I have this process, nothing extremely large but it's not small either, maybe 25 activities in all, mostly calling into subprocess. For some reason when this process is checked in and deployed, the new version is discarded. If I make a recording and replay it, it shows me the old version of the process. I have tried everything I can think of; undeploy, redeploy, delete and get the application again, nothing seems to work. I even went as far as reinstalling workbench (reinstalling the server isn't an option). Seems like this process just can't be updated, this happens on all our machines trying to modify this process. The only thing that works so far is to save this process under a different name and modify that copy, but as soon as this new version is deployed and has been run, I end up at the same point. I went as far as recreating the entire process from scratch and still no luck. We do have a second server where I have tried this process same thing happens there.
So I don't know what to do anymore, my instincts tells me there is something that the deployment engine doesn't like in my process and once it's deployed it can't delete the object anymore during subsequent deploys. I snooped around in the LiveCycle internal database and saw references to the old process name, the one I deleted from the server while saving under a different name (don't know if this is normal or not). Something is definitely not being cleaned up properly.
Where is the process source stored on the server when I deploy an application? Is it in the database, is it copiled into a jar (or other container file)?
LiveCycle ES2 Server
Version: 18.104.22.168, GM
Patch Version: 2.73;2.38;SP2
Service Pack Version: ???2.73;2.38;SP2_BuildVersion???
my guess would be something not right with Application server and how the LC application started
I would try restarting the server and make sure you dont have any errors in the log
And also if the application server is taking too long not all the LC services are started correctly not much info in the log
hope it helps
Thanks for the suggestions, believe me we've restarted the server many, many times, we just can't find any errors in the log that pertain to this problem.
I did some more tests on this problem just recently and I did find that saving the process as a new process and comparing the two process xml (from the metadata) shows differences in the indentation and the <?xml...?> tag which is now missing. I replaced the old process with the new one and this managed to get me past the issue for now. But this is the second time this has happened to this process and last time renaming the process just wouldn't work past the first check-in.
Thanks for the reply, I appreciate any ideas!