| Author |
Message
|
| DaifaN |
 Posted: Tue Aug 06, 2013 12:21 am Post subject: Anything I missed steup ma0z for CLSSDR channel? |
 |
|
Newbie
Joined: 05 Aug 2013
Posts: 8
|
I did the following testing on 3 servers, I can receive message from testing.q QM2.QUEUE.MANAGER, everything works fine.
but no logs in /tmp/mlivt.log appeared, anything I missed?
-------------------------------------------------
mqserver1:(solaris10)
bash-3.2$ ls -l /var/mqm/exits64/
total 334
-rwxrwx--- 1 mqm mqm 1338 Aug 5 17:22 mlconfig.ivt
-rwxrwx--- 1 mqm mqm 1675 Feb 26 2010 mlconfig.sample
-rwxrwx--- 1 mqm mqm 9060 Feb 26 2010 mlutil
-rw-r--r-- 1 mqm mqm 383 Aug 5 17:22 mlutil.history
-rw-r--r-- 1 mqm mqm 722 Feb 26 2010 wmqcml.readme
-rwxrwx--- 1 mqm mqm 146152 Feb 26 2010 wmqcml64.so
bash-3.2$
In QM1.QUEUE.MANAGER:
alter CHANNEL(TO.MQ2) CHLTYPE(CLUSSDR) MSGEXIT('wmqcml64.so(wmqcml)') MSGDATA('LF=/tmp/mlivt.log;LO=0,3,5')
then restart this CLUSSDR channel
--------------------------------------------------
mqserver2: (solaris9)
In QM2.QUEUE.MANAGER, there is a CLUSRCVR channel named TO.MQ2.
I created a cluster queue in QM2.
define ql('testing.q') CLUSTER('CLUSTER')
---------------------------------------------------
client server:(solaris10)
I send message to testing.q QM1.QUEUE.MANAGER:
bash-3.2$ export MQSERVER="SYSTEM.DEF.SVRCONN/TCP/192.168.3.4(1414)"
bash-3.2$ /opt/mqm/samp/bin/amqsputc testing.q QM1.QUEUE.MANAGER
---------------------------------------------------- |
|
| Back to top |
|
 |
| exerk |
| Posted: Tue Aug 06, 2013 12:28 am Post subject: |
|
|
Jedi Council
Joined: 02 Nov 2006
Posts: 6339
|
Post moved from News/Updates...
DaifaN, please take note of where you post, thank you. And, what is it that you're trying to achieve with the message exit?
_________________
It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys. |
|
| Back to top |
|
|
| DaifaN |
| Posted: Tue Aug 06, 2013 12:49 am Post subject: |
|
|
Newbie
Joined: 05 Aug 2013
Posts: 8
|
| exerk wrote: |
Post moved from News/Updates...
DaifaN, please take note of where you post, thank you. And, what is it that you're trying to achieve with the message exit? |
Thanks exerk.
I just want to check the biggest message that passed from the CLSSDR channel. This is my first time testing message exit, I want to make sure it can works.
MQ server is 7.0.1.9, client is 7.0.1.5 |
|
| Back to top |
|
|
| exerk |
| Posted: Tue Aug 06, 2013 3:08 am Post subject: |
|
|
Jedi Council
Joined: 02 Nov 2006
Posts: 6339
|
Is your exit expecting the log to already exist, or create it if not?
_________________
It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys. |
|
| Back to top |
|
|
| JosephGramig |
| Posted: Tue Aug 06, 2013 5:48 am Post subject: |
|
|
Grand Master
Joined: 09 Feb 2006
Posts: 1244
Location: Gold Coast of Florida, USA
|
| Well, for an explicit CLUSSDR channel, it gets replaced by an auto definition from the CLUSRCVR definition at the other end. So I don't think your exit will get called. For cluster channels, you will need auto channel exits. Tricky! |
|
| Back to top |
|
|
| gbaddeley |
| Posted: Tue Aug 06, 2013 4:29 pm Post subject: |
|
|
Jedi Knight
Joined: 25 Mar 2003
Posts: 2538
Location: Melbourne, Australia
|
| exerk wrote: |
| Is your exit expecting the log to already exist, or create it if not? |
It appears the OP is using IBM SupportPac MA0Z. It creates log files as needed.
