MQSeries.net :: View topic - [SOLVED]Message Retry Limit

CHF · Posted: Mon Apr 19, 2004 1:19 pm Post subject:

We specified a Reply2Q for response to ProcessTemplateCreateAndStartInstance. Then our execution server died because it couldn't put reply message back to Reply2Q specified (When I look up at the Fmcsys.log, the reason was the reply message is bigger than the maximum message length in the Queue). We manually started the execution server, and again it went down because it tried to process the same message.

I did some research in Administration guide and programming guide on how to handle situations like this. I have few doubts as follows. Someone please help me in understanding these.

(1) What happens in general when some thing happens like I said above? For instance, execution server died because it couldn't process an XML message, then after 5 min admin server brought back execution server automatically, again execution server tried to process that XML message and died again........ Like this how many times will it happen? Is it going to happen like a loop until someone manually clears up the messages from the XML input queue (or) Is Workflow clever enough to clear the message after certain number of tries?

(2) I saw in administration guide about RTMessageRetryLimit. What I understood from this is that Workflow will only try a message for 5 times (Which is default). After that it pus the message to Hold Queue. Am I thinking right on this?

(3) How does WF keeps track in such that it tried a message for 5 times? What I mean to ask here is that, Is there any other variable in Workflow such as RetryCount which might have initial value of 0, and for each try it incremented by 1 and when it reaches the RTMessageRetryLimit (5) , then it puts that message to Hold Queue?

(4) I saw about backout count in programming guide. Is there any relation between backout count and RTMessageRetryLimit?

If I am not clear about any thing above, please let me know.

Thanks for your time
CHF