-
Notifications
You must be signed in to change notification settings - Fork 1.3k
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
[performance] - TopicOperator capacity test case #10050
Conversation
/packit test --labels performance-topic-operator-capacity |
@strimzi-ci run tests --cluster-type=ocp --cluster-version=4.15 --install-type=bundle --profile=performance --testcase=TopicOperatorPerformance#testCapacity |
|
/packit test --labels performance-topic-operator-capacity |
✔️ Test Summary ✔️TEST_PROFILE: performance |
Signed-off-by: see-quick <[email protected]>
Signed-off-by: see-quick <[email protected]>
Signed-off-by: see-quick <[email protected]>
Signed-off-by: see-quick <[email protected]>
Signed-off-by: see-quick <[email protected]>
Signed-off-by: see-quick <[email protected]>
/packit test --labels performance-topic-operator-capacity |
@strimzi-ci run tests --cluster-type=ocp --cluster-version=4.15 --install-type=bundle --profile=performance --testcase=TopicOperatorPerformance#testCapacity --env=STRIMZI_USE_KRAFT_IN_TESTS=true |
|
@strimzi-ci run tests --cluster-type=ocp --cluster-version=4.15 --install-type=bundle --profile=performance --testcase=TopicOperatorPerformance#testCapacity --env=STRIMZI_USE_KRAFT_IN_TESTS=true |
|
✔️ Test Summary ✔️TEST_PROFILE: performance |
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.
LGTM, just one question - do we have some summary readme or something about the perf tests? I think it might be useful not just for us but also for users.
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.
Thanks for adding these additional performance tests.
A KafkaTopic creation event is much heavier to handle than a configuration change event. I was wondering if we should use a workload with both event types in parallel. With that, I get the best performance with batch.size=100 and linger.ms=10. Wdyt?
Not yet. I can write some overview about them in the |
So as we have in the
Such questions are answered in my findings in the description. This could be potentially documented somewhere (same with User Operator findings).
This is an interesting one and I think it's another use case right? Currently, I am just trying to hit the Note 1: Also, I think it's important to see other Note 2: There seems to be some issue when we hit around 4k in [1] - #10054 |
Ok this makes sense, thanks.
Yes, this is the one I used for my custom tests to simulate the workload of a busy cluster, that should give more general tuning recommendation. If you confirm my finding, then we may even think about changing the default value of linger.ms.
I'll look into that. Thanks for raising. |
Type of change
Description
TLDR; This PR adds a capacity test case to check and understand performance under varying conditions of the Topic Operator.
Findings:
Multi-node (ZK-based)
Multi-node (KRaft-based)
Testing farm ARM:
Testing farm Intel
Checklist