_________________
Glenn |
|
| Back to top |
|
|
| DaifaN |
| Posted: Tue Aug 06, 2013 5:36 pm Post subject: |
|
|
Newbie
Joined: 05 Aug 2013
Posts: 8
|
| JosephGramig wrote: |
| Well, for an explicit CLUSSDR channel, it gets replaced by an auto definition from the CLUSRCVR definition at the other end. So I don't think your exit will get called. For cluster channels, you will need auto channel exits. Tricky! |
Thanks for your reply.
you mean no need alter CLUSSDR channel, alter CLUSRCVR, the CLUSSDR channel will replaced automaticly, right?
As you said, I change back CLUSSDR channel and alter CLUSRCVR to add (MSGEXIT,MSGDATA), but after restart CLUSSDR channel, it can not startup, the status is retrying. |
|
| Back to top |
|
|
| DaifaN |
| Posted: Tue Aug 06, 2013 5:41 pm Post subject: |
|
|
Newbie
Joined: 05 Aug 2013
Posts: 8
|
| gbaddeley wrote: |
| exerk wrote: |
| Is your exit expecting the log to already exist, or create it if not? |
It appears the OP is using IBM SupportPac MA0Z. It creates log files as needed. |
Yes, if the log file didn't exist, it will be created. I also tried create the log file and change permission to 777, owner is mqm. Still no logs appeared. |
|
| Back to top |
|
|
| gbaddeley |
| Posted: Wed Aug 07, 2013 3:30 pm Post subject: |
|
|
Jedi Knight
Joined: 25 Mar 2003
Posts: 2538
Location: Melbourne, Australia
|
| DaifaN wrote: |
| I also tried create the log file and change permission to 777, owner is mqm. Still no logs appeared. |
Check the MQ error logs, the MCA may not be able to load the exit module. The exit produces diagnostics on the system log if it cannot create or open the exit log file. See the user guide for details.
_________________
Glenn |
|
| Back to top |
|
|
| gbaddeley |
| Posted: Wed Aug 07, 2013 3:39 pm Post subject: |
|
|
Jedi Knight
Joined: 25 Mar 2003
Posts: 2538
Location: Melbourne, Australia
|
| DaifaN wrote: |
| JosephGramig wrote: |
| Well, for an explicit CLUSSDR channel, it gets replaced by an auto definition from the CLUSRCVR definition at the other end. So I don't think your exit will get called. For cluster channels, you will need auto channel exits. Tricky! |
Thanks for your reply.
you mean no need alter CLUSSDR channel, alter CLUSRCVR, the CLUSSDR channel will replaced automaticly, right?
As you said, I change back CLUSSDR channel and alter CLUSRCVR to add (MSGEXIT,MSGDATA), but after restart CLUSSDR channel, it can not startup, the status is retrying. |
Refer to http://www.mqseries.net/phpBB2/viewtopic.php?t=48149
To use an exit on a CLUSSDR channel, it must be defined on the corresponding CLUSRCVR channel of the same name. The exit module must be installed on all hosts in the cluster and they must all use the same exit module name (eg. you can't mix Windows and UNIX in the cluster). To overcome this limitation you would need to implement a CHAD exit that modifies the channel exit atttributes of the CLUSDR channel when it is auto defined.
_________________
Glenn |
|
| Back to top |
|
|
| JosephGramig |
| Posted: Thu Aug 08, 2013 6:06 am Post subject: |
|
|
Grand Master
Joined: 09 Feb 2006
Posts: 1244
Location: Gold Coast of Florida, USA
|
| DaifaN wrote: |
| JosephGramig wrote: |
| Well, for an explicit CLUSSDR channel, it gets replaced by an auto definition from the CLUSRCVR definition at the other end. So I don't think your exit will get called. For cluster channels, you will need auto channel exits. Tricky! |
Thanks for your reply.
you mean no need alter CLUSSDR channel, alter CLUSRCVR, the CLUSSDR channel will replaced automaticly, right?
