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

CPU reduction while using ADRDSSU back up


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

New User


Joined: 07 Oct 2006
Posts: 6
Location: Chennai

PostPosted: Wed May 02, 2007 3:10 pm
Reply with quote

Hi,
I take a back up of around 1500 datasets using ADRDSSU backup step into TAPE, which is consuming large CPU times. Are there any options to reduce the CPU consumption while using ADRDSSU step.
Back to top
View user's profile Send private message
expat

Global Moderator


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

PostPosted: Wed May 02, 2007 3:43 pm
Reply with quote

What do you call large CPU times ?
Back to top
View user's profile Send private message
Tamanna

New User


Joined: 07 Oct 2006
Posts: 6
Location: Chennai

PostPosted: Wed May 02, 2007 4:09 pm
Reply with quote

By Large CPU times , i meant to say the TOTAL CPU TIME taken by the job
Back to top
View user's profile Send private message
expat

Global Moderator


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

PostPosted: Wed May 02, 2007 4:14 pm
Reply with quote

Quote:
By Large CPU times , i meant to say the TOTAL CPU TIME taken by the job

Yes, I rather gathered that, so I did mean for you to quote actual CPU consumption figures and the elapsed time too.

Have you considered the size of these 1,500 datasets.
Are they different HLQ's.
What paramteres have you specified in your DUMP job.
Back to top
View user's profile Send private message
Tamanna

New User


Joined: 07 Oct 2006
Posts: 6
Location: Chennai

PostPosted: Wed May 02, 2007 4:38 pm
Reply with quote

Hi expat,


The backup job i was using takes the backup of all the PS and VSAM files i have created. Please find the jobs below( where TSTPS : corresponds to PS files HLQ and TSTTV corresponds to VSAM files HLQ)

//SB1PR02 EXEC PGM=ADRDSSU
//SYSPRINT DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//TAPE1 DD DSN=TSTPS.ONSITE.D070501,
// DISP=(NEW,CATLG,DELETE),
// LABEL=RETPD=15,
// UNIT=TAPE3,
// VOL=(,,,15)
//SB1PR02.SYSIN DD *
DUMP OUTDD(TAPE1) -
TOL(ENQF) -
DS(INCL(TSTPS.** -
TSTTV.** ))
/*
Please find the CPU statistics below:

TOTAL CPU TIME= 3.71
TOTAL ELAPSED TIME= 183.95
Back to top
View user's profile Send private message
expat

Global Moderator


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

PostPosted: Wed May 02, 2007 4:52 pm
Reply with quote

Doing a quick calculation, you are only using about 2% CPU during your job.

Do you really mean that the elapsed time is too long ?

I would try adding two parameters to your job

OPTIMIZE(4)
Which reads one cylinder at a time, and may help speed your job up.

SPHERE
Always advisable for VSAM datasets, as this includes any associated AIX and PATHs for the base cluster. However, this may in fact slow down your processing.

If you are lucky and have either a product called FAVER (CA ) or FDR ( Innovation Data Processing) installed, use those in preference to DFdss. These will far outperform Dfdss, But then again so would my grandmother icon_confused.gif

FAVER is specifically for VSAM datasets and is a bloody marvellous product which offers so much flexibility and spped of processing.

FDR is a generic DASD backup/dump/restore tool, and once again goes much faster than DFdss.
Back to top
View user's profile Send private message
HappySrinu

Active User


Joined: 22 Jan 2008
Posts: 194
Location: India

PostPosted: Fri Jun 06, 2008 11:11 am
Reply with quote

Hi Expat,,
know it's very late reply but when searching about FDR i come across this post

We got FDR in our shop.

Can you provide me sample jcl, documentation on it to copy large number of datasets from one volume to another which doesn't have Cicon_biggrin.gif link.
Back to top
View user's profile Send private message
expat

Global Moderator


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

PostPosted: Fri Jun 06, 2008 11:31 am
Reply with quote

HappySrinu wrote:
Hi Expat,,
know it's very late reply but when searching about FDR i come across this post

We got FDR in our shop.

Can you provide me sample jcl, documentation on it to copy large number of datasets from one volume to another which doesn't have Cicon_biggrin.gif link.

No ...... I use DFdss here icon_biggrin.gif
Back to top
View user's profile Send private message
gcicchet

Senior Member


Joined: 28 Jul 2006
Posts: 1702
Location: Australia

PostPosted: Fri Jun 06, 2008 4:30 pm
Reply with quote

Hi Tamanna,

can you tell how long from the time the job commenced executing to the time it accepted the tape, ie.it started writing to it.

It's very time consuming when masks/hlq are used to dump files.



Gerry
Back to top
View user's profile Send private message
Pedro

Global Moderator


Joined: 01 Sep 2006
Posts: 2546
Location: Silicon Valley

PostPosted: Fri Jun 06, 2008 10:20 pm
Reply with quote

Quote:
1,500 datasets.

How much storage are you talking about? The number of datasets does not tell enough. If each one is very large, perhaps you are actually getting good performance.
Back to top
View user's profile Send private message
Pedro

Global Moderator


Joined: 01 Sep 2006
Posts: 2546
Location: Silicon Valley

PostPosted: Sat Jun 07, 2008 3:22 am
Reply with quote

Quote:
// VOL=(,,,15)

Your JCL allows for 15 tapes. How many are actually used? If all are used, that is about 12 minutes per tape, including mounting and rewinding, which I think is fairly good.
Back to top
View user's profile Send private message
Douglas Wilder

Active User


Joined: 28 Nov 2006
Posts: 305
Location: Deerfield IL

PostPosted: Sat Jun 07, 2008 3:35 am
Reply with quote

The question was asked over a year ago. Do you think Tamanna is still waiting for the answer?
Back to top
View user's profile Send private message
dick scherrer

Moderator Emeritus


Joined: 23 Nov 2006
Posts: 19244
Location: Inside the Matrix

PostPosted: Sat Jun 07, 2008 5:28 am
Reply with quote

Hello,

Quote:
Do you think Tamanna is still waiting for the answer?
No, this is a new-life-for-the-topic discussion with a bit of the traffic on the new question and some on the original icon_smile.gif

I probably should have split this, but haven't yet. . . Just being slack. . .
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 How do I SFTP a ADRDSSU file from mai... IBM Tools 4
No new posts Back Page print direction (Duplex Pri... JCL & VSAM 3
No new posts adrdssu for copy all data JCL & VSAM 5
No new posts adrdssu tool to dump a files from a PDS JCL & VSAM 5
No new posts ADRDSSU BY(REFDT... not selecting cor... IBM Tools 3
Search our Forums:

Back to Top