| Author |
Message
|
| EddieA |
 Posted: Tue Apr 19, 2005 8:10 pm Post subject: |
 |
|
Jedi
Joined: 28 Jun 2001
Posts: 2453
Location: Los Angeles
|
| Quote: |
| 64-bit AMD servers |
Why is that a problem for MQ.
| Quote: |
| 64-bit queue managers for AIX, HP-UX & Solaris |
No mention of Intel here. Move along, nothing to see.
Cheers,
_________________
Eddie Atherton
IBM Certified Solution Developer - WebSphere Message Broker V6.1
IBM Certified Solution Developer - WebSphere Message Broker V7.0 |
|
| Back to top |
|
 |
| kirani |
| Posted: Tue Apr 19, 2005 9:53 pm Post subject: |
|
|
Jedi Knight
Joined: 05 Sep 2001
Posts: 3779
Location: Torrance, CA, USA
|
Maybe I wasn't clear in my post. We were suppose to buy 64-bit AMD servers, but someone didn't like that idea. Finally, we ended up purchasing 32-bit Intel servers. Now I will not be able to use 64-bit MQ on a 32-bit servers.
_________________
Kiran
IBM Cert. Solution Designer & System Administrator - WBIMB V5
IBM Cert. Solutions Expert - WMQI
IBM Cert. Specialist - WMQI, MQSeries
IBM Cert. Developer - MQSeries
|
|
| Back to top |
|
|
| Tibor |
| Posted: Wed Apr 20, 2005 12:17 am Post subject: |
|
|
Grand Master
Joined: 20 May 2001
Posts: 1033
Location: Hungary
|
| RogerLacroix wrote: |
| Actually, since WMQ v6 cannot co-exist with v5.3, it means I will need to buy 3 new boxes (AIX, HP-UX & Solaris), since it will take companies a year or two (or longer) to upgrade to the new release. |
Just a hint: your Unix boxes can't handle any virtualization? Like LPAR on AIX.
Tibor |
|
| Back to top |
|
|
| Tibor |
| Posted: Wed Apr 20, 2005 10:48 pm Post subject: |
|
|
Grand Master
Joined: 20 May 2001
Posts: 1033
Location: Hungary
|
I read into the security manul and I'm pleased that IBM dropped the platform specific SSL handling on Windows. We had a lot of problems with it, mostly on older OSes (WinNT).
From v6 the keystore handling will be unified with GSKit7 on distributed platforms 
Tibor |
|
| Back to top |
|
|
| Nigelg |
| Posted: Wed Apr 20, 2005 11:29 pm Post subject: |
|
|
Grand Master
Joined: 02 Aug 2004
Posts: 1046
|
| Let's see how pleased you are after you have wrestled with GSKit for a while. |
|
| Back to top |
|
|
| Nigelg |
| Posted: Wed Apr 20, 2005 11:38 pm Post subject: |
|
|
Grand Master
Joined: 02 Aug 2004
Posts: 1046
|
| Quote: |
| New attributes for queues, queue manager channels, and cluster channels extend the range of cluster workload algorithms. |
Did you spot this? There are big changes so that you can have much more control over the way workload balancing is done.
For example:
It can choose local and remote queues
You can rank qmgrs so that msgs always go to higher ranked qmgrs.
You can have a weighting so that msgs can be sent to qmgrs in proportion to the weighting value.
You can limit the number of qmgrs in the cluster that a cluster qmgr can send to.
Best of all...
The workload balancing algorithm is actually published in the CLusters manual. No more speculation about how it does it! |
|
| Back to top |
|
|
| fschofer |
| Posted: Thu Apr 21, 2005 1:42 am Post subject: |
|
|
Knight
Joined: 02 Jul 2001
Posts: 524
Location: Mainz, Germany
|
| Quote: |
| 64-bit AMD servers |
Its not a question of 64-bit processors (AMD or latest version auf Pentium 4 with EMT 64).
Its a question of 32-bit or 64 bit Windows.
A 32-bit Windows on a 64-bit processor beheave like
a 32-bit Windows on a 32-bit processor.
