| Author |
Message
|
| wschutz |
 Posted: Thu Sep 29, 2005 6:46 am Post subject: |
 |
|
Jedi Knight
Joined: 02 Jun 2005
Posts: 3316
Location: IBM (retired)
|
The term "CSD" (Corrective Service Diskette) is no longer in use, for what should be obvious reasons.
_________________
-wayne |
|
| Back to top |
|
 |
| jefflowrey |
| Posted: Thu Sep 29, 2005 6:50 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
Sure, I'd expect that it would be FixPack 1 of MQ 6.0. Or is everything going to be Refresh Packs instead?
But there weren't any APARs listed on that page he posted...
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| ramires |
| Posted: Thu Sep 29, 2005 6:52 am Post subject: |
|
|
Knight
Joined: 24 Jun 2001
Posts: 523
Location: Portugal - Lisboa
|
Jeff,
sorry to ask... but what do you mean by "placeholder page"... english isn't my primary language |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Thu Sep 29, 2005 6:56 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
I meant "a page that exists only to hold stuff later".
The page you listed looked like it was a skeleton for a page, with more data to come later.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| ramires |
| Posted: Thu Sep 29, 2005 7:12 am Post subject: |
|
|
Knight
Joined: 24 Jun 2001
Posts: 523
Location: Portugal - Lisboa
|
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Thu Sep 29, 2005 12:29 pm Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7723
|
For what its worth, we ran some tests today against my Linux 6.0 QM. We thought we were OK, as with amqsgetc, MO71 and some Java apps connected in client mode the QM came down cleanly. But then we added some VB.NET apps to the mix. endmqm hung. Something with the .NET clients, because they were the only ones that seemed to hang the endmqm.
The solution was to kill -9 the endmqm from another telnet session, whch allowed our shutdown script to complete. At that point ps-ef|grep mq showed the amqrmppa process orphaned. We kill -9 that, and the QM is down for good. It starts back up normally. Not a show stopper, but a pain in the arse nevertheless.
I'm gonna open up a ticket to see if I can get an Interim fix ahead of CSD1.
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| wschutz |
| Posted: Thu Sep 29, 2005 2:30 pm Post subject: |
|
|
Jedi Knight
Joined: 02 Jun 2005
Posts: 3316
Location: IBM (retired)
|
| PeterPotkay wrote: |
For what its worth, we ran some tests today against my Linux 6.0 QM. We thought we were OK, as with amqsgetc, MO71 and some Java apps connected in client mode the QM came down cleanly. But then we added some VB.NET apps to the mix. endmqm hung. Something with the .NET clients, because they were the only ones that seemed to hang the endmqm.
The solution was to kill -9 the endmqm from another telnet session, whch allowed our shutdown script to complete. At that point ps-ef|grep mq showed the amqrmppa process orphaned. We kill -9 that, and the QM is down for good. It starts back up normally. Not a show stopper, but a pain in the arse nevertheless.
I'm gonna open up a ticket to see if I can get an Interim fix ahead of CSD1. |
Peter...were they executing an mq api call at the time? I can repro this problem with amsputc hanging on console input. Once I enter text (ie MQPUT), the qmgr shuts down.
_________________
-wayne |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Thu Sep 29, 2005 2:39 pm Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7723
|
The apps (I wrote them) were either doing an MQGET with wait, or an MQPUT under syncpoint but before the MQCMIT.
We thought, Oooh, its the outstanding MQGET with wait that is doing it, but we could not reproduce it with amqsgetc.
Just go this responce for my PMR:
| Quote: |
I looked into defect 93717, and the change team has said that it is dependant on too many changes, and wont be available until the release of CSD01. They have suggested avoiding the "-i" until the CSD is available.
|
So avoid using -i, and have the QM wait until all clients decide to end (i.e. forever) or use -i, and wait forever. 
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| bbburson |
| Posted: Fri Sep 30, 2005 1:12 pm Post subject: |
|
|
Partisan
Joined: 06 Jan 2004
Posts: 378
Location: Nowhere near a queue manager
|
| PeterPotkay wrote: |
| So avoid using -i, and have the QM wait until all clients decide to end (i.e. forever) or use -i, and wait forever. |
What about -p? Is it still available to force the queue manager to come down? My shutdown script for v5.3 goes through endmqm, endmqm -i, endmqm -p, kill -9, allowing a certain amount of time at each stage before proceeding. |
|
| Back to top |
|
|
| wschutz |
| Posted: Fri Sep 30, 2005 2:44 pm Post subject: |
|
|
Jedi Knight
Joined: 02 Jun 2005
Posts: 3316
Location: IBM (retired)
|
| bbburson wrote: |
| PeterPotkay wrote: |
| So avoid using -i, and have the QM wait until all clients decide to end (i.e. forever) or use -i, and wait forever. |
What about -p? Is it still available to force the queue manager to come down? My shutdown script for v5.3 goes through endmqm, endmqm -i, endmqm -p, kill -9, allowing a certain amount of time at each stage before proceeding. |
endmqm -p is still available. A quick test here shows that it too hangs in the same circumstances that -i hangs
_________________
-wayne |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Fri Sep 30, 2005 2:55 pm Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7723
|
I even tried to do an endmqlsr -f -m QMNAME before trying the endmqm, thinking it might help. Not.
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| hopsala |
| Posted: Fri Sep 30, 2005 5:58 pm Post subject: |
|
|
Guardian
Joined: 24 Sep 2004
Posts: 960
|
This is the funniest thread ever "the story of the MQ Admin who couldn't".
It's the new-and-improved IBM approach to EAI - MQ v6.0 is so bloody amazing, you'll never want it to end  |
|
| Back to top |
|
|
| EddieA |
| Posted: Fri Sep 30, 2005 10:20 pm Post subject: |
|
|
Jedi
Joined: 28 Jun 2001
Posts: 2453
Location: Los Angeles
|
| Quote: |
| MQ v6.0 is so bloody amazing, you'll never want it to end |
That's the problem, you can't end it. Haven't you been reading the last few posts. 
Cheers,
_________________
Eddie Atherton
IBM Certified Solution Developer - WebSphere Message Broker V6.1
IBM Certified Solution Developer - WebSphere Message Broker V7.0 |
|
| Back to top |
|
|
| SAFraser |
| Posted: Mon Oct 03, 2005 1:19 pm Post subject: |
|
|
Shaman
Joined: 22 Oct 2003
Posts: 742
Location: Austin, Texas, USA
|
1. Love the term "refresh pack". Is that like calling a defect an "undocumented feature"?
2. I had just about decided to go ahead with v6.0 on a development server..... But I guess I should do so only if I never want it to end. <sigh> I'll let you all know how it goes when I screw up the courage to try it.
This has been a great thread.
Shirley |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Mon Oct 03, 2005 1:26 pm Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7723
|
my pmr was just updated and they said this endmqm bug is only present in Linux and AS/400.
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
|
|
