roachtest: improve the cluster-init roachtest
#69940
Conversation
Release justification: non-production code changes Release note: None
See discussion here; cockroachdb#67105 (comment) Release justification: non-production code changes Release note: None
|
This change is |
knz
requested review from
a team and
stevendanna
and removed request for
a team
Release justification: test-only change Release note: None
Give it a progress bar and a rudimentary context cancellation policy. Release justification: testing-only changes Release note: None
There was a problem hiding this comment.
I just pushed two commits, one of which makes this test use c.Start (the other is a minor cleanup). I verified that the test works using
TEST=acceptance/cluster-init LOCAL=y SHORT=y COUNT=1 ./pkg/cmd/roachtest/roachstress.sh
As far as I'm concerned the result is ready for the merge, but since I wrote it perhaps someone else on this review (server or test-eng) wants to sign off officially.
I would also suggest closing #67105 (tests should not manage their own CRDB processes, and when they do it's fair that they would wipe the cluster before termination) and also #66516 (since the test is now idiomatic).
Reviewed 1 of 1 files at r1.
Reviewable status:complete! 0 of 0 LGTMs obtained (waiting on @stevendanna and @tbg)
There was a problem hiding this comment.
Reviewed 1 of 1 files at r3, 1 of 1 files at r4, all commit messages.
Reviewable status:
|
bors r=knz |
|
Build succeeded: |
This is an edulcorated backport of cockroachdb#69940, which cannot be backported as-is due to severe roachtest infra changes. Release justification: test-only changes Release note: None
Fixes #67105
Fixes #66516.
Fixes #69949.
Fixes #70003
Release justification: non-production code changes
Release note: None