|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
 |
|
[mainframe] queue internals |
« View previous topic :: View next topic » |
Author |
Message
|
pgorak |
Posted: Tue Oct 21, 2003 4:37 am Post subject: [mainframe] queue internals |
|
|
 Disciple
Joined: 15 Jul 2002 Posts: 158 Location: Cracow, Poland
|
Hi all,
As we know, MQSeries queue is physically stored as a VSAM on the mainframe. Supposing we have a local queue with INDXTYPE set to NONE and several thousands of persistent messages in the queue, should this influence the performance of MQGET, when issued without MatchOptions?
My point is whether the organisation of the data set itself can possibly appear as a problem in such a situation.
I'm informed that this was the case in one of the environments I work on.
Piotr |
|
Back to top |
|
 |
pgorak |
Posted: Fri Oct 24, 2003 2:16 am Post subject: |
|
|
 Disciple
Joined: 15 Jul 2002 Posts: 158 Location: Cracow, Poland
|
It turns out that what was observed is rather connected with workload than with queue depth.
After constantly putting and getting messages on/from a particular queue, the duration of MQGET increases about ten times, on the average. My idea is that the underlying data set becomes defragmented (if I can use this word speaking of VSAM) and hence longer times are needed to obtain a message.
Surely, some of you are working with heavy-loaded mainframe environments, so perhaps someone is able to provide some useful information on this subject?
Piotr |
|
Back to top |
|
 |
LuisFer |
Posted: Mon Oct 27, 2003 3:34 am Post subject: |
|
|
 Partisan
Joined: 17 Aug 2002 Posts: 302
|
We see the same problem & we detect the problem when the VSAM takes extensions. (on V5.3.0).
LuisFer |
|
Back to top |
|
 |
pgorak |
Posted: Mon Oct 27, 2003 4:39 am Post subject: |
|
|
 Disciple
Joined: 15 Jul 2002 Posts: 158 Location: Cracow, Poland
|
Thanks for the reply. Provided this is the only reason, allocations of extensions could be avoided by defining primary size that is large enough.
Piotr |
|
Back to top |
|
 |
LuisFer |
Posted: Mon Oct 27, 2003 4:51 am Post subject: |
|
|
 Partisan
Joined: 17 Aug 2002 Posts: 302
|
yes, we redefined the PSID's with more space on the 1st exts & we see a better performance (we have about 60 Persistent Messages by second)
dzien´dobre.
PD: Sorry english is not my native lenguage.
LuisFer. |
|
Back to top |
|
 |
|
|
 |
|
Page 1 of 1 |
|
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
|
|
|
|