chia7712 commented on code in PR #16499:
URL: https://github.com/apache/kafka/pull/16499#discussion_r1679937441
##########
core/src/test/java/kafka/test/junit/ClusterTestExtensions.java:
##########
@@ -119,7 +138,21 @@ public Stream<TestTemplateInvocationContext>
provideTestTemplateInvocationContex
return generatedContexts.stream();
}
+ @Override
+ public void beforeEach(ExtensionContext context) {
+ detectThreadLeak = DetectThreadLeak.of(thread ->
+ SKIPPED_THREAD_PREFIX.stream().noneMatch(prefix ->
thread.getName().startsWith(prefix)));
+ }
+ @Override
+ public void afterEach(ExtensionContext context) throws
InterruptedException {
+ if (detectThreadLeak == null) {
+ return;
+ }
+ TestUtils.waitForCondition(() ->
detectThreadLeak.newThreads().isEmpty(),
Review Comment:
> I think this waitForCondition will cover up certain thread leaks where the
test started shutdown of the thread, but didn't call join() to ensure the
thread was completely stopped. I think these are leaks that would be good to
find and fix.
that is a great idea, but I guess the related errors will spring up. For
example, I apply the idea on my local, and Zk SessionTracker show up. We can
fix them one by one to avoid random failed tests, but that will make this PR
pending for a while.
Maybe we can open a jira as follow-up to fix ignored "join()", and
eventually we can remove the waiting from thread leak detection.
@gharris1727 WDYT?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
