cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jonathan Ellis (Resolved) (JIRA)" <>
Subject [jira] [Resolved] (CASSANDRA-3923) Cassandra Nodetool Cleanup causing sstable corruption.
Date Wed, 04 Apr 2012 18:11:23 GMT


Jonathan Ellis resolved CASSANDRA-3923.

       Resolution: Cannot Reproduce
    Fix Version/s:     (was: 1.1.0)

As noted on CASSANDRA-3065, I don't think we'll be able to make progress on this without a
pre-corruption snapshot to reproduce the problem with.

(We have also observed bad hardware and buggy virtualization layers to cause corruption; see

CASSANDRA-4051 is also open to improve streaming reliability which may be relevant.
> Cassandra Nodetool Cleanup causing sstable corruption.
> ------------------------------------------------------
>                 Key: CASSANDRA-3923
>                 URL:
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Tools
>    Affects Versions: 0.8.2
>            Reporter: Samarth Gahire
>              Labels: cleanup, compaction, corruption, nodetool, scrub
>   Original Estimate: 48h
>  Remaining Estimate: 48h
> We recently doubled size of our cluster. 
> After that,we ran the cleanup on the old machines in the cluster.
> After that we loaded the data which triggered minor compaction which did not complete
due to corrupt sstables
> This was the case with those machines only on which we ran the cleanup.
> As the cleanup is unavoidable after node addition, what is the way to avoid this problem?
> Is this issue fixed in newer versions of Cassandra(As we are using cassandra-0.8.2)?
> OR 
> Are there any steps / procedure that will avoid need of cleanup after node addition.
> The same issue is reported here: CASSANDRA-3065

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:!default.jspa
For more information on JIRA, see:


View raw message