ASG
IBM
Zystems
Cressida
Icon
Netflexity
 
  MQSeries.net
Search  Search       Tech Exchange      Education      Certifications      Library      Info Center      SupportPacs      LinkedIn  Search  Search                                                                   FAQ  FAQ   Usergroups  Usergroups
 
Register  ::  Log in Log in to check your private messages
 
RSS Feed - WebSphere MQ Support RSS Feed - Message Broker Support

MQSeries.net Forum Index » General IBM MQ Support » Multi-instance Qmgr issue (svrconn)

Post new topic  Reply to topic
 Multi-instance Qmgr issue (svrconn) « View previous topic :: View next topic » 
Author Message
Gideon
PostPosted: Thu Aug 29, 2013 7:54 am    Post subject: Multi-instance Qmgr issue (svrconn) Reply with quote

Chevalier

Joined: 18 Aug 2009
Posts: 403

I was assigned an NFSv4 share, which passed all the MQ
verification tests (POSIX, locking, and concurrent writes).
I then created a multi-instance qmgr.

I have a java test program using a connectlist to send to either the Active or Standby Instance. This program can connect to the MI qmgr with either server being the active
instance. While I have not yet tested the failover with a "endmqm -s -r" option yet, the qmgr has been working
flawlessly, until now

I was running a test, and sent 20,000 messages to the qmgr. It was running very well and had sent many messages, but I got this message from the server

Code:
From STANDBY instance:
-bash-4.1$ 08/29/2013 05:24:18 AM AMQ9999: Channel 'SYSTEM.ADMIN.SVRCONN' to host '10.14.100.2 (10.14.100.2)'
ended abnormally.
08/29/2013 05:24:18 AM AMQ9508: Program cannot connect to the queue manager.
08/29/2013 05:25:42 AM AMQ9999: Channel 'SYSTEM.ADMIN.SVRCONN' to host '10.14.100.2 (10.14.100.2)'
ended abnormally.
08/29/2013 05:25:42 AM AMQ9508: Program cannot connect to the queue manager.


So the channel went down for some reason, and the clients can no longer connect. AND I get this message from the standby instance, not the active instance. Strange

So I go to the Active instance (btvm2) and try to end the qmgr instance, or the qmgr itself


Code:
-bash-4.1$ hostname
btvm2
-bash-4.1$
-bash-4.1$ dspmq -x -o standby -o status
QMNAME(WQM1)                                              STATUS(Status not available) STANDBY(Not applicable)
QMNAME(mqha2)                                             STATUS(Running elsewhere) STANDBY(Permitted)
    INSTANCE(btvm2) MODE(Active)
    INSTANCE(btvm1) MODE(Standby)
-bash-4.1$
-bash-4.1$ endmqm -x mqha2
AMQ8146: WebSphere MQ queue manager not available.
-bash-4.1$
-bash-4.1$ dspmq
QMNAME(WQM1)                                              STATUS(Status not available)
QMNAME(mqha2)                                             STATUS(Running elsewhere)
-bash-4.1$


So the dspmq long status, says the btvm2 is active, but when I try to end the qmgr, it says its not available, and when I display the sort status, it says its running elsewhere

Kind of looks like it did a partial failover

I go to the standby instance, and see:

Code:
-bash-4.1$ hsotnae
-bash: hsotnae: command not found
-bash-4.1$
-bash-4.1$ hostname
btvm1
-bash-4.1$
-bash-4.1$ dspmq
QMNAME(WQM1)                                              STATUS(Status not available)
QMNAME(mqha2)                                             STATUS(Running as standby)
-bash-4.1$



So the standby thinks he is the standby, and the active is confused. And I can not end the qmgr in any fashion from the active instnace


Code:
-bash-4.1$ endmqm -i mqha2
AMQ8146: WebSphere MQ queue manager not available.
-bash-4.1$


How can I recover from this, any ideas on what to do next?

Thanks
Back to top
View user's profile Send private message Send e-mail
JosephGramig
PostPosted: Thu Aug 29, 2013 8:47 am    Post subject: Reply with quote

Grand Master

Joined: 09 Feb 2006
Posts: 1244
Location: Gold Coast of Florida, USA

Did you try this?
http://pic.dhe.ibm.com/infocenter/wmqv7/v7r5/topic/com.ibm.mq.con.doc/q017880_.htm
Back to top
View user's profile Send private message AIM Address
Gideon
PostPosted: Thu Aug 29, 2013 9:16 am    Post subject: Reply with quote

Chevalier

Joined: 18 Aug 2009
Posts: 403

I tried it on the Active, no luck:


Code:
-bash-4.1$ ./killqmgr.sh mqha2 60
Attempting immediate end of queue manager 'mqha2'
AMQ8146: WebSphere MQ queue manager not available.
Waiting for immediate end of queue manager 'mqha2'
Waiting for immediate end of queue manager 'mqha2'
Waiting for immediate end of queue manager 'mqha2'
Waiting for immediate end of queue manager 'mqha2'
Waiting for immediate end of queue manager 'mqha2'
Waiting for immediate end of queue manager 'mqha2'
Waiting for immediate end of queue manager 'mqha2'
Waiting for immediate end of queue manager 'mqha2'
Waiting for immediate end of queue manager 'mqha2'
Waiting for immediate end of queue manager 'mqha2'
Attempting brutal end of queue manager 'mqha2'
[1]+  Killed                  runmqlsr -m mqha2 -t tcp -p 1414
-bash-4.1$ dspmq
QMNAME(WQM1)                                              STATUS(Status not available)
QMNAME(mqha2)                                             STATUS(Running elsewhere)
-bash-4.1$
-bash-4.1$ endmqm mqha2
AMQ8146: WebSphere MQ queue manager not available.



I tried it on the standby, better luck:

Code:
-bash-4.1$ ./killqmgr.sh mqha2 60
Attempting immediate end of queue manager 'mqha2'
WebSphere MQ queue manager 'mqha2' ending.
WebSphere MQ queue manager 'mqha2' ended.
Waiting for immediate end of queue manager 'mqha2'
-bash-4.1$
-bash-4.1$ dspmq
QMNAME(WQM1)                                              STATUS(Status not available)
QMNAME(mqha2)                                             STATUS(Ended immediately)


I can now bring up the qmgr in both active and standby modes

THanks, I appreciate all of your help
Back to top
View user's profile Send private message Send e-mail
JosephGramig
PostPosted: Thu Aug 29, 2013 9:42 am    Post subject: Reply with quote

Grand Master

Joined: 09 Feb 2006
Posts: 1244
Location: Gold Coast of Florida, USA

You probably always want to execute that script (or one modified to your liking) every time the machine is shutdown. If you don't, you will get FDCs about your Qmgr being shutdown because your machine was shutdown.

I would add logging to trace the effective and real user that caused invocation.

Consider what you want to do on start up... Probably auto starting the Qmgr is not the right answer for a MI or HA configuration.
Back to top
View user's profile Send private message AIM Address
Display posts from previous:   
Post new topic  Reply to topic Page 1 of 1

MQSeries.net Forum Index » General IBM MQ Support » Multi-instance Qmgr issue (svrconn)
Jump to:  



You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
Protected by Anti-Spam ACP
 
 


Theme by Dustin Baccetti
Powered by phpBB © 2001, 2002 phpBB Group

Copyright © MQSeries.net. All rights reserved.