| Author |
Message
|
| Fmoosakutty |
 Posted: Sat May 27, 2006 12:10 am Post subject: Message Id and Correaltion Id getting changed |
 |
|
Newbie
Joined: 26 May 2006
Posts: 2
|
We are trying to set up a System which having two MQ servers one resides on
Windows And other on Mainframe CICS. Windows side language is CICS
and Mainframe side PL/I. Messages are getting passed properly and triggers also
working fine.
When we are trying to implement reply request logic we are not able to match
the message Ids and correlation Ids.
Process in Mainframe
1) Request Send to Mainframe
2) Reads the queue and processed
3) ReplyQueue.MessageId = Request.CorrelationId
After getting reply from mainframe in dot net side the Ids are not matching
Request.CorrealtionId is not matching with Reply.messageId.
We tried passing correlation Id also.That is also not matching.
We are not able to display that Ids in Pl/I. Some Junks are displaying
But from windows side we could see that internal format is having minor difference
Between two Ids. May be in the process of send –receive some where data is not
pssing correctly or getting converted or changed.As per manuals all MQMD structure
conversion and integrity between different operating systems and encodings are ensured by MqSeries.But we are getting this change..?How?
Is that sender or receiver channel making some conversions? If yes how to disable this?
Can anybody help us on this?
One more Analysis point: As we could see a minor chane in hex for mat of these ids is it becasue message ids are created using some charecter sets which is not supported by Pl/I?But we are just moving the elemets not doing any process ?
Regards
Firoze |
|
| Back to top |
|
 |
| zpat |
| Posted: Sat May 27, 2006 2:28 am Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5849
Location: UK
|
It's more normal (for the server application) to set the reply message correlid value from the request message msgid value.
Most request Correlids are null, unless especially set by the application whereas the Msgid will always be set.
The client application saves the msgid of the request message and uses it for MATCH_CORRELID on the get of the reply. Also always use GMO WAIT, CONVERT options. |
|
| Back to top |
|
|
| markt |
| Posted: Sat May 27, 2006 2:47 am Post subject: |
|
|
Knight
Joined: 14 May 2002
Posts: 504
|
| Sounds like you're expecting msgid/correlid to be printable text strings. They are not - MQBYTE24. On Windows make sure you're using memcpy, not strcpy. |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Sat May 27, 2006 3:48 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
| Fmoosakutty wrote: |
Windows side language is CICS
and Mainframe side PL/I. |
Does CICS uses memcpy or strcpy?
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| wschutz |
| Posted: Sat May 27, 2006 4:19 am Post subject: |
|
|
Jedi Knight
Joined: 02 Jun 2005
Posts: 3316
Location: IBM (retired)
|
Maybe he meant C#...haven't seen CICS on intel in a spell....
_________________
-wayne |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Sat May 27, 2006 11:58 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
| wschutz wrote: |
| Maybe he meant C#...haven't seen CICS on intel in a spell.... |
I haven't seen MQSI 2.1 in quite a spell. Apparently, a lot of people are still using it.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| wschutz |
| Posted: Sat May 27, 2006 12:02 pm Post subject: |
|
|
Jedi Knight
Joined: 02 Jun 2005
Posts: 3316
Location: IBM (retired)
|
I'd wager (rum) that you've seen MQSI 2.1 more recently than I've seen CICS/NT or CICS/OS2.
_________________
-wayne |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Sat May 27, 2006 4:39 pm Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
| wschutz wrote: |
| I'd wager (rum) that you've seen MQSI 2.1 more recently than I've seen CICS/NT or CICS/OS2. |
You keep trying to bet me my own rum, and somehow expecting that to succeed.
I'll bet you some of someone else's beer that I've seen CICS/NT more recently than you have.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| tleichen |
| Posted: Thu Jun 01, 2006 6:57 am Post subject: |
|
|
Yatiri
Joined: 11 Apr 2005
Posts: 663
Location: Center of the USA
|
| jefflowrey wrote: |
I'll bet you some of someone else's beer that I've seen CICS/NT more recently than you have. |
I think you guys would say anything for beer.... 
_________________
IBM Certified MQSeries Specialist
IBM Certified MQSeries Developer |
|
| Back to top |
|
|
| Bill.Matthews |
| Posted: Thu Jun 01, 2006 8:14 am Post subject: |
|
|
Master
Joined: 23 Sep 2003
Posts: 232
Location: IBM (Retired)
|
Since Forize said
| Quote: |
After getting reply from mainframe in dot net side the Ids are not matching
Request.CorrealtionId is not matching with Reply.messageId. |
ie a reference is made to dot net - so C# is a much better bet than CICS on Intel.
As for me - my userid is "cicsos2"
Cheers
Bill
_________________
Bill Matthews |
|
| Back to top |
|
|
| JoePanjang |
| Posted: Thu Jun 01, 2006 3:20 pm Post subject: |
|
|
Voyager
Joined: 10 Jul 2002
Posts: 88
Location: Dengkil MALAYSIA
|
check your channel converstion is set to Y or N. I have a similar problem last time with .NET & CICS Bridge and it was resolved when I set converstion with Y (at the channel level).
hope this will help you.. 
_________________
Every good deed is charity... |
|
| Back to top |
|
|
| wschutz |
| Posted: Thu Jun 01, 2006 3:29 pm Post subject: |
|
|
Jedi Knight
Joined: 02 Jun 2005
Posts: 3316
Location: IBM (retired)
|
except that conversion (either gmo_convert or channel conversion) does NOT convert the msgID or correlID, they are both MQBYTE fields.
_________________
-wayne |
|
| Back to top |
|
|
| Fmoosakutty |
| Posted: Mon Jun 05, 2006 9:26 pm Post subject: Thanks |
|
|
Newbie
Joined: 26 May 2006
Posts: 2
|
Thaks for all of your sugessions and its it was due to some
problems in cics trigger monitor.I am starting a new question on same. |
|
| Back to top |
|
|
|
|
