| Author |
Message
|
| HugoB |
 Posted: Thu Feb 20, 2003 8:08 am Post subject: SYSTEM.DEF.CLUSSDR always in retry |
 |
|
Acolyte
Joined: 26 Jun 2001
Posts: 67
|
When i ask the channel status of the QMGR , i always get the
result that the SYSTEM.DEF.CLUSSDR channel is in retry.
Here is the AMQERR01.LOG extract of it;
02/20/03 07:10:15
AMQ9519: Channel 'SYSTEM.DEF.CLUSSDR' not found.
EXPLANATION:
The requested operation failed because the program could not find a definition
of channel 'SYSTEM.DEF.CLUSSDR'.
ACTION:
Check that the name is specified correctly and the channel definition is
available.
-------------------------------------------------------------------------------
02/20/03 07:10:16
AMQ9002: Channel program started.
EXPLANATION:
Channel program 'SYSTEM.DEF.CLUSSDR' started.
ACTION:
None.
-------------------------------------------------------------------------------
02/20/03 07:10:16
AMQ9203: A configuration error for TCP/IP occurred.
EXPLANATION:
Error in configuration for communications to host ''. Allocation of a TCP/IP
conversation to host '' was not possible.
ACTION:
The configuration error may be one of the following:
1.If the communications protocol is LU 6.2, it may be that one of the
transmission parameters (Mode, or TP Name) is incorrect. Correct the error
and try again. The mode name should be the same as the mode defined on host
''. The TP name on '' should be defined.
2.If the communications protocol is LU 6.2, it may be that an LU 6.2 session
has not been established. Contact your systems administrator.
3.If the communications protocol is TCP/IP, it may be that the host name
specified is incorrect. Correct the error and try again.
4.If the communications protocol is TCP/IP, it may be that the host name
specified cannot be resolved to a network address. The host name may not be
in the nameserver.
The return code from the TCP/IP (gethostbyname) call was 78 (X'4E').
Record the error values and tell the system administrator.
-------------------------------------------------------------------------------
02/20/03 07:10:16
AMQ9999: Channel program ended abnormally.
EXPLANATION:
Channel program 'SYSTEM.DEF.CLUSSDR' ended abnormally.
ACTION:
Look at previous error messages for channel program 'SYSTEM.DEF.CLUSSDR' in the
error files to determine the cause of the failure.
-------------------------------------------------------------------------------
The weird thing is that there is nothing on the cluster transmit queue.
Evenmore weird, we don't use any MQ Cluster functionality. !!!
Should i rebuild all the default objects again of the QMGR, with the
crtmqm -c command or similar ??
BTW all the other channels do work perfect, so no TCP/IP issue to
my opinion !!
Ah oh yes almost forgot, AIX MQ 5.2 + CSD05
ThnX All |
|
| Back to top |
|
 |
| bbeardsley |
| Posted: Thu Feb 20, 2003 8:57 am Post subject: |
|
|
Acolyte
Joined: 17 Dec 2001
Posts: 52
Location: Dallas, TX, USA
|
The first thing I would do is issue a command to stop the channel. Then investigate why it tried starting up in the first place.
If you find that your qmgr is part of a cluster and you want to remove it, there is a section in the cluster manual that will walk you through it. (Chapter 9)
-Bridgette Beardsley |
|
| Back to top |
|
|
| HugoB |
| Posted: Fri Feb 21, 2003 12:43 am Post subject: |
|
|
Acolyte
Joined: 26 Jun 2001
Posts: 67
|
Thnx Bridgette,
I did issue a stop channel of course.
And guess what, it did stop. And it even dit not try to start again so far.
I will verify it in a moment.
But now i want to know why it tried to start.
Maybe i should stop the qmgr, and restart it, to see whether
it will try to start it up again.
But as i said, non of the qmgrs is participating in a cluster.
We do not use clustering as of yet.
And yes extracting a qmgrs out of a cluster is a delicate piece of
work, otherwise you risk messing up the reposittory databases. |
|
| Back to top |
|
|
| TonyD |
| Posted: Tue Feb 25, 2003 1:55 pm Post subject: |
|
|
Knight
Joined: 15 May 2001
Posts: 540
Location: New Zealand
|
| Several people have had this problem, always so far in a clustering environment, but I have never seen an explanation. Stopping the channel and leaving it stopped, since it is never actually used for traffic, seems to be the best way around it; at least this stops amqerr01.log filling up with unwanted messages. |
|
| Back to top |
|
|
| mrlinux |
| Posted: Thu Feb 27, 2003 11:52 am Post subject: |
|
|
Grand Master
Joined: 14 Feb 2002
Posts: 1261
Location: Detroit,MI USA
|
Well if a channel is manually stopped it will never start without being manually restarted with START CHANNEL or from MQ Explorer
_________________
Jeff
IBM Certified Developer MQSeries
IBM Certified Specialist MQSeries
IBM Certified Solutions Expert MQSeries |
|
| Back to top |
|
|
| HugoB |
| Posted: Fri Feb 28, 2003 1:05 am Post subject: |
|
|
Acolyte
Joined: 26 Jun 2001
Posts: 67
|
Well i found the problem, at least the most probable reaon.
They have, during an emergency, started all channels they could find
in the system.
So they did not know what they were doing.
Since they don't use any clustering i leave them stopped now.
And during a maintenance interval it well be solved i hope.
Thnx all |
|
| Back to top |
|
|
| rjl_state |
| Posted: Fri Feb 28, 2003 8:31 am Post subject: |
|
|
Apprentice
Joined: 04 Oct 2002
Posts: 48
Location: Des Moines, IA
|
I am afraid that won't fix it. I don't have a solution, I just know that my Cluster channel that I still don't know how it got started, is just always in a stopped state. But since I don't use Clusters, it is not a problem.
The only solution I can come up with is to go ahead and delete the QMGR and recreate everything from scratch.
Unless someone knows if you can delete the channel itself and then add it back in. Of course if you don't use clusters, why add it back in. |
|
| Back to top |
|
|
| HugoB |
| Posted: Mon Mar 03, 2003 4:04 am Post subject: |
|
|
Acolyte
Joined: 26 Jun 2001
Posts: 67
|
Yes a rebuild was my intentio too in a maintenance interval.
You could delete the channels and keep the rest as it is.
And then do a strmqm -c <QMGR>, this wil start the queuemanager, recreate the default objects, en it will stop.
Then just do a strmqm <QMGR>
Ah yes of course the first step is to stop the queuemanager.
Another option is to dltmqm the queuemanger and do a full
manual rebuild. |
|
| Back to top |
|
|
|
|
