IBM Mainframe Forum Index
 
Log In
 
IBM Mainframe Forum Index Mainframe: Search IBM Mainframe Forum: FAQ Register
 

how to find the cpu utilization in OS/390 machine?


IBM Mainframe Forums -> All Other Mainframe Topics
Post new topic   Reply to topic
View previous topic :: View next topic  
Author Message
sathish kumar

New User


Joined: 06 Jul 2004
Posts: 9
Location: chennai

PostPosted: Thu Sep 01, 2005 9:43 pm
Reply with quote

Hi

How to find the CPU utilization in OS/390 Machine?

If there is any command to display the cpu utilization.

help me friends
sathish
Back to top
View user's profile Send private message
priyesh.agrawal

Senior Member


Joined: 28 Mar 2005
Posts: 1448
Location: Chicago, IL

PostPosted: Thu Sep 01, 2005 10:32 pm
Reply with quote

Hi Sathish,

Cant we see it in S.DA....

There is a column CPU% showing the jobs involving CPU.

Regards,

Priyesh.
Back to top
View user's profile Send private message
spanda

New User


Joined: 31 Aug 2005
Posts: 48
Location: U.K.

PostPosted: Fri Sep 02, 2005 5:30 pm
Reply with quote

For active jobs, the CPU time utilised may be seen on the DA panel of SDSF, under the CPU% column (as already said by priyesh). For completed jobs, the CPU usage can be seen in the JESMSGLG dataset at the bottom of the data. The message normally appears with the words 'JOB END' followed by the job name and the job parameters.

Hope this helps.
Panda.
Back to top
View user's profile Send private message
Ravi gaur

New User


Joined: 12 Jul 2005
Posts: 38

PostPosted: Mon Sep 05, 2005 3:34 pm
Reply with quote

There are lot of ways to know the CPU utilisation.

One of them is RMF..

If you want long Performance report of CPU use Postprocessor of RMF.

for this u need to collect record type 70:78 of smf which are RMF type..

Let me know If you need anything else.


These days i m working on the same project
Back to top
View user's profile Send private message
sundayblooms

New User


Joined: 09 Jun 2006
Posts: 4
Location: India

PostPosted: Thu Sep 28, 2006 5:41 pm
Reply with quote

Hi ravi , can you tell me how do we switch on the record types in SMF
Back to top
View user's profile Send private message
kcs4u2004
Warnings : 1

New User


Joined: 21 Apr 2008
Posts: 56
Location: Chennai

PostPosted: Wed Oct 06, 2010 11:40 am
Reply with quote

I want a email once the cpu utilization reaches 90%..can anybody please help me on this regard..
Thanks in advance..it wud be of great help to us and wud be highly appreciated..

Thanks..
KCS
Back to top
View user's profile Send private message
expat

Global Moderator


Joined: 14 Mar 2007
Posts: 8797
Location: Welsh Wales

PostPosted: Wed Oct 06, 2010 11:55 am
Reply with quote

sundayblooms wrote:
Hi ravi , can you tell me how do we switch on the record types in SMF
You ask your systems admin people to do this.

kcs4u2004 wrote:
want a email once the cpu utilization reaches 90%..can anybody please help me on this regard..
Thanks in advance..it wud be of great help to us and wud be highly appreciated..
And why would that be, why not let you performance and capacity planning people do their jobs and produce usage and projection reports.

So what are you going to do when cpu utilisation reaches 90%, please do tell.
Back to top
View user's profile Send private message
PeD

Active User


Joined: 26 Nov 2005
Posts: 459
Location: Belgium

PostPosted: Wed Oct 06, 2010 11:56 am
Reply with quote

Try this ( I will not code all of your tool )

Code:
        /* REXX */                                                     
/*  TRACE ?R   */                                                       
/*                                                                     
NUMERIC DIGITS 12                                                       
*/                                                                     
CVT  = C2D(STORAGE(10,4))                                               
RMCT     = C2D(STORAGE(D2X(CVT+604),4))      /* POINT TO RMCT        */
RCT      = C2D(STORAGE(D2X(RMCT+228),4))   /* RESOURCE CONTROL TBL */   
RCTLACS  = C2D(STORAGE(D2X(RCT+196),4))    /* 4 HR MSU AVERAGE     */   
RCTIMGWU = C2D(STORAGE(D2X(RCT+28),4))     /* IMAGE DEFINED MSUS   */   
RCTCECWU = C2D(STORAGE(D2X(RCT+32),4))     /* CEC MSU CAPACITY     */   
SAY  'THE DEFINED MSU CAPACITY FOR THIS LPAR IS' RCTIMGWU'.'           
SAY  'THE 4 HOUR MSU AVERAGE USAGE IS' RCTLACS'.'                       
                                                                       
