[ 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