[
https://issues.apache.org/jira/browse/CASSANDRA-14014?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16254779#comment-16254779
]
Kurt Greaves commented on CASSANDRA-14014:
------------------------------------------
That mutation/write would never have got through due to the error. In fact you
can't send a mutation bigger than 1/2 the {{commitlog_segment_size}} or
{{max_mutation_size}}. The stacktrace you received is not a bug, however
bq. EOF after xxx bytes out of YYY",
may indicate a bug. The other errors are simply because cluster/node became
overloaded running repair. To diagnose corruption we'd need a lot more
information. Potentially more logs and probably a copy of the corrupt SSTable...
> Inserting of larger columns leads to exceptions and failing repairs
> -------------------------------------------------------------------
>
> Key: CASSANDRA-14014
> URL: https://issues.apache.org/jira/browse/CASSANDRA-14014
> Project: Cassandra
> Issue Type: Bug
> Components: Repair
> Environment: 7 node cluster, debian stretch, 32GB ram, 2TB disk, C*
> 3.11.1
> Reporter: Michael Wohlwend
> Attachments: cassandra-log.txt
>
>
> Inserting records with an unusual large text field (hundreds of MB instead of
> a few KB) caused exceptions of the kind "mutation too large":
> ===========================
> ERROR [MutationStage-3] 2017-11-07 08:32:52,712 StorageProxy.java:1414 -
> Failed to apply mutation locally : {}
> java.lang.IllegalArgumentException: Mutation of 185.439MiB is too large for
> the maximum size of 64.000MiB
> at
> org.apache.cassandra.db.commitlog.CommitLog.add(CommitLog.java:255)
> ~[apache-cassandra-3.11.1.jar:3.11.1]
> at org.apache.cassandra.db.Keyspace.applyInternal(Keyspace.java:585)
> ~[apache-cassandra-3.11.1.jar:3.11.1]
> at org.apache.cassandra.db.Keyspace.apply(Keyspace.java:462)
> ~[apache-cassandra-3.11.1.jar:3.11.1]
> at org.apache.cassandra.db.Mutation.apply(Mutation.java:227)
> ~[apache-cassandra-3.11.1.jar:3.11.1]
> at org.apache.cassandra.db.Mutation.apply(Mutation.java:232)
> ~[apache-cassandra-3.11.1.jar:3.11.1]
> at org.apache.cassandra.db.Mutation.apply(Mutation.java:241)
> ~[apache-cassandra-3.11.1.jar:3.11.1]
> at
> org.apache.cassandra.service.StorageProxy$8.runMayThrow(StorageProxy.java:1408)
> ~[apache-cassandra-3.11.1.jar:3.11.1]
> at
> org.apache.cassandra.service.StorageProxy$LocalMutationRunnable.run(StorageProxy.java:2647)
> [apache-cassandra-3.11.1.jar:3.11.1]
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> [na:1.8.0_141]
> at
> org.apache.cassandra.concurrent.AbstractLocalAwareExecutorService$FutureTask.run(AbstractLocalAwareExecutorService.java:162)
> [apache-cassandra-3.11.1.jar:3.11.1]
> at
> org.apache.cassandra.concurrent.AbstractLocalAwareExecutorService$LocalSessionFutureTask.run(AbstractLocalAwareExecutorService.java:134)
> [apache-cassandra-3.11.1.jar:3.11.1]
> at org.apache.cassandra.concurrent.SEPWorker.run(SEPWorker.java:109)
> [apache-cassandra-3.11.1.jar:3.11.1]
> at java.lang.Thread.run(Thread.java:748) [na:1.8.0_141]
> WARN [MutationStage-145] 2017-11-07 08:33:08,291
> AbstractLocalAwareExecutorService.java:167 - Uncaught exception on thread
> Thread[MutationStage-145,5,main]: {}
> ====================================
> Trying to repair the table fails. In the end, the cassandra shuts down on the
> node.
> The exceptions thrown during repair are "EOF after xxx bytes out of YYY",
> "leak detected",
> "failed to create merkle tree ... OutOfMemory" and "self-supression not
> permitted".
> See the attached log file (edited and shortened).
> Michael
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
