I spent ages figuring this out and to be sure that I’m not ever going to waste time on this issue again I decided to write this post.
I created a workflow for Sharepoint 2007 using visual studio. Deployed it to a server and attached it to at list. It all worked fine. Weeks later I made some smaller changes in the workflow and redeployed it to the server.
After that whenever I tried starting the workflow it returned “Failed on start (retrying)” – What!!
I tried everything:
- Triplechecked correlation tokens
- Restated Timer job service
- Restarted server
- Deactivated/Activated the feature
- Updated the Performance Counter (http://blogs.msdn.com/b/wael/archive/2008/01/11/workflow-failed-on-start-retrying.aspx)
- Looked into the logs for f.. hours
- Googled my ass off reading everything about every possible workflow problem
Note: In some cases the workflow actually started anyway and the actions in the workflow for ex. sending a mail to the assigned user worked, however the workflow status was still “Failed on start”
Finally I accidentally stumbled across this article by Jeremy Hancock: http://blog.ozippy.com/2008/02/sharepoint-2007-workflow-on-start.html
For some reason regarding moving a site from one sitecollection to another I’d changes the Application Pool account running my site from Network service to an Active Directory account via the IIS Manager (inetmgr), sometime between the first and the second deployment. But I hadn’t changed the service account in Central administration 😦
After changing this all my workflows started running smoothly as ever.