AnsweredAssumed Answered

Process Start Delay

Question asked by TDBecker on Jan 9, 2013
Latest reply on Dec 11, 2013 by Suman Pramanik
Hi, I've been fighting this one for a while and am hoping somebody has run into it before.

I upgraded my Clarity version from 12.0.6 to 12.1.1 last December. Part of the upgrade was to move my application to new VM servers. (I should also add that I have the nsa and niku ear files managed through WebLogic).

The upgrade seemed to go well, but my users started complaining that the processes which are triggered from actions taken on the screen where being delayed up to 3 minutes. For example, we have a custom field "Gate" which triggers a process to update othe fields on the project. I have witnessed the delay and can understand their frustration. It seems like the process is in a running status, but then just sits there for up to 3 minutes. Once it starts, it completes very quickly.

I have two servers in QA and three in PROD. Both environments exhibit the same symptoms.

My theory was that the trigger is taking place on the node that doesn't have BG running on it, and the delay is cause by the message never reaching BG. Howerver, I tested that out by triggering the process from the screen on the same server as BG, and the process didn't run any quicker.

I have confirmed that BEACON is communicating between the nodes, NSA sees all nodes and I can see all servers using the Tower command. I do notice that the status of my processes engine always displays that it cannot be determined.

My network folks are going to help me capture traffic on the network tomorrow just in case there are packets being dropped.

Well, that's the sad story. Thanks for any help.