Author |
Message
|
MonkeyDoo |
Posted: Wed Sep 03, 2014 4:47 am Post subject: PSID and dataset name |
|
|
Novice
Joined: 05 Aug 2013 Posts: 17
|
OK, I see the define, alt and delete but I don't see how to display what dataset backs a PSID. What is the magic?
I need to verify STGCLASS points to the right PSID, points to the right dataset, data set has X amount of space.
BTW, how much is a cylinder? GB please.
When I was a pup, it depended on the device... |
|
Back to top |
|
 |
MQsysprog |
Posted: Wed Sep 03, 2014 6:13 am Post subject: |
|
|
Centurion
Joined: 24 Feb 2014 Posts: 116
|
A idcams listc on the linear dataset of the pageset, will give you the required information. |
|
Back to top |
|
 |
MonkeyDoo |
Posted: Wed Sep 03, 2014 8:45 am Post subject: |
|
|
Novice
Joined: 05 Aug 2013 Posts: 17
|
Moderator, please move this to mainframe forum. |
|
Back to top |
|
 |
Vitor |
Posted: Wed Sep 03, 2014 9:04 am Post subject: |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
MonkeyDoo wrote: |
Moderator, please move this to mainframe forum. |
Done.  _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
gbaddeley |
Posted: Wed Sep 03, 2014 3:40 pm Post subject: |
|
|
 Jedi Knight
Joined: 25 Mar 2003 Posts: 2538 Location: Melbourne, Australia
|
The page datasets are identified using DEFINE PSID commands in the CSQINP1 initialization dataset. The CSQINP1 dsname(s) should appear in the proc JCL for the MQ MSTR address space. The JCL and commands should be viewable in SDSF. _________________ Glenn |
|
Back to top |
|
 |
MonkeyDoo |
Posted: Mon Sep 08, 2014 6:22 am Post subject: |
|
|
Novice
Joined: 05 Aug 2013 Posts: 17
|
OK, I see them after:
Code: |
//******************************************************************
//* PAGE SET DATA SETS *
//* *
//* This sample shows four page set data sets. *
//* You must have page set 00. *
//****************************************************************** |
Seems a little clunky... |
|
Back to top |
|
 |
bruce2359 |
Posted: Mon Sep 08, 2014 8:43 am Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
MonkeyDoo wrote: |
OK, I see them after:
Code: |
//******************************************************************
//* PAGE SET DATA SETS *
//* *
//* This sample shows four page set data sets. *
//* You must have page set 00. *
//****************************************************************** |
Seems a little clunky... |
Exactly what were you expecting? A voice reading the dataset name out loud? _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
MonkeyDoo |
Posted: Mon Sep 08, 2014 12:13 pm Post subject: |
|
|
Novice
Joined: 05 Aug 2013 Posts: 17
|
I was expecting an MQ object definition. Not punched cards. |
|
Back to top |
|
 |
Vitor |
Posted: Mon Sep 08, 2014 12:23 pm Post subject: |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
MonkeyDoo wrote: |
Not punched cards. |
Hey, they're electronically rendered punched cards in a PDS; it's progress.
Why should WMQ be the only z/OS software that doesn't have it's definitions in 80 column format? Everything works like this. You'll be claiming JCL should be object orientated next....
MonkeyDoo wrote: |
how much is a cylinder? GB please.
When I was a pup, it depended on the device... |
At about the same point in time I was a full grown dog and even today it still depends on the device. Especially as many of the DASD devices don't actually have 15 spinning plattens any more. Start either here or speak to the people running your SMS software.
Actually, speak to them first. You may find you've less control over how much space you can allocate than you think. You may well find that your calculations are fruitless because you can't control the device type the allocation is performed on. _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
bruce2359 |
Posted: Mon Sep 08, 2014 2:45 pm Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
MonkeyDoo wrote: |
I was expecting an MQ object definition. Not punched cards. |
I guess this is Windows/UNIX think: If I keep navigating down the filesystem, I'll find something/everything there is to find. _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
MonkeyDoo |
Posted: Tue Sep 09, 2014 4:27 am Post subject: |
|
|
Novice
Joined: 05 Aug 2013 Posts: 17
|
Perhaps I'm not so up on z/OS stuff.
I see now that Buffer Pools are connected to the PSID. The PSID is connected to the STGCLASS. And the STGCLASS is part of the Queue definitions.
Like anything mechanical, if you get something wrong in this linkage, it won't work (well, not the way you want).
So, I see Buffer Pools are rather important to get right (lesson learned with pain). There is a suggestion in the KC and I've made those changes. Need to observe for a little bit to make sure it is affective.
I would guess this is equivalent to UNIX stanza LogBufferPages? |
|
Back to top |
|
 |
MQsysprog |
Posted: Tue Sep 09, 2014 4:51 am Post subject: |
|
|
Centurion
Joined: 24 Feb 2014 Posts: 116
|
Hi ,
The log buffer is a real storage area used by the logger to store the log records before actually writing them to disk on the log file.
The size of the log buffer is given in pages of 4k ,on logbufferpages .
The buffers pool are used to access the messages in memory ,and the
scope is minimizing i/o activities to the pagesets .
A buffer pool could map several pagesets ,depending on the definitions
on the stgclass .
The critical resource is always the pageset 0 ,which hosts the Mq internal catalog ,and therefore normally uses a dedicated buffer pool on its own .
You will find a good amount of documentation on these topics on MQ performance reports site :
http://www-01.ibm.com/support/docview.wss?rs=171&uid=swg27007150&loc=en_US&cs=utf-8&lang=en |
|
Back to top |
|
 |
Vitor |
Posted: Tue Sep 09, 2014 4:54 am Post subject: |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
MonkeyDoo wrote: |
I would guess this is equivalent to UNIX stanza LogBufferPages? |
No. Used for different things.
The z/OS implementation doesn't map onto the distributed architecture in the same way most z/OS software doesn't map to their distributed equivalents. Witness WMQ's cousin DB2, which is arranged a completely different way on z/OS (for good & sufficient reason) to DB2 on a distributed platform, even though they look identical from the front where an application is looking. _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
bruce2359 |
Posted: Tue Sep 09, 2014 5:31 am Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
MQsysprog wrote: |
Hi ,
The log buffer is a real storage area ... |
A small nit: Everything is virtual in z/OS.
MQ buffer pools are allocated in virtual storage in the qmgr virtual address space. Yes, virtual pages are backed by real storage (RAM-equivalent). _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
MQsysprog |
Posted: Tue Sep 09, 2014 11:05 pm Post subject: |
|
|
Centurion
Joined: 24 Feb 2014 Posts: 116
|
Oh yes Bruce2359 ,it was a typo in my mind ,i was tinking of Unix main memory when i wrote the post...
With the new V8 for Z/OS, the buffers could be allocated over the 32 bit line to the 64 bit ,so at last the old low limit of 1.6 Gb addressing
below is removed .
With the Buffers on 64 bit, there is also a big increase in the number of buffer pools available for the pagesets in the range from 0 to 99 . |
|
Back to top |
|
 |
|