View previous topic :: View next topic
|
Author |
Message |
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Hi,
If I use these two different SORT cards on a single data set...
Code: |
SORT FIELDS=(01,10,CH,A,
11,10,CH,A,
21,08,CH,A,
29,02,CH,A) |
After modification :
Code: |
SORT FIELDS=(01,30,CH,A) |
Does that make any difference in CPU cost of JOB?
P.S.: I tried to test it but with a small data it does not show really a difference. But here at my work place a debate is going on that 'it might'
...please suggest. My shop uses Syncsort, but I think this functionality should be same irrespective of SORT product so post is in DFSORT forum. |
|
Back to top |
|
|
Frank Yaeger
DFSORT Developer
Joined: 15 Feb 2005 Posts: 7129 Location: San Jose, CA
|
|
|
|
I wouldn't expect it to make any difference for DFSORT as we combine contiguous fields internally when possible as in this case. |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Thanks for the confirmation Frank. However, can we've some technical explanation for this instead of running JOBs with these SORT cards & then compairing their respective costs? |
|
Back to top |
|
|
Frank Yaeger
DFSORT Developer
Joined: 15 Feb 2005 Posts: 7129 Location: San Jose, CA
|
|
|
|
I don't know what you mean by "some technical explanation". I thought I gave a technical explanation that "we combine contiguous fields internally". What I mean is that the "key" is handled in the same way (as one contiguous 30-byte key) for both SORT statements. Of course, I can only speak for DFSORT whereas you said you're using Syncsort. |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hi Anuj,
A technical answer is in Frank's reply
Quote: |
we combine contiguous fields internally when possible as in this case. |
Before the actual sorting is started, the code looks at the sort fields and when they are contiguous, it internally creates a single sort field (just like the shorter syntax that was posted), rather than working with 4 (from your example). |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Frank Yaeger wrote: |
"we combine contiguous fields internally". |
I was beaten by the word "we".. I could not corelate it with SORT ..apologies.. "kernal King" .
Quote: |
What I mean is that the "key" is handled in the same way (as one contiguous 30-byte key) for both SORT statements. Of course, I can only speak for DFSORT whereas you said you're using Syncsort. |
Thanks for explaing again, got it now.
Thankyou Dick, I got the answer now to deal with my counter parts at my work place.. . Let see what they have in store to say now... |
|
Back to top |
|
|
Alissa Margulies
SYNCSORT Support
Joined: 25 Jul 2007 Posts: 496 Location: USA
|
|
|
|
Anuj,
In the example you provided, you should not see any difference in CPU usage in SyncSort either. |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Thankyou Alissa for confirming.. |
|
Back to top |
|
|
|