IBM FileNet P8, Version 5.2.1            

Replication failed, multiple records trapped in the replication queue

Replication fails when multiple records are trapped in the replication queue.

Symptoms

If several records are trapped in the replication queue, including a stop replication, the document is never de-replicated. Consider the following scenario:

Diagnosing the problem

Whenever a journal is marked as corrupted, all subsequent journals for that same object are trapped in the replication queue. You must fix the problem, but more importantly, you must reprocess or delete the journal.

When you de-replicate the document, only the replication group in the document is reset immediately. A journal is created for the de-replicate during the journal processing, which can be much later. The journal processing resets the external identity in the document and the replication group and external identity in the associated annotations.

Because there is a corrupted journal for this document, the de-replicate journal remains in the queue. For this reason, the f_ce_os_id value in the FileNet Image Services document is not removed. Because the de-replicate journal was not processed, the replication group is still in the annotation, any updates or deletions to the annotation replicate to FileNet Image Services, and the annotation journals are not blocked by the corrupted journal for the document. In general, to correct any failed situation, the first step is to look at and understand the reason why the journal failed, then reprocess or delete the corrupted journal. That corrupted journal must be gone before you make further changes to the document.

Resolving the problem

Remove all journals in the replication queue for the specific document before you make further updates to FileNet Image Services.


Last updated: October 2015
p8pco002.htm

© Copyright IBM Corporation 2015.