Greetings
Frank |
|
| Back to top |
|
|
| Tibor |
| Posted: Thu Apr 21, 2005 1:47 am Post subject: |
|
|
Grand Master
Joined: 20 May 2001
Posts: 1033
Location: Hungary
|
| Nigelg wrote: |
| Let's see how pleased you are after you have wrestled with GSKit for a while. |
Nigel,
We handle the keystores with GSKit for some time past. Isn't excellent but working 
Then again I had a lot of pesky experience with the windows schannel API.
Tibor |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Thu Apr 21, 2005 3:19 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
| Nigelg wrote: |
| Let's see how pleased you are after you have wrestled with GSKit for a while. |
You mean you haven't been wanting a lot of PMRs for usability against GSKit?
I think that, regardless of how bad GSkit may or may not be, this opens up the window for centralized management of the SSL infrastructure across the enterprise.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| Tibor |
| Posted: Thu Apr 21, 2005 4:02 am Post subject: |
|
|
Grand Master
Joined: 20 May 2001
Posts: 1033
Location: Hungary
|
How can I send a failure description of MQv6? The new MQExplorer causes an FDC file on the server side:
| Code: |
+-----------------------------------------------------------------------------+
| |
| WebSphere MQ First Failure Symptom Report |
| ========================================= |
| |
| Date/Time :- Thursday April 21 13:03:37 MEST 2005 |
| Host Name :- wfmsgate (SunOS 5.9) |
| PIDS :- 5724B4103 |
| LVLS :- 530.6 CSD06 |
| Product Long Name :- WebSphere MQ for Sun Solaris |
| Vendor :- IBM |
| Probe Id :- VP009006 |
| Application Name :- MQM |
| Component :- vwb_admin |
| Build Date :- Feb 11 2004 |
| CMVC level :- p530-06-L040211 |
| Build Type :- IKAP - (Production) |
| UserID :- 00000150 (mqm) |
| Program Name :- amqpcsea |
| Process :- 00028319 |
| Thread :- 00000001 |
| QueueManager :- WFMS |
| Major Errorcode :- vrcE_BAD_VERSION |
| Minor Errorcode :- OK |
| Probe Type :- MSGAMQ7925 |
| Probe Severity :- 2 |
| Probe Description :- AMQ7925: Message version 3 is not supported. |
| FDCSequenceNumber :- 0 |
| Arith1 :- 3 3 |
| |
+-----------------------------------------------------------------------------+ |
If anyone is interesting...
Tibor |
|
| Back to top |
|
|
| Michael Dag |
| Posted: Thu Apr 21, 2005 4:10 am Post subject: |
|
|
Jedi Knight
Joined: 13 Jun 2002
Posts: 2607
Location: The Netherlands (Amsterdam)
|
|
| Back to top |
|
|
| Nigelg |
| Posted: Thu Apr 21, 2005 7:33 am Post subject: |
|
|
Grand Master
Joined: 02 Aug 2004
Posts: 1046
|
| Look at the FFST. It is produced from the command server on 5.3 CSD06. Naturally that is not going to know about new version of the MQCFH (just guessing) introduced at V6. |
|
| Back to top |
|
|
| Tibor |
| Posted: Thu Apr 21, 2005 11:15 am Post subject: |
|
|
Grand Master
Joined: 20 May 2001
Posts: 1033
Location: Hungary
|
Nigel,
I understand what caused the problem, just I want to help the mq developers.
Tibor |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Thu Apr 21, 2005 12:59 pm Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7723
|
| Nigelg wrote: |
Did you spot this? There are big changes so that you can have much more control over the way workload balancing is done.
For example:
It can choose local and remote queues
.
.
|
Finally! I never understood this limitation. Now we don't have to have a dedicated Gateway QM that can't house any local queues.
Or should we still keep that design in place, because the round robining is so intensive you are better off not requiring the QM to do anything but round robining?
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| Nigelg |
| Posted: Fri Apr 22, 2005 12:02 am Post subject: |
|
|
Grand Master
Joined: 02 Aug 2004
Posts: 1046
|
It is not so intensive. The algorithm always collected a list of all queues of the right name in the cluster, and then gave them all a score. If you ever looked at a trace, this is done in rfiChooseOne, and against each qmgr there is a hex number. The highest number wins the msg!
It is no more intensive now, i.e. it does not collect any more data, but the rules on how the data is interpreted to determine the destination are more complicated.
A gateway qmgr is still a good idea to determine the destination of msgs originating outside the cluster. |
|
| Back to top |
|
|
|
|
