changefeedccl: fix test flake by waiting for rangefeed start #152689
Conversation
|
This change is |
| {`regularUser`, false}, | ||
| {`adminUser`, true}, | ||
| {`viewClusterMetadataUser`, true}, | ||
| {`feedCreator`, false}, |
There was a problem hiding this comment.
What's the point of reordering the cases? I ask because changes like this have the cost of adding another commit to the git blame history (for these lines) so there needs to be some kind of benefit that outweighs that
There was a problem hiding this comment.
It's possible that the feed creator case was possible not because they lacked the permissions but because no one was able to see the table since the rangefeed hadn't started up yet (this is the reason the admin user case was failing). Putting the true cases before the false cases ensures that the tables are visible to someone when the feed creator can't see them. Do you think it's worth including? I don't feel strongly
There was a problem hiding this comment.
IMO it's not really worth switching since with either order, the subtest that would fail would still be adminUser right? We might actually want to randomize the order by using a map instead of a slice so there's more variance, up to you if you want to make that change
Previously, this test would flake because a table that we were asserting the admin user could see hadn't yet become visible. Now we assert the changefeed and rangefeed have started by waiting for a highwater mark. Epic: none Fixes: cockroachdb#152060 Release note: None
aerfrei
force-pushed
the
group-test-failure
branch
from
eed0d7a to
9f48bed
Compare
|
bors r=andyyang890 |
celeste-cockroachdb
bot
added
v25.4.0-prerelease
and removed
target-release-25.4.0
labels
|
blathers backport 24.3 25.2 25.3 |
|
Based on the specified backports for this PR, I applied new labels to the following linked issue(s). Please adjust the labels as needed to match the branches actually affected by the issue(s), including adding any known older branches. Issue #152060: branch-release-24.3, branch-release-25.2, branch-release-25.3. 🦉 Hoot! I am a Blathers, a bot for CockroachDB. My owner is dev-inf. |
3 participants
Previously, this test would flake because a table that we were asserting the admin user could see hadn't yet become visible. Now we assert the changefeed and rangefeed have started by waiting for a highwater mark.
Epic: none
Fixes: #152060
Release note: None