| Author |
Message
|
| rajasri |
 Posted: Wed Jun 10, 2009 4:10 am Post subject: Message Flows Failing |
 |
|
Centurion
Joined: 11 Jun 2006
Posts: 114
|
Hi All,
We have a problem in Production where messages are failing in message flows with error
CWF Logical Tree - Message Definition Mismatch
TDS writing error - Incoming message not defined. , CWF Data Overflow Error
RM Worker Not Found.
But when we replay the same messages again, they go fine. We are assuming that this has started after the Broker and its database are bounced. To add, the message flows have been running fine from 2 years. The broker details are as below.
BIPv500 en US
ucnv Console CCSID 1208 dft ucnv CCSID 1208
OS ISO8859-1 ICUW ISO-8859-1 ICUA ISO-8859-1
BIP8996I: Version: 500
BIP8997I: Product: WebSphere MQ Integrator
BIP8998I: CMVC Level: S500-L40713
Did any one of you have faced the same problem and resolved it?. Thanks in advance.
regards,
Pradeep Raju. |
|
| Back to top |
|
 |
| AkankshA |
| Posted: Thu Jun 11, 2009 8:00 pm Post subject: |
|
|
Grand Master
Joined: 12 Jan 2006
Posts: 1494
Location: Singapore
|
upgrade to latest patch level...
May be , just may be there are 2 different versions of message sets / flows deployed (one was stopped and other used ) and broker bounce brought them both in existance...
How about removing all flows and redeploying all the flows in PM time window..
_________________
Cheers |
|
| Back to top |
|
|
| rajasri |
| Posted: Thu Jun 11, 2009 11:26 pm Post subject: |
|
|
Centurion
Joined: 11 Jun 2006
Posts: 114
|
Hi Akanksha,
I can confirm you that ther are no 2 different versions of flows or sets. I double checked that. And I have redeployed the failing message SETS. but this also dint solve the problem. Any other solution before we conclude that deploying message FLOWS might work.
Regards,
Pradeep. |
|
| Back to top |
|
|
| Hari |
| Posted: Sat Jun 13, 2009 1:58 pm Post subject: |
|
|
Centurion
Joined: 21 Nov 2002
Posts: 117
Location: USA
|
Are you sure there is just one message coming in?
IMO there is one more (probably blank) message coming in....or a bug in the code which is causing a blank/wrong message being propagated. |
|
| Back to top |
|
|
| rajasri |
| Posted: Mon Jun 15, 2009 1:34 am Post subject: |
|
|
Centurion
Joined: 11 Jun 2006
Posts: 114
|
| The messages which are failing are valid, and when I replay the message, they go fine. The code is bug free and this has been running with out the mentioned problems for 2 continous years. |
|
| Back to top |
|
|
| rajasri |
| Posted: Mon Jun 22, 2009 11:54 pm Post subject: |
|
|
Centurion
Joined: 11 Jun 2006
Posts: 114
|
| Hi All, does any one know the solution for this?? |
|
| Back to top |
|
|
| kimbert |
| Posted: Tue Jun 23, 2009 12:18 am Post subject: |
|
|
Jedi Council
Joined: 29 Jul 2003
Posts: 5542
Location: Southampton
|
You are getting four different errors, not one intermittent error.
| Code: |
CWF Logical Tree - Message Definition Mismatch
TDS writing error - Incoming message not defined. , CWF Data Overflow Error
RM Worker Not Found. |
The first exception is thrown by the CWF writer when it cannot match the contents of the message tree to the message definition.
The second is thrown by the TDS writer when it cannot find a message definition corresponding to the 'Message Type' value ( it set on the node or supplied in the mcd folder of an RFH2 header )
The third is thrown by the CWF writer while outputting data - not sure of the details for that one
The fourth is complaining that the 'Message Format' setting does not identify a physical format in the specified message set ( it is set on the node or supplied in the mcd folder of an RFH2 header ). It might be caused by the 'Message Set' value being incorrect, of course.
My guess is that you have some kind of random corruption of your input messages/flow/message sets. I suggest that you redeploy everything.
It would also be a really good idea to move up to a supported version of the product, but I expect you know that already  |
|
| Back to top |
|
|
| rajasri |
| Posted: Wed Jun 24, 2009 2:19 am Post subject: |
|
|
Centurion
Joined: 11 Jun 2006
Posts: 114
|
| Thanks Kimbert. We have the same problem in our test environment as well. So I am going to redeploy all the artifacts and see if its fine or not. |
|
| Back to top |
|
|
| sumar |
| Posted: Fri Jun 26, 2009 6:15 am Post subject: |
|
|
Newbie
Joined: 26 Jun 2009
Posts: 1
|
We have the same problem, I think. (v5.0 FP9)
About 10 000 messages are done OK and another one makes error like "RM Dictionary Not Found".
After reading of message from queue a filter node makes error - in trace node I can see this.
(0x01000000):Properties = (
(0x03000000):MessageSet = NULL
(0x03000000):MessageType = 'sip_message'
(0x03000000):MessageFormat = 'XML'
[Sometimes MessageType = NULL, MessageFormat = XXX.]
Values of MessageSet, MessageType MessageFormat are constants filled in default tab of input queue.
It looks like broker lost information in its header.
I have also tried removing and redeploying flows and msgsets.
Is it possible that problem is there?
When starting, broker prints this message:
"Malloc Buckets: Value specified for bucket_sizing_factor:n configuration
option is invalid. This value must be a multiple of 16. Using default
value of 64 for bucket sizing factor." |
|
| Back to top |
|
|
| Vitor |
| Posted: Sat Jun 27, 2009 2:36 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| sumar wrote: |
| We have the same problem, I think. (v5.0 FP9) |
You certainly have the same unsupported software problem!
| sumar wrote: |
When starting, broker prints this message:
"Malloc Buckets: Value specified for bucket_sizing_factor:n configuration
option is invalid. This value must be a multiple of 16. Using default
value of 64 for bucket sizing factor." |
This points to a configuration error that certainly isn't helping. I would theorise that after 10,000 messages the broker is running out of memory for some reason (possibly the way your flows are constructed) and corrupting the in-core copies of the message set objects.
Fix this configuration and re-try. Or better still upgrade to a supported version and re-try; memory handling in better in v6.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
|
|
