| Author |
Message
|
| EB |
 Posted: Tue Aug 10, 2004 11:19 pm Post subject: Message sequence error after IPL |
 |
|
Acolyte
Joined: 19 Mar 2004
Posts: 70
|
Hi
For some time now we experience problems with channels after IPL. After IPL when the channel internally in mainframe between LPARs and between distributed and a specific LPAR are to resume connection they fail and sender channel goes into retry.
Investigations show that the reason is inconsistencies in message sequence (see logmessages below).
This leads me to believe that at some point before or after IPL channels are reset on this specific LPAR.
This is where I am stuck, how do I continue the search? Is there a parameter which is reset within MQ?
Any advice is greatly apreciated.
-------------------------------------------------------------------------------
> CSQX526E ^ CSQXRCTL Message sequence error for channel MRQ5.MRT5,
> sent=1 expected=18
> CSQX506E ^ CSQXRCTL Message receipt confirmation not received for
> channel MRQ5.MRT5
> CSQX599E ^ CSQXRCTL Channel MRQ5.MRT5 ended abnormally
-------------------------------------------------------------------------------
Regards
EB |
|
| Back to top |
|
 |
| offshore |
| Posted: Wed Aug 11, 2004 3:52 am Post subject: |
|
|
Master
Joined: 20 Jun 2002
Posts: 222
|
EB,
IBM has a APAR (APAR Identifier ...... PQ48057) that could be of benefit to you.
Depending on what version of MQ series you are using.
Offshore |
|
| Back to top |
|
|
| EB |
| Posted: Wed Aug 11, 2004 4:00 am Post subject: |
|
|
Acolyte
Joined: 19 Mar 2004
Posts: 70
|
Thanks for your reply offshore.
We are using WMQ 5.3.1 on z/OS 1.4, on distributed systems we are mainly running WMQ 5.3.06 and WMQ 5.3.07 althoug on some servers we still run MQ 5.2.
Could I bother you to explain in more detail how APAR Identifier could benefit us in the matter at hand?
Thanks alot.
EB |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Wed Aug 11, 2004 5:26 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
An APAR is an Authorized Problem Analysis Report. (or some such)
offshore is telling you that IBM may already have coded a fix for your problem, and if you use the APAR identifier to search IBM's support site, you will find the relevant information to that particular problem and it's fix (if there is one), and will be able to determine for yourself if this is the same problem you are having.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| offshore |
| Posted: Wed Aug 11, 2004 8:54 am Post subject: |
|
|
Master
Joined: 20 Jun 2002
Posts: 222
|
EB,
The APAR I mentioned isn't going help. That APAR works with 5.2.0, but now that you gave a little more detail I see you're running 5.3.1.
I mentioned it(after doing a quick search) that it help with channel sequence errors after and IPL.
In the MVS QMGR startup proc do you have any members in the CSQINP2 step? I read on IBM's site that if you have the SYSTEM.DEF.RECEIVER channel being defined again, that is with the REPLACE option then it resets the chl seq numbers over each time. By replacing the exisiting definition the flags get overwritten, so MQ has no idea what number it should be on.
Do a search on IBM's site & here on mqseries.net for CSQX506E and CSQX526E see what hits come back.
OFFSHORE |
|
| Back to top |
|
|
| JT |
| Posted: Wed Aug 11, 2004 3:25 pm Post subject: |
|
|
Padawan
Joined: 27 Mar 2003
Posts: 1564
Location: Hartford, CT.
|
EB,
Does this out-of-sync condition also occur when the queue manager is stopped and re-started, not just at IPL time?
Does the queue manager have a script that re-creates the MQ resources at start-up? I worked at a shop where they did this for their mainframe queue managers. |
|
| Back to top |
|
|
| EB |
| Posted: Wed Aug 11, 2004 11:53 pm Post subject: |
|
|
Acolyte
Joined: 19 Mar 2004
Posts: 70
|
Hi
Thanks for the replies.
Offshore:
There are members but none with REPLACE only with DEFINE which results in Channel already exist.
