In a cluster where the messaging engine has failed over between nodes, a situation could arise where the transaction log (’tranlog’) has become corrupt or is unrecoverable. In this situations, you may see messages on SIB Queues that are in the LOCKED state but are unprocessed. If you factor in a sequential delivering of messages, we may see that the queue refuses to process any messages at all.
In this case, we have to look at the SIB Data store and run the following query:

select count(*) from sib000 where XID is not null

if you discover existing values in XID you should update XID field to null:

update sib000 set xid = null where xid is not null

This situation would only occur if the ‘tranlog’ on a server was unrecoverable and should not be considered a ‘normal’ step of server recovery or server startup.

Related Posts

• Event Sequencing Message Locking - Resolution!
• WebSphere Application Server: -recovery flag
• Tip: WebSphere Process Server - Do not delete the ‘tranlog’ on a production server
• ProcessTemplateNotFoundException: CWWBA0008E: Process template ‘myProcess, -valid_from_date-’ is not found
• Asynchronous SCA One-Way Operations And How To Recover

You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.