| Author |
Message
|
| Alan P |
 Posted: Wed Oct 11, 2006 8:13 am Post subject: TimeoutControl/TimeoutNotifcation node issues |
 |
|
Newbie
Joined: 20 Apr 2006
Posts: 4
|
Hi,
Im having a bit of an issue with TimeoutControl nodes and TimeoutNotification nodes which I wondered if anyone has seen before.
We are trying to delay a message at a point in the flow (for 60 seconds for example). We have a timeoutcontrol node setup at the place in the flow we wish to perform the delay and we have a timeoutnotification node paired up with it located at the point we wish to continue the process.
I have left all values in the Message section of the TimeoutControl node property sheet empty (though have tried different playing about with the Message Domain).
The message going into the timeoutcontrol node looks as follows:
Message
>Properties
>><usual message property info, should I be looking for any specific under here which isnt present?>
>XMLNSC
>>TestData
>>><testdata contents>
LocalEnvironment
>TimeoutRequest
>>Count = 1
>>Startime = 16:54:36
>>IgnoreMissed = false
>>Identifier = TimeReq1
>>Action = SET
Environment
><Various environment variables we are using>
ExceptionList (nothing under here)
As soon as the message enters the timeoutcontrol node (it doesnt do the required delay), a 'message' goes down the failure route with the following information:
Message (nothing here)
LocalEnvironment (nothing here)
Environment (nothing here)
ExceptionList
>RecoverableException
>>File = TimeoutControl\\ImbTimeoutNotifcationNode.cpp
>>Line = 874
>>Function = ImbTimeoutNotificationNode::propagateFailureDeleteMsg
>>Type = ComIbmTimeoutNotificationNode
>>Catalog = BIPv600
>>Severity = 3
>>Number = 4619
Has anyone seen this? I wondered if its anything to do with parsing the message incorrectly but if its a parsing failure it usually gives you a more useful error message.
Ive played around with the parser options in the TimoutControl node but the exact same error happens each time. Does the 'Message' section of the timeoutrequest message need anything else?
Im using Broker v6.
Thanks
Alan |
|
| Back to top |
|
 |
| wschutz |
| Posted: Wed Oct 11, 2006 8:37 am Post subject: |
|
|
Jedi Knight
Joined: 02 Jun 2005
Posts: 3316
Location: IBM (retired)
|
|
| Back to top |
|
|
| Alan P |
| Posted: Thu Oct 12, 2006 12:13 am Post subject: |
|
|
Newbie
Joined: 20 Apr 2006
Posts: 4
|
Yes, ive been using this scenario to get as far as i have.
Its probably just something simple im missing! Does a message need a MQMD to go through a TimeoutControl node? |
|
| Back to top |
|
|
| AkankshA |
| Posted: Thu Oct 12, 2006 4:28 am Post subject: |
|
|
Grand Master
Joined: 12 Jan 2006
Posts: 1494
Location: Singapore
|
Yes
You need MQMD header for that
TimeOutcontrol node puts the message in SYSTEM.BROKER.TIMEOUT.QUEUE and TimeOutNotification node picks it from there. |
|
| Back to top |
|
|
| AlanP |
| Posted: Thu Oct 12, 2006 5:44 am Post subject: |
|
|
Newbie
Joined: 05 Jan 2006
Posts: 2
Location: London
|
| That makes sense. The messages coming into our flows dont have an MQMD so if I construct an MQMD and create a unique MessageId and CorrelationId in a compute node just before the timeoutcontrol, will this be adequate for the timeoutcontrol node purpose? |
|
| Back to top |
|
|
| AkankshA |
| Posted: Thu Oct 12, 2006 7:29 pm Post subject: |
|
|
Grand Master
Joined: 12 Jan 2006
Posts: 1494
Location: Singapore
|
yes it should work....
one of the issue which you would face is that you would loose all the environment data/variables. |
|
| Back to top |
|
|
|
|
