cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Stefania (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-12100) Compactions are stuck after TRUNCATE
Date Wed, 03 Aug 2016 09:00:29 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-12100?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15405595#comment-15405595
] 

Stefania commented on CASSANDRA-12100:
--------------------------------------

I'll write a distributed test with 3 nodes to see if we can reproduce this.

>From a code inspection, I can already see a problem at this line [here|https://github.com/stef1927/cassandra/blob/cassandra-3.0/src/java/org/apache/cassandra/db/compaction/CompactionTask.java#L171].
If we interrupt the compaction here, then {{collector.finishCompaction()}} is never called
as it should be in the {{finally}} block of the outer {{try}} block. We should definitely
fix this. This only affects the compaction stats however, the actual compaction operation
should have been aborted. If you can reproduce it, you can verify this by checking that there
are no compaction log files in the table directories.

Regarding the {{Unexpected exception during request;}}, I have no idea how it could be related.
How do you issue the truncate command, via cqlsh or via nodetool?

> Compactions are stuck after TRUNCATE
> ------------------------------------
>
>                 Key: CASSANDRA-12100
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-12100
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Compaction
>            Reporter: Stefano Ortolani
>            Assignee: Stefania
>             Fix For: 3.0.x
>
>         Attachments: node3_jstack.log
>
>
> Hi,
> since the upgrade to C* 3.0.7 I see compaction tasks getting stuck when truncating the
column family. I verified this on all nodes of the cluster.
> Pending compactions seem to disappear after restarting the node.
> {noformat}
> root@node10:~# nodetool -h localhost compactionstats
> pending tasks: 6
>                                      id   compaction type              keyspace     
          table   completed      total    unit   progress
>    24e1ad30-3cac-11e6-870d-5de740693258        Compaction   		     schema		         table_1
          0   57558382   bytes      0.00%
>    2be2e3b0-3cac-11e6-870d-5de740693258        Compaction   		     schema		         table_2
          0   65063705   bytes      0.00%
>    54de38f0-3cac-11e6-870d-5de740693258        Compaction   		     schema		         table_3
          0     187031   bytes      0.00%
>    31926ce0-3cac-11e6-870d-5de740693258        Compaction   		     schema		         table_4
          0   42951119   bytes      0.00%
>    3911ad00-3cac-11e6-870d-5de740693258        Compaction   		     schema		         table_5
          0   25918949   bytes      0.00%
>    3e6a8ab0-3cac-11e6-870d-5de740693258        Compaction   		     schema		         table_6
          0   65466210   bytes      0.00%
> Active compaction remaining time :   0h00m15s
> {noformat}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message