JT:
Yes, the offsynch situation also occurs when restarting the QMGR not only during IPL.
Further advice is greatly apreciated.
Regards
EB |
|
| Back to top |
|
|
| JT |
| Posted: Thu Aug 12, 2004 7:21 am Post subject: |
|
|
Padawan
Joined: 27 Mar 2003
Posts: 1564
Location: Hartford, CT.
|
EB,
What, if anything, do you have defined to the JCL statement CSQINP2 ? |
|
| Back to top |
|
|
| EB |
| Posted: Fri Aug 13, 2004 12:19 am Post subject: |
|
|
Acolyte
Joined: 19 Mar 2004
Posts: 70
|
There are no definitions made in CSQINP2 members.
Regards
EB |
|
| Back to top |
|
|
| tkane |
| Posted: Fri Aug 13, 2004 12:43 pm Post subject: |
|
|
Voyager
Joined: 23 Dec 2002
Posts: 82
Location: Kansas City
|
There is one more place to look similar to INP2.
CSQINPX is a DD in the CHIN.
You can put commands in there as well, and the output will show up in your CHIN STC.
Good Luck
Tom |
|
| Back to top |
|
|
| EB |
| Posted: Sun Aug 15, 2004 10:22 pm Post subject: |
|
|
Acolyte
Joined: 19 Mar 2004
Posts: 70
|
Hi
Thanks for the suggestion. There was only one command in CSQINPX and that was display channels after initiator starts, which should not result in resetting the channels.
Regards
EB |
|
| Back to top |
|
|
| ulli |
| Posted: Mon Aug 16, 2004 1:59 am Post subject: |
|
|
Novice
Joined: 16 Aug 2004
Posts: 20
|
Hi,
there is another IBM Apar PQ82416, maybe it is helpful. Did you check this one already?
Regards, Ulli |
|
| Back to top |
|
|
| EB |
| Posted: Tue Aug 31, 2004 12:34 am Post subject: Help needed |
|
|
Acolyte
Joined: 19 Mar 2004
Posts: 70
|
Hi again
I am still struggling with this issue.
I think I have found something that might cause the problem mentioned earlier.
The setup guide states:
----------------------------------------------------------
You must include the required queue definitions in your WebSphere MQ queue
manager. Distributed queuing requires a queue for use with sequence numbers
and
logical units of work identifiers (LUWID). You must ensure that a queue is
available with the name SYSTEM.CHANNEL.SEQNO.
----------------------------------------------------------
The SYSTEM.CHANNEL.SEQNO queue is not defined in any off my mainframe QMGRs thus I suspect that after IPL the channels have no other option than to start at seqno 1.
Does this make sense? And if so should this queue not have been created on installation?
Assuming I am correct, I think the next logical step would be to manually create the queue on my QMGRs. Can anyone help me with the settings this queue should have (i.e. persistence, depth etc)?
Any advice is GREATLY appreciated.
Regards
EB
|
|
| Back to top |
|
|
| ulli |
| Posted: Tue Aug 31, 2004 2:19 am Post subject: |
|
|
Novice
Joined: 16 Aug 2004
Posts: 20
|
Hi,
are you using CICS ICS for distributed queuing? Because the mentioned queue SYSTEM.CHANNEL.SEQNO ist just requiered if you are using CICS ICS - as described in the setup guide.
Did you check your SYSTEM.CHANNEL.SYNCQ for duplicate messages? |
|
| Back to top |
|
|
| EB |
| Posted: Tue Aug 31, 2004 2:52 am Post subject: |
|
|
Acolyte
Joined: 19 Mar 2004
Posts: 70
|
Hi Ulli
No we are not using CICS ICS.
What implications will I experience when deleting from SYSTEM.CHANNEL.SYNCQ ?
The problem occurs on every QMGRs on mainframe, it's unlikely that there should be problems with the SYSTEM.CHANNEL.SYNCQ on every QMGR is it not?
Regards
EB |
|
| Back to top |
|
|
|
|
