Livelink service crashes

[madhuusa]

Team,

We upgrade one of our LL installations to 9.7.1 from 9.5.1. (this is our second installation. We have 3 installations due to huge setup here).

In this upgrade, things went smooth. We have 2 front end with 3 LL instances and one seperate admin machine.

In the admin machine though livelink service crashes every one hour (not the admin server but livelink service).

We are in March 2009 patch.

We see similar kind of logs (below) when it crashes. OT support is also scrambling to find a reason.

Gurus, can you help me here. It looks like extractor process is in the middle of extraction when this happens.

It says bad allocation. I guess memory is being killed somewhere. We are monitoring memory also but of no use since it does not spike huge before the crash.

Log[7] : [-1187642304]_WriteObject content ref
loc[\\exfilestore\pd\exfile\0052\259\52259846.dat] Log[7] :
IndexLibrary.UpdateLibraryNodeIterator::Next() processing
R<'Seq'=52259845,'Operation'=1,'DataID'=52259845,'VersionNum'=0> Log[7] :
[-1187642288]ExportToIPool: AddOrReplaceataId=52259845&Version=1
RMClassificationMetaData Error selecting against rimsNodeClassification for
nodeId 52259845 Log[7] : [-1187642226]_WriteObject content ref
loc[\\exfilestore\pd\exfile\0052\259\52259845.dat] Log[7] :
IndexLibrary.UpdateLibraryNodeIterator::Next() processing
R<'Seq'=52259844,'Operation'=1,'DataID'=52259844,'VersionNum'=0> Log[7] :
[-1187642210]ExportToIPool: AddOrReplaceataId=52259844&Version=1 07/26/2009
15:37:21 [2092808424] LLServer caught an exception: bad allocation;
terminating. 07/26/2009 15:37:21 [2092808569] LLI thread 4 shutdown (thread
exiting) Socket object auto-closed

 

[ggriffit]

Is there anything special about the node id 52259844 compared to any others ? Do you have any TRACE / THREAD files to got with this ?

Greg Griffiths
Livelink Certified Developer & ECM Global Star Champion 2005 & 2006

http://www.greggriffiths.org/livelink/

 

[LiveLinkPRO]

Look at the complete logs, especially threads and connects. Also check that DataID as mentioned.

Make sure that no other apps mess up your memory or other applications access your files / dataflow as the processes are running. An EFS on a UNC share is also not the best thing to do, network peaks or outages are not that rare.

Apart from that, this is really a Support thing to look at and not something somebody would tell you in a forum. A lot more information and analysis activities would be required.

Jack

 

[madhuusa]

Thanks Greg and Jack for your response.

The issue is still there and various things are being thrown around from oracle 9i client vs oracle 10g db, backup software, vertias software, antivirus software, router issue, java issue, etc, etc

If we find anything I will post.

 

[LiveLinkPRO]

...from oracle 9i client vs oracle 10g db...."

Are you asserting that you are using an Oracle 9i Client against an Oracle 10g database? This is totally unsupported and will result in all kinds of issues, frequent crashes and instabilities included, but not limited to that.

You MUST use the 10.2 Client (the full regular client, not instant or express), otherwise crashes are to be expected.

You better install the correct client and test again. I bet this will resolve your issue.

Jack

 

[LiveLinkPRO]

So did the new Oracle Client resolve the problem?

C'mon, commit to it. It's not a shame to overlook something