[
https://issues.apache.org/jira/browse/CASSANDRA-18710?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17777830#comment-17777830
]
Branimir Lambov commented on CASSANDRA-18710:
---------------------------------------------
It looks like the reason for the unexpected flush is the commit log:
{code:java}
[junit-timeout] INFO [OptionalTasks:1] 2023-10-12 21:55:11,095
ColumnFamilyStore.java:1017 - Enqueuing flush of
cql_test_keyspace_alt.table_01, Reason: COMMITLOG_DIRTY, Usage: 74.752KiB (0%)
on-heap, 3.777KiB (0%) off-heap
[junit-timeout] INFO [PerDiskMemtableFlushWriter_0:2] 2023-10-12 21:55:11,103
Flushing.java:154 - Writing Memtable-table_01@1180822937(6.854KiB serialized
bytes, 242 ops, 74.916KiB (0%) on-heap, 3.781KiB (0%) off-heap), flushed range
= [null, null)
[junit-timeout] INFO [PerDiskMemtableFlushWriter_0:2] 2023-10-12 21:55:11,128
Flushing.java:180 - Completed flushing
/tmp/cassandra/build/test/cassandra/data/cql_test_keyspace_alt/table_01-03e61210694a11eeb4091bdb4ac3170b/nc-1-big-Data.db
(6.839KiB) ... {code}
which is flushing just 242 out of the 1000 ops that the test needs per table.
We need to understand what causes these {{COMMITLOG_DIRTY}} flushes, because
there are quite a few tests that will fail if a flush happens at the wrong
time. Or maybe somehow disable commitlog-driven flushing for tests (e.g. by
setting a really large commit log space limit).
> Test failure:
> org.apache.cassandra.io.DiskSpaceMetricsTest.testFlushSize-.jdk17 (from
> org.apache.cassandra.io.DiskSpaceMetricsTest-.jdk17)
> ------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-18710
> URL: https://issues.apache.org/jira/browse/CASSANDRA-18710
> Project: Cassandra
> Issue Type: Bug
> Components: Test/unit
> Reporter: Ekaterina Dimitrova
> Assignee: Brandon Williams
> Priority: Normal
> Fix For: 5.0.x, 5.x
>
> Attachments: org.apache.cassandra.io.DiskSpaceMetricsTest.txt
>
>
> Seen here:
> [https://ci-cassandra.apache.org/job/Cassandra-trunk/1644/testReport/org.apache.cassandra.io/DiskSpaceMetricsTest/testFlushSize__jdk17/]
> h3.
> {code:java}
> Error Message
> expected:<7200.0> but was:<1367.83970468544>
> Stacktrace
> junit.framework.AssertionFailedError: expected:<7200.0> but
> was:<1367.83970468544> at
> org.apache.cassandra.io.DiskSpaceMetricsTest.testFlushSize(DiskSpaceMetricsTest.java:119)
> at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
> Method) at
> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:77)
> at
> java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> {code}
>
--
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
