[
https://issues.apache.org/jira/browse/CASSANDRA-7803?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Marcus Eriksson updated CASSANDRA-7803:
---------------------------------------
Attachment: 0001-dont-try-to-remove-fake-FINAL-files.patch
When opening the compaction result early we fake the descriptor to be FINAL,
then, when we abort the compaction we try to remove the FINAL file which does
not exist. Patch checks if the sstable is opened early and removes the TMPLINK
files instead.
> When compaction is interrupted, it leaves locked, undeletable files
> -------------------------------------------------------------------
>
> Key: CASSANDRA-7803
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7803
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Environment: RedHat, xfs4, JNA enabled, JBOD,
> Reporter: Scooletz
> Assignee: Marcus Eriksson
> Labels: compaction
> Fix For: 2.1.0
>
> Attachments: 0001-dont-try-to-remove-fake-FINAL-files.patch
>
>
> During tests of new features of 2.1 like:
> - incremental repairs
> - leveled compaction
> I interrupted a compaction, which left the following ERROR in the _system.log_
> {quote}
> org.apache.cassandra.db.compaction.CompactionInterruptedException: Compaction
> interrupted: Compaction@152e6e70-1975-11e4-ba09-61f0d75c60c6(xx, xxx,
> 378505918/1993581634)bytes
> at
> org.apache.cassandra.db.compaction.CompactionTask.runWith(CompactionTask.java:174)
> at
> org.apache.cassandra.io.util.DiskAwareRunnable.runMayThrow(DiskAwareRunnable.java:48)
> at
> org.apache.cassandra.utils.WrappedRunnable.run(WrappedRunnable.java:28)
> at
> org.apache.cassandra.db.compaction.CompactionTask.executeInternal(CompactionTask.java:74)
> at
> org.apache.cassandra.db.compaction.AbstractCompactionTask.execute(AbstractCompactionTask.java:59)
> at
> org.apache.cassandra.db.compaction.CompactionManager$BackgroundCompactionTask.run(CompactionManager.java:235)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
> ~[na:1.7.0_09]
> at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:334)
> ~[na:1.7.0_09]
> at java.util.concurrent.FutureTask.run(FutureTask.java:166)
> ~[na:1.7.0_09]
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1110)
> ~[na:1.7.0_09]
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:603)
> [na:1.7.0_09]
> at java.lang.Thread.run(Thread.java:722) [na:1.7.0_09]
> {quote}
> Right after that, a cascade of reoccurring errors was being emitted till the
> restart:
> {quote}
> ERROR [NonPeriodicTasks:1] 2014-08-19 13:38:41,258
> SSTableDeletingTask.java:81 - Unable to delete
> /grid/data04/cassandra/data/xx/xxx-152e6e70197511e4ba0961f0d75c60c6/xx-xxx-ka-55058-Data.db
> (it will be removed on server restart; we'll also retry after GC)
> {quote}
> which made this node blinking (noted from the other nodes gossiper log
> entries).
> After restart, the node is healthy and fully operational.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
