[ 
https://issues.apache.org/jira/browse/CASSANDRA-4078?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

paul cannon updated CASSANDRA-4078:
-----------------------------------

    Attachment: 4078.patch2.txt

Ok. We'll keep that info around in case we ever see something like this again. 
Thanks.

I do recommend that these asserts (patch attached, also at my pending/4078-2 
tag in github) get committed to mainline Cassandra, though, to help protect 
against this sort of situation. The asserts in this patch will not catch all 
possible updates to SSTable.last, but they surround some of the more important 
(and less frequent) places.

Also, the asserts inside the IntervalTree creation should avoid getting into 
confusing StackOverflowError situations.
                
> StackOverflowError when upgrading to 1.0.8 from 0.8.10
> ------------------------------------------------------
>
>                 Key: CASSANDRA-4078
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-4078
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 0.8.10
>         Environment: OS: Linux xps.openfin 2.6.35.13-91.fc14.i686 #1 SMP Tue 
> May 3 13:36:36 UTC 2011 i686 i686 i386 GNU/Linux
> Java: JVM vendor/version: Java HotSpot(TM) Server VM/1.6.0_31
>            Reporter: Wenjun
>            Assignee: paul cannon
>            Priority: Critical
>             Fix For: 0.8.10
>
>         Attachments: 4078.add-asserts.txt, 4078.patch2.txt, 
> cassandra.yaml.1.0.8, cassandra.yaml.8.10, keycheck.txt, system.log, 
> system.log.0326, system.log.0326-02
>
>
> Hello
> I am trying to upgrade our 1-node setup from 0.8.10 to 1.0.8 and seeing the 
> following exception when starting up 1.0.8.  We have been running 0.8.10 
> without any issues.
>  
> Attached is the entire log file during startup of 1.0.8.  There are 2 
> exceptions:
> 1. StackOverflowError (line 2599)
> 2. InstanceAlreadyExistsException (line 3632)
> I tried "run scrub" under 0.8.10 first, it did not help.  Also, I tried 
> dropping the column family which caused the exception, it just got the same 
> exceptions from another column family.
> Thanks

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Reply via email to