-
Notifications
You must be signed in to change notification settings - Fork 577
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ducky: pin to the latest version #6003
Conversation
7a30a9c
to
bbb9ea2
Compare
The change made |
If this change is making failures more deterministic then that is good, but please let's not merge it until those tests are either fixed or marked |
The wait_for_version_sync was correct for waiting for the configuration to propagate across the cluster, but it was _not_ correct for waiting for the configuration status to be symmetric on all nodes (i.e. for all nodes to know the status of all other nodes) For test cases that query the status via arbitrary notes, a stricter wait is needed. This issue becomes visible once wait_until is improved to avoid spurious extra sleeps, in redpanda-data#6003 Fixes redpanda-data#6010
The wait_for_version_sync was correct for waiting for the configuration to propagate across the cluster, but it was _not_ correct for waiting for the configuration status to be symmetric on all nodes (i.e. for all nodes to know the status of all other nodes) For test cases that query the status via arbitrary notes, a stricter wait is needed. This issue becomes visible once wait_until is improved to avoid spurious extra sleeps, in redpanda-data#6003 Fixes redpanda-data#6010
The wait_for_version_sync was correct for waiting for the configuration to propagate across the cluster, but it was _not_ correct for waiting for the configuration status to be symmetric on all nodes (i.e. for all nodes to know the status of all other nodes) For test cases that query the status via arbitrary notes, a stricter wait is needed. This issue becomes visible once wait_until is improved to avoid spurious extra sleeps, in redpanda-data#6003 Fixes redpanda-data#6010
The wait_for_version_sync was correct for waiting for the configuration to propagate across the cluster, but it was _not_ correct for waiting for the configuration status to be symmetric on all nodes (i.e. for all nodes to know the status of all other nodes) For test cases that query the status via arbitrary notes, a stricter wait is needed. This issue becomes visible once wait_until is improved to avoid spurious extra sleeps, in redpanda-data#6003 Fixes redpanda-data#6010
The wait_for_version_sync was correct for waiting for the configuration to propagate across the cluster, but it was _not_ correct for waiting for the configuration status to be symmetric on all nodes (i.e. for all nodes to know the status of all other nodes) For test cases that query the status via arbitrary notes, a stricter wait is needed. This issue becomes visible once wait_until is improved to avoid spurious extra sleeps, in redpanda-data#6003 Fixes redpanda-data#6010
The wait_for_version_sync was correct for waiting for the configuration to propagate across the cluster, but it was _not_ correct for waiting for the configuration status to be symmetric on all nodes (i.e. for all nodes to know the status of all other nodes) For test cases that query the status via arbitrary notes, a stricter wait is needed. This issue becomes visible once wait_until is improved to avoid spurious extra sleeps, in redpanda-data#6003 Fixes redpanda-data#6010 (cherry picked from commit b7d4f2c)
147c1bd
to
e38344f
Compare
|
PrefixTruncateRecoveryUpgradeTest.test_recover_during_upgrade #5589 |
|
|
|
|
PartitionBalancerTest.test_full_nodes - #5884 |
|
c066a0a
to
0db1bdc
Compare
0db1bdc
to
b9186d0
Compare
b9186d0
to
a7e822d
Compare
|
|
77b6628
to
e7a512c
Compare
The latest version improved performance of wait_until and reduced the full test run by 30%. See details in redpanda-data/ducktape#16
e7a512c
to
b10d1f1
Compare
BLOCKER: PrefixTruncateRecoveryUpgradeTest.test_recover_during_upgrade #5589 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nice
self.logger.exception("%s: something is wrong" % | ||
self.who_am_i(node)) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Should this be conditional on !self._stopping.is_set()
?
Oh. Why? |
I saw that there was a commit updating ducktape here
Would have been nicer to have a comment explaining why this PR was closed though, I was initially confused. @rystsov last time you updated this PR, were there tests that were still unstable with the wait_until change? I guess any tests that were in that state will now be unstable on |
Cover letter
The latest version improved performance of
wait_until
and reduced the full test run by 30%. See details thereFaster
wait_until
dormant flakiness hidden by extra sleep. See this commentBackport Required
UX changes
Not a user visible change
Release notes