[
https://issues.apache.org/jira/browse/CASSANDRA-16681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17613928#comment-17613928
]
Piotr Kolaczkowski edited comment on CASSANDRA-16681 at 10/7/22 7:41 AM:
-------------------------------------------------------------------------
The `testPoolAllocateWithRecyclePartially` fails quite reliably if I exclude
the last patch:
{noformat}
INFO [main] 2022-10-07 09:31:39,811 LongBufferPoolTest.java:329 - 2022/10/07
09:31:39 - testing 16 threads for 2m
INFO [main] 2022-10-07 09:31:39,811 LongBufferPoolTest.java:330 - Testing
BufferPool with memoryUsageThreshold=16777216 and enabling BufferPool.DEBUG
INFO [test:5] 2022-10-07 09:31:39,829 NoSpamLogger.java:92 - Maximum memory
usage reached (16,000MiB), cannot allocate chunk of 8,000MiB
java.lang.AssertionError: Last recycled 5 < last acquired 6
at
org.apache.cassandra.utils.memory.LongBufferPoolTest$Debug.check(LongBufferPoolTest.java:122)
at
org.apache.cassandra.utils.memory.LongBufferPoolTest.testAllocate(LongBufferPoolTest.java:353)
at
org.apache.cassandra.utils.memory.LongBufferPoolTest.testPoolAllocate(LongBufferPoolTest.java:159)
at
org.apache.cassandra.utils.memory.LongBufferPoolTest.testPoolAllocateWithRecyclePartially(LongBufferPoolTest.java:147)
{noformat}
The other test `testPoolAllocateWithoutRecyclePartially` works fine with only 6
patches.
So the 7-th patch fixes partial recycling. Not sure how serious this is - looks
like some chunks would simply be not recycled, so that could cause a memory
leak worst case?
was (Author: pkolaczk):
The {noformat}testPoolAllocateWithRecyclePartially{noformat} fails quite
reliably if I exclude the last patch:
{noformat}
INFO [main] 2022-10-07 09:31:39,811 LongBufferPoolTest.java:329 - 2022/10/07
09:31:39 - testing 16 threads for 2m
INFO [main] 2022-10-07 09:31:39,811 LongBufferPoolTest.java:330 - Testing
BufferPool with memoryUsageThreshold=16777216 and enabling BufferPool.DEBUG
INFO [test:5] 2022-10-07 09:31:39,829 NoSpamLogger.java:92 - Maximum memory
usage reached (16,000MiB), cannot allocate chunk of 8,000MiB
java.lang.AssertionError: Last recycled 5 < last acquired 6
at
org.apache.cassandra.utils.memory.LongBufferPoolTest$Debug.check(LongBufferPoolTest.java:122)
at
org.apache.cassandra.utils.memory.LongBufferPoolTest.testAllocate(LongBufferPoolTest.java:353)
at
org.apache.cassandra.utils.memory.LongBufferPoolTest.testPoolAllocate(LongBufferPoolTest.java:159)
at
org.apache.cassandra.utils.memory.LongBufferPoolTest.testPoolAllocateWithRecyclePartially(LongBufferPoolTest.java:147)
{noformat}
The other test ({noformat}testPoolAllocateWithoutRecyclePartially{noformat})
works fine with only 6 patches.
So the 7-th patch fixes partial recycling. Not sure how serious this is - looks
like some chunks would simply be not recycled, so that could cause a memory
leak worst case?
> org.apache.cassandra.utils.memory.LongBufferPoolTest - tests are flaky
> ----------------------------------------------------------------------
>
> Key: CASSANDRA-16681
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16681
> Project: Cassandra
> Issue Type: Bug
> Components: CI
> Reporter: Ekaterina Dimitrova
> Assignee: Piotr Kolaczkowski
> Priority: Normal
> Fix For: 3.11.x, 4.0.x, 4.x
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> Jenkins history:
> [https://jenkins-cm4.apache.org/job/Cassandra-4.0/50/testReport/junit/org.apache.cassandra.utils.memory/LongBufferPoolTest/testPoolAllocateWithRecyclePartially/history/]
> Fails being run in a loop in CircleCI:
> https://app.circleci.com/pipelines/github/ekaterinadimitrova2/cassandra/844/workflows/945011f4-00ac-4678-89f6-5c0db0a40169/jobs/5008
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
