Page MenuHomeFreeBSD

makefs: tests: Double the timeout of ZFS compression test
ClosedPublic

Authored by jlduran on Wed, Dec 31, 1:37 AM.
Tags
None
Referenced Files
Unknown Object (File)
Fri, Jan 23, 12:52 PM
Unknown Object (File)
Fri, Jan 23, 11:49 AM
Unknown Object (File)
Wed, Jan 21, 9:34 AM
Unknown Object (File)
Tue, Jan 20, 2:04 AM
Unknown Object (File)
Mon, Jan 19, 11:23 PM
Unknown Object (File)
Sun, Jan 18, 5:48 AM
Unknown Object (File)
Fri, Jan 16, 4:22 AM
Unknown Object (File)
Fri, Jan 9, 8:58 PM
Subscribers

Details

Summary

The test makefs_zfs_tests:compression is timing out on ci.freebsd.org.
Double the default timeout to allow it to finish.

While here, check if the file exists before cleaning up, otherwise, cat
may fail.

Reported by: Jenkins

Diff Detail

Repository
rG FreeBSD src repository
Lint
Lint Not Applicable
Unit
Tests Not Applicable

Event Timeline

asomers added a subscriber: asomers.

This looks fine. But as the default timeout is already 300s, I suggest increasing it even more. If it were me, I would double it.

This revision is now accepted and ready to land.Wed, Dec 31, 3:30 PM

This looks fine. But as the default timeout is already 300s, I suggest increasing it even more. If it were me, I would double it.

Thank you, I'll double the amount of time (600s) upon commit.

This comment has been deleted.
usr.sbin/makefs/tests/makefs_zfs_tests.sh
137–138

Is a bug filed for this issue? Have we been close to hitting the default limit for a while? This failure might hint at a problem with ZFS on main if the test hasn't been around for a while.

jlduran retitled this revision from makefs: tests: Extend the timeout of ZFS compression test to makefs: tests: Double the timeout of ZFS compression test.Fri, Jan 2, 8:23 PM
jlduran edited the summary of this revision. (Show Details)
usr.sbin/makefs/tests/makefs_zfs_tests.sh
137–138

Is a bug filed for this issue?

I did not see any bug filed after a quick search.

Have we been close to hitting the default limit for a while?

I think I understand what you mean, yes the test sometimes pass:
~220s (for example), sometimes barely making the cutoff.
The tests occasionally timeout only when ran on an emulated architecture, on amd64, they finish perfectly on time. No drastic increase in the time has been historically seen, as to presume a regression in ZFS itself.