| Author |
Message
|
| mar |
 Posted: Wed Jan 25, 2006 8:01 am Post subject: AMQ9526 error handling |
 |
|
Apprentice
Joined: 18 Oct 2004
Posts: 26
|
Hi.
I would like to ask the following:
When I get the AMQ9526: Message sequence number error for channel 'XXXX', how can I handle it without user intervention to reset the channel to the specific message sequence number?
I know it seems silly to ask somethning like this but has anyone handle a situation like this, without his intervention? A script maybe, an event monitor etc.
Thank you in advance for your help. |
|
| Back to top |
|
 |
| tkane |
| Posted: Wed Jan 25, 2006 12:05 pm Post subject: |
|
|
Voyager
Joined: 23 Dec 2002
Posts: 82
Location: Kansas City
|
I don't think that you really want to do this. I've been working with MQ for about 8 years now and way back in the late 90's we used to do that kind of thing. I've learned a lot about MQ since then and MQ has improved 100x what it was back then. And I have almost 100 x as many machines running MQ as I did back then.
Ask yourself why are you getting channel sequence errors?
Are there 2 machines trying to use the same channel name? We see this on occasion when we have a DR drill and the OS people bring up our zOS qmgrs that somehow try and connect to our other queue managers.
Was a queue manger rebuilt?
Those are the most frequent reasons why I have seen channel sequence errors. I've got over 400 queue managers on Unix and zOS and a handfull on Windows. Although I don't do monitoring, I haven't heard from Operations in a long time that they had a sequence error.
Good Luck
Tom |
|
| Back to top |
|
|
| mvic |
| Posted: Wed Jan 25, 2006 5:01 pm Post subject: |
|
|
Jedi
Joined: 09 Mar 2004
Posts: 2080
|
| tkane wrote: |
| Ask yourself why are you getting channel sequence errors? |
I agree with Tom; the reason why MQ does not solve this automatically is that this is a situation that needs manual intervention. |
|
| Back to top |
|
|
| mar |
| Posted: Thu Jan 26, 2006 2:28 am Post subject: |
|
|
Apprentice
Joined: 18 Oct 2004
Posts: 26
|
Thanks both of you for replying me.
I have a windows queue manager and a remote CICS queue manager. I don't have access to the CICS one.
I have searched all the cases I could think. I was assured that in the remote queue manager they don't do any changes although I heard that they use a JCL every hour to start and stop the channels because their CICS is so old that does not support channel triggering.
The last case is instant network failure.
Anyway, you mentioned:
-----
I've been working with MQ for about 8 years now and way back in the late 90's we used to do that kind of thing.
------
How did you used to do that?
Thanks a lot in advance. |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Thu Jan 26, 2006 4:50 pm Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
| mar wrote: |
| I was assured that in the remote queue manager they don't do any changes although I heard that they use a JCL every hour to start and stop the channels because their CICS is so old that does not support channel triggering. |
That makes zero sense. Why do they stop a channel every hour just to start it? If they do something as silly as that, then are they also doing something as silly as resetting the sequence #s every time too?
What version of MQ do they have that doesn't support channel triggering? And if that is truly the case, then just tell them to set DISCINT to 0, start the channel, and be done with it.
You automatically resetting sequence #s is just masking one problem, and sure to cause other ones eventually. If sequence #s could be automatically reset, MQ would just do it for you anyway, or not even have them in the first place.
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| kevinf2349 |
| Posted: Thu Jan 26, 2006 6:52 pm Post subject: |
|
|
Grand Master
Joined: 28 Feb 2003
Posts: 1311
Location: USA
|
| Quote: |
| What version of MQ do they have that doesn't support channel triggering? |
Peter IIRC 5.3.0 on z/OS didn't. We had to write our own trigger monitor to do it. I think it has been able to trigger channels for a good few years now though.
Could be that thye are on the older release....however, I totally agree with you about why stop the channel just to restart it.
Sounds like another case of MQ ignorance winning the day on the behalf of the mainframe folks.  |
|
| Back to top |
|
|
| mar |
| Posted: Wed Feb 08, 2006 11:54 pm Post subject: new info |
|
|
Apprentice
Joined: 18 Oct 2004
Posts: 26
|
Thanks all for replying me.
New evidence is found in application log just before the channel ends abnormally.
An unexpected error occurred in communications.
The return code from the CPI-C (cmrcv) call was 27 (X'1B'). Record these values and tell the systems administrator.
SNA APPC Application Information None 3017 CPI-C conversation_ID unknown, 01000000
SNA APPC Application Information None 91 APPC conversation failure:
TP_ID = 00000000B018C700
Conv_ID = 00C719A8
EXPLANATION
APPC conversation failed because of a connection failure. The connection failure is reported to the local TP with a return code of AP_CONV_FAILURE_RETRY.
One possible cause of this error is that the partner TP issued a TP_ENDED(HARD).
Considered that LU6.2 channels are of conversation type (as far I know), could that be a reason for getting message sequence error after channels restart?
Any opinion is appreciated. Thanks in advance. |
|
| Back to top |
|
|
|
|
