Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@cassandra.apache.org
Received-SPF: pass (nike.apache.org: domain of cbertu81@libero.it designates
 212.52.84.101 as permitted sender)
Message-ID: <18298044.684051311260048926.JavaMail.defaultUser@defaultHost>
Date: Thu, 21 Jul 2011 16:54:08 +0200 (CEST)
From: "cbertu81@libero.it" <cbertu81@libero.it>
Reply-To: "cbertu81@libero.it" <cbertu81@libero.it>
To: user@cassandra.apache.org
Subject: Can not repair
MIME-Version: 1.0
Content-Type: text/plain;charset="UTF-8"
Content-Transfer-Encoding: 7bit

Hi all,
I can't get the repair in my production.
We are out since 6 months but before we did not perform any delete do we 
didn't need to run repair.
Now we are out since 2 weeks with a new version of our software that performs 
delete but we can not get the nodetool repair working,

The first problem I see in the log is a: 

ERROR 16:34:49,790 Fatal exception in thread Thread[CompactionExecutor:1,1,
main]
java.io.IOException: Keys must be written in ascending order.
        at org.apache.cassandra.io.sstable.SSTableWriter.beforeAppend
(SSTableWriter.java:111)
        at org.apache.cassandra.io.sstable.SSTableWriter.append(SSTableWriter.
java:128)
        at org.apache.cassandra.db.CompactionManager.doCompaction
(CompactionManager.java:451)
        at org.apache.cassandra.db.CompactionManager$1.call(CompactionManager.
java:124)
        at org.apache.cassandra.db.CompactionManager$1.call(CompactionManager.
java:94)
        at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
        at java.util.concurrent.FutureTask.run(FutureTask.java:138)
        at java.util.concurrent.ThreadPoolExecutor$Worker.runTask
(ThreadPoolExecutor.java:886)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run
(ThreadPoolExecutor .java:908)
        at java.lang.Thread.run(Thread.java:662)


I've read something concerning a change of the partitioner but I've never 
modified it. There is to say that we were online with 0.6.5 and now we are with 
0.7.5: to migrate from this version to another we followed the documentation 
(conversion of the yaml, nodetool drain and so on) ... 

The system is working even if there is this "problem" but not the repair. The 
limit of FD of the user running the repair is unlimited but everytime we get a 
"Too many open files".

I'm little bit worried cause if some delete reappears the webapp will export 
wrong data ...

Best regards

Carlo