[
https://issues.apache.org/jira/browse/CASSANDRA-6822?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13975353#comment-13975353
]
Ananthkumar K S commented on CASSANDRA-6822:
--------------------------------------------
Just one doubt. We have been running 2.0.3 for so long and have not found any
external issue so far other than a private link failure causing a stall in
operation.
Can you tell how stable is 2.0.7? Up gradation can be done on the run right
?(by just bringing down the node running 2.0.3 by nodetool drain and starting
off new one with updates in cassandra.yaml? Shall I transfer the SSTables
directly from 2.0.3 to 2.0.7 or allow it to sync its data after starting? Am
actually using vnodes.
> Never ending batch replay after dropping column family
> ------------------------------------------------------
>
> Key: CASSANDRA-6822
> URL: https://issues.apache.org/jira/browse/CASSANDRA-6822
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Environment: Cassandra 2.0.5
> Reporter: Duncan Sands
> Assignee: Aleksey Yeschenko
> Fix For: 1.2.17, 2.0.7, 2.1 beta2
>
> Attachments: 6822.txt
>
>
> After dropping a column family, I got an infinite number of messages like
> this:
> WARN [BatchlogTasks:1] 2014-03-07 17:50:08,962 BatchlogManager.java (line
> 235) Skipped batch replay of bc3916e0-a533-11e3-9c41-55f3810000b5 due to
> org.apache.cassandra.db.UnknownColumnFamilyException: Couldn't find
> cfId=2fdad07d-0cc3-373a-8aa0-15837cd9bc6c
> I say an infinite number but I only waited about 12 hours before I decided I
> had had enough and truncated system.batchlog. So perhaps it would have gone
> away by itself.
> Presumably the batch replay is skipped but not discarded, so will keep on
> turning up forever.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