As you said, I change back CLUSSDR channel and alter CLUSRCVR to add (MSGEXIT,MSGDATA), but after restart CLUSSDR channel, it can not startup, the status is retrying. |
gbaaddeley gave you a great answer you must use for cluster channels. A CLUSSDR is only used the first time it is started or if you refresh the cluster on *this* Qmgr with REPOS(YES). Doing the refresh that way is like removing the Qmgr from the cluster and adding it back. Before you do that, make sure the one explicit CLUSSDR can start with a channel ping. You can also have an explicit CLUSSDR to the other Full Repository.
If you have more then one FR, then I'm not sure which FRs gets tried. I am sure this is against "Best Practices" and you will most likely end up with fragmented Full Repositories for the cluster. This is because an FR will only forward updates it receives to the other FRs in this cluster if it also has an explicit CLUSSDR. I'm pretty sure an FR that receives an update from another FR will not forward that update to any other FR. The "Best Practice" rule clearly came to be, because nobody could follow the rules...
Final advise about FRs, don't use them also as application Qmgrs. Build separate ones on an existing box if you cannot afford dedicated equipment and licenses.
Now that I stop and think about it, this is one of the cases where I can see the value of multiple instances of WMQ being installed... You always want the FRs to be at the highest level of version/maintenance in your infrastructure (and always go first). |
|
| Back to top |
|
|
| gbaddeley |
| Posted: Thu Aug 08, 2013 4:33 pm Post subject: |
|
|
Jedi Knight
Joined: 25 Mar 2003
Posts: 2538
Location: Melbourne, Australia
|
| JosephGramig wrote: |
| You always want the FRs to be at the highest level of version/maintenance in your infrastructure (and always go first). |
.. and on the most reliable platform. z/OS and AIX at the top of the heap, Windows at the bottom. Other UNIX platforms with HA implementations fit in between.
_________________
Glenn |
|
| Back to top |
|
|
| JosephGramig |
| Posted: Fri Aug 09, 2013 5:31 am Post subject: |
|
|
Grand Master
Joined: 09 Feb 2006
Posts: 1244
Location: Gold Coast of Florida, USA
|
| gbaddeley wrote: |
| .. and on the most reliable platform. z/OS... |
Hmmm, I would not pick z/OS Qmgrs for Full Repositories unless I had no other choice. Surely, you can afford two Linux VMs... And now I will get the "Airplane" movie wise cracks...
I'm not saying it isn't reliable. I'm saying z/OS will complicate the cluster channel configuration especially if it is in a SysPlex. |
|
| Back to top |
|
|
| Vitor |
| Posted: Fri Aug 09, 2013 5:45 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| JosephGramig wrote: |
| And now I will get the "Airplane" movie wise cracks... |
Who would do such a thing? 
| JosephGramig wrote: |
| I'm not saying it isn't reliable. I'm saying z/OS will complicate the cluster channel configuration especially if it is in a SysPlex. |
It's not that bad. Remember that you can simply use the z/OS queue manager as a reliable FR & not have it hosting any queues/applications which would cause multiple auto-generated channels into and out of z/OS.
I await incomming fire from the school of thought which says that even distributed FRs should be dedicated & not host applications. My defence is that Surely doesn't always have as much budget (or management will to spend what there is) as she'd like and occassionally you need to compromise.
The customer is always right. Even if the customer is a pointy haired, penny pinching contact admin who's wasting oxygen by breathing it.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| JosephGramig |
| Posted: Fri Aug 09, 2013 6:08 am Post subject: Ode on Isolation of Full Repositories from Late Night Pages |
|
|
Grand Master
Joined: 09 Feb 2006
Posts: 1244
Location: Gold Coast of Florida, USA
|
There was a time when pager, cell, and disrupted sleep,
The earth, and every common MQ cluster
To me did seem
Apparelled in fragmented repository,
The glory and the freshness of a REFRESH CLUSTER(...) REPOS(YES).
It is not now as it hath been of yore;--
Turn wheresoe'er I may,
By night or day,
The things which I have seen I now can see no more.
Because some IGMO did this at the FR! |
|
| Back to top |
|
|
|
|
