatically, (for example) after a specific date 🤔

When run on an F36 host using netavark/aardvark-dns, for whatever
underlying reason most checkpoint/restore tests are emitting an error
similar to:

`criu: Symbol `__rseq_offset' has different size in shared object,
consider re-linking`

This extraneous output is causing the basic checkpoint system test to
fail.  Since, all other testing of checkpoint/restore feature is
passing (also with the extraneous message) loosen the system test
sensitivity to match.

Signed-off-by: Chris Evich <cevich@redhat.com>
@cevich cevich force-pushed the fix_system_criu_relink branch from 71035bc to 93ecafc Compare April 21, 2022 21:53
@edsantiago
Copy link
Copy Markdown
Member

I can add a FIXME but they tend to be ignored

They're actually not - I have a script that I run periodically to check for them. (It prefers FIXMEs with actual issue numbers which it can cross-check against github, and we don't have that here, but I'll still find it one day. Without a FIXME, I'll never see it again).

@cevich
Copy link
Copy Markdown
Member Author

cevich commented Apr 21, 2022

I have a script

Of course you do! 🤣

@TomSweeneyRedHat
Copy link
Copy Markdown
Member

LGTM

@edsantiago
Copy link
Copy Markdown
Member

/lgtm

thanks for the comment!

@openshift-ci openshift-ci bot added the lgtm Indicates that a PR is ready to be merged. label Apr 21, 2022
@Luap99
Copy link
Copy Markdown
Member

Luap99 commented Apr 22, 2022

/approve

@openshift-ci
Copy link
Copy Markdown
Contributor

openshift-ci bot commented Apr 22, 2022

[APPROVALNOTIFIER] This PR is APPROVED

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

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

The pull request process is described here

Details Needs approval from an approver in each of these files:

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

@openshift-ci openshift-ci bot added the approved Indicates a PR has been approved by an approver from all required OWNERS files. label Apr 22, 2022
@openshift-merge-robot openshift-merge-robot merged commit 2627345 into containers:main Apr 22, 2022
@mihalicyn
Copy link
Copy Markdown

@edsantiago thanks for reporting!

It looks related to this change in Glibc:
https://sourceware.org/pipermail/libc-alpha/2022-February/136024.html

You have CRIU built on the system with one Glibc version but run it on the system with a different version. Is it a common scenario for you? If yes, then we need to rework our approach by getting the __rseq_offset offset value using dlopen.

@rst0git
Copy link
Copy Markdown
Contributor

rst0git commented Apr 23, 2022

You have CRIU built on the system with one Glibc version but run it on the system with a different version. Is it a common scenario for you?

@mihalicyn I think in this case we have CRIU installed from a package and the CRIU package is not automatically rebuild when the glibc package is updated.

@fweimer-rh
Copy link
Copy Markdown

I should have checked and rebuilt criu in Fedora after changing the symbol size in glibc.

We can't easily fix this right now because Fedora 36 is in final freeze before the release. I can try to get a freeze exception, but it's probably easier to wait a few days until the release if that's okay.

Just to repeat, this a Fedora-specific problem which will go away once the criu package is rebuilt (without any source code changes).

@mihalicyn
Copy link
Copy Markdown

@rst0git @fweimer-rh Thanks, guys. Please, ping me If I need to rework or fix something :)

@cevich cevich deleted the fix_system_criu_relink branch April 18, 2023 14:47
@github-actions github-actions bot added the locked - please file new issue/PR Assist humans wanting to comment on an old issue or PR with locked comments. label Aug 30, 2023
@github-actions github-actions bot locked as resolved and limited conversation to collaborators Aug 30, 2023
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

approved Indicates a PR has been approved by an approver from all required OWNERS files. lgtm Indicates that a PR is ready to be merged. locked - please file new issue/PR Assist humans wanting to comment on an old issue or PR with locked comments.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

8 participants