Stopped CPIC Update Workprocess

[Mark Pyc Workflow]

Hi,
 
This is in reference to a question raised a few days ago...
I did not receive any replies about this one, but it's seems I have fixed
it. The problem appears to be when a workflow has a syncronous update
background method as the first step. If sufficient concurrent users process
the original transaction (which in this case starts the workflow via Change
Document event raise) to hold all update tasks the system ends in a deadlock
situation. It seems the update WP is not released until the WF does it's
thing, which it can't if all Update WPs are already filled.
 
By changing the coding of this method to use a CALL TRANSACTION in Local
update rather than syncronous, the problem seems to be overcome (no
guarantees, but haven't broken it yet). An Asyncronous update may have also
worked, but would have added extra load on the system in terms of a
terminating event.
 
I'm very surprised by this if it is true. Has anyone had similar problems.
Has anyone seen any warnings about using syncronous background methods? I
believe a requested start delay would also have been an effective
workaround, but the business process could not accept delays in this case.
 
Any thoughts or comments are greatly appreciated.
Mark Pyc
pyc_workflow at hotmail.com
 
_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp