Author |
Message
|
sumansengupta07 |
Posted: Fri Nov 15, 2013 3:28 am Post subject: If any transaction got abended what happened to that message |
|
|
Novice
Joined: 14 Nov 2013 Posts: 23
|
In MQ if a queue triggers any transaction and the corresponding transaction got abended then what happened to that message? |
|
Back to top |
|
 |
Vitor |
Posted: Fri Nov 15, 2013 6:02 am Post subject: Re: If any transaction got abended what happened to that mes |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
sumansengupta07 wrote: |
In MQ if a queue triggers any transaction and the corresponding transaction got abended then what happened to that message? |
That depends on your transaction handling, doesn't it?  _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
sumansengupta07 |
Posted: Fri Nov 15, 2013 7:48 am Post subject: |
|
|
Novice
Joined: 14 Nov 2013 Posts: 23
|
If transaction got abended then it will be backed out and all the resources will get released. According to me, it should go back to that local queue from where it comes from if backout queue is not defined?? Correct me if i m wrong??????? |
|
Back to top |
|
 |
Vitor |
Posted: Fri Nov 15, 2013 8:08 am Post subject: |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
sumansengupta07 wrote: |
If transaction got abended then it will be backed out and all the resources will get released. According to me, it should go back to that local queue from where it comes from if backout queue is not defined?? Correct me if i m wrong??????? |
If an MQ transaction abends and you take no action it'll roll onto the original queue even if a backout queue is defined unless you have some kind of transaction handler or handling in play. _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
sumansengupta07 |
Posted: Fri Nov 15, 2013 8:16 am Post subject: |
|
|
Novice
Joined: 14 Nov 2013 Posts: 23
|
But if any transaction abends consecutively 3 times in 3 separate MQ call then 3 messages will go back again to the original queue. And if we create a backout queue with threshold value 3 then only 1 message will go to that backout queue. By defining this backout queue, storage space is less consumed in mq. Am I r8 or wrong?? |
|
Back to top |
|
 |
mqjeff |
Posted: Fri Nov 15, 2013 8:18 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008 Posts: 17447
|
You are right.
Or you are wrong.
Have you considered experimenting? |
|
Back to top |
|
 |
Vitor |
Posted: Fri Nov 15, 2013 9:41 am Post subject: |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
sumansengupta07 wrote: |
Am I r8 or wrong?? |
You're wrong about it consuming less space.
You're wrong about how WMQ handles transactional boundaries.
You're wrong about what backout threshold means.
You're wrong about how backout threshold is used. _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
bruce2359 |
Posted: Fri Nov 15, 2013 10:14 am Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
Are you asking about a CICS transaction? An IMS transaction? Some other kind of transaction? _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
sumansengupta07 |
Posted: Sun Nov 17, 2013 6:09 pm Post subject: |
|
|
Novice
Joined: 14 Nov 2013 Posts: 23
|
then can you please tell me the logic means how it works??? |
|
Back to top |
|
 |
bruce2359 |
Posted: Sun Nov 17, 2013 8:31 pm Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
sumansengupta07 wrote: |
then can you please tell me the logic means how it works??? |
How what works?
Do you know what a transaction is? Search here and Google for 'syncpoint' and 'unit of recovery' and 'unit of work'.
Search also for 'computer+acid test'. _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
elkinsc |
Posted: Tue Dec 03, 2013 9:27 am Post subject: Chiming in late - as usual |
|
|
 Centurion
Joined: 29 Dec 2004 Posts: 138 Location: Indy
|
If a transaction under the control of the resource managers available on z/OS (CICS, IMS, RRS) abends and one or more messages have been put or gotten and are under syncpoint control, then they will be backed out. If the messages have been processed with the no syncpoint option, they are not.
By themselves the transaction managers on z/OS (CICS, RRS, IMS) do not do anything with the backout count of backout re-queue queue. They expect the application to check the count and determine if the messages should be moved to the re-queue queue (poisoned or politician messages).
If you are using the CICS bridge to initiate the transaction, the action is different. The CICS Bridge, which is a supplied CICS and MQ application program, does honor the the backout count and re-queue queue. If an error occurs, the message that initiated the transaction is put to the specified re-queue queue. Please see the CICS InfoCenter for additional information (CICS supplies the code). A link to the CICS 4.1 description of the behavior can be found here:
http://pic.dhe.ibm.com/infocenter/cicsts/v4r1/topic/com.ibm.cics.ts.wmq.bridge.doc/topics/zs11430_.html
What you have asked is very broad, and the response to those types of questions is always 'it depends'. |
|
Back to top |
|
 |
|