EXIT 0                                                                 
Back to top
View user's profile Send private message
expat

Global Moderator


Joined: 14 Mar 2007
Posts: 8797
Location: Welsh Wales

PostPosted: Wed Oct 06, 2010 12:08 pm
Reply with quote

Thanks PeD,

Just cut and pasted that one

Fancy a beer one evening ?
Back to top
View user's profile Send private message
PeD

Active User


Joined: 26 Nov 2005
Posts: 459
Location: Belgium

PostPosted: Wed Oct 06, 2010 12:15 pm
Reply with quote

Why not, with pleasure, the unique problem is I am based at Braine.
But when I "have" to come to Brussels I will drop you a note.
Still at Fortis ......BNP........?
Back to top
View user's profile Send private message
expat

Global Moderator


Joined: 14 Mar 2007
Posts: 8797
Location: Welsh Wales

PostPosted: Wed Oct 06, 2010 12:21 pm
Reply with quote

Yeah, drop me a note and I'll happily buy you a beer or three icon_lol.gif
Back to top
View user's profile Send private message
kcs4u2004
Warnings : 1

New User


Joined: 21 Apr 2008
Posts: 56
Location: Chennai

PostPosted: Wed Oct 06, 2010 12:24 pm
Reply with quote

if the cpu utilization is close to 100% i.e some jobs are taking all the cpu resource,so that other jobs are not getting processed..we have some time dependent jobs,hence if the jobs will not get processed accordingly in time we will get a major(severity issue)..hence if we will get an eamil once it reaches more than 95%,so that we can be alert that something may go wrong adn will ask the operation to HOLD the job manually so that it wont be abended..
Hence need the code for an alert email if cpu utilization is more than 95%
Back to top
View user's profile Send private message
expat

Global Moderator


Joined: 14 Mar 2007
Posts: 8797
Location: Welsh Wales

PostPosted: Wed Oct 06, 2010 12:40 pm
Reply with quote

It sounds to me that your system needs adjustment to allow higher priority jobs to be given the required resources when cpu utilisation is heavy.

Why fight the fire when you can put it out sooner
Back to top
View user's profile Send private message
dbzTHEdinosauer

Global Moderator


Joined: 20 Oct 2006
Posts: 6966
Location: porcelain throne

PostPosted: Wed Oct 06, 2010 2:25 pm
Reply with quote

Most good shops always run about 95 to 100%,
otherwise they are paying for hardware that they are not using.

they have people in capacity planning and operations that know their jobs,
are proficient,
and with the proper input from departments
are able to plan and implement the necessary priorities
as well as scheduling
to accomplish this.

putting out fires in a poorly planned/managed environment
is a never ending task.
it is impossible when everyone is sitting around
using resources to monitor the system
using resources to generate eMails - that may or may not be read -
to request that tweaks be made to assure that their application runs.

this is typical bureaucratic nonsense.
instead of doing their job in the first place,
they figure out how to cya.
Back to top
View user's profile Send private message
Robert Sample

Global Moderator


Joined: 06 Jun 2008
Posts: 8696
Location: Dubuque, Iowa, USA

PostPosted: Wed Oct 06, 2010 4:54 pm
Reply with quote

Your problem is not that you need an email when CPU utilization exceeds 95% (or whatever arbitrary number you want to use). Your problem is that your Workload Manager (WLM) policy is not set up correctly to recognize the time sensitive jobs. You need to work with your site support group to update the WLM policy.

Furthermore, your site should already have the tools in place to handle this issue without adding to the system overhead by putting in an email notification -- your scheduling package should be flagging jobs that are late and the operators should be adjusting priority processing to get the work complete. If this is not happening, it's not a technical issue to address by a technical solution, it is a management problem that needs management attention.
Back to top
View user's profile Send private message
View previous topic :: :: View next topic  
Post new topic   Reply to topic View Bookmarks
All times are GMT + 6 Hours
Forum Index -> All Other Mainframe Topics

 


Similar Topics
Topic Forum Replies
No new posts To find whether record count are true... DFSORT/ICETOOL 6
No new posts Find the size of a PS file before rea... COBOL Programming 13
No new posts Find the occurrence of Key Field (Par... DFSORT/ICETOOL 6
No new posts Find a record count/numeric is multip... COBOL Programming 1
No new posts Need to find a specific STRING COBOL Programming 11
Search our Forums:

Back to Top