In fa4b9472-7337-4ad5-a094-0458df706...@comcast.net, on 12/17/2011
at 12:59 AM, Ed Gould edgould1...@comcast.net said:
I was remembering MVS.
A 64 KiB region sounds very small for MVS. Could that have been OS/360
but the conditional GETMAIN for 1 MiB been MVS?
--
Shmuel (Seymour J.)
Back in the day, IEBCOPY used to do a conditional getmain for 1meg (1.5
meg?).
if it got the storage, SYSUT3 and SYSUT4 were not used, thus reducing
I/O and elapsed time.
Also (at the time) IEBCOPY would run in a region much smaller than 1
meg. ISTR 64K, but am not positive.
I do not know if
In 45e5f2f45d7878458ee5ca679697335502e25...@usdaexch01.kbm1.loc, on
12/16/2011
at 07:27 AM, Staller, Allan allan.stal...@kbmg.com said:
Also (at the time) IEBCOPY would run in a region much smaller than 1
meg. ISTR 64K, but am not positive.
That would not have been in the same timeframe as a
Seymour:
I was remembering MVS. I was using numbers in my original reply from
SMF and IEFACTRT accounting in sysout. I don't recall if a
conditional getmain would show up as used. I do not think so as I
specifically remember seeing sizes less than 320K (I want to say
256K) on the job
Has anyone compared the performance of the new IEBCOPY in z/OS 1.13 with
the older version?
Regards,
John K
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message
Has anyone compared the performance of the new IEBCOPY in z/OS 1.13 with
the
older version?
Regards,
John K
John
At Share Aug 2011 there was a session dedicated to IEBCOPY and it had some
comparison data in there. You might want to check that out.
Lizette
: יום ד 14 דצמבר 2011 15:25
To: IBM-MAIN@bama.ua.edu
Subject: Re: IEBCOPY in z/OS 1.13
Has anyone compared the performance of the new IEBCOPY in z/OS 1.13
with
the
older version?
Regards,
John K
John
At Share Aug 2011 there was a session dedicated to IEBCOPY and it had some
comparison data
On Wed, 14 Dec 2011 15:57:21 +0200, Yifat Oren wrote:
Thank you for the tip, Lizette.
http://share.confex.com/share/117/webprogram/Handout/Session9940/SHARE%20994
0_IEBCOPY%20New%20Tricks.pdf
Grrr... Wrap!
Gil, I can't speak for all of them of course, but:
IEBCOPY warning messages - Issue W and/or E suffix messages for severity 4
errors. Show return code
for each COPY/COPYMOD.
Is still an open requirement in the share system.
http://reqs4.share.org/Display.jsp?r=SSMVSE11069
SSMVSE11069
Status:
On Wed, 14 Dec 2011 11:07:22 -0600, Mary Anne Matyaz wrote:
Gil, I can't speak for all of them of course, but:
IEBCOPY warning messages - Issue W and/or E suffix messages for severity 4
errors. Show return code
for each COPY/COPYMOD.
Is still an open requirement in the share system.
Gil, That's what was in Volume 7, where the IEBCOPY messages are, so that's
what I used. FYI, even though that requirement shows up as mine, it was
originally written by someone else. I simply rewrote it as part of the Share
requirements committee. IBM's response had indicated that the message
John, I ran a few tests today and I'm seeing pretty good results, similar to
what are in the presentation. Here's some data on one job that copies five
large and one small PDS:
T 12.02.3312.05.48 3 min 15 seconds
J12.11.13 12.15.46 4 min 35 seconds
MA
Mary Anne Matyaz wrote:
Gil, That's what was in Volume 7, where the IEBCOPY messages are, so that's
what I used. FYI, even though that requirement shows up as mine, it was
originally written by someone else. I simply rewrote it as part of the Share
requirements committee. IBM's response had
Mary Anne:
Way back when and continuing forward I always put on parm=size=1000K
on all IEBCOPY (especially in SMPE) and got great throughput usually
50 percent (or more) faster runtimes. I think I started doing this
in the 1970's. I know we had to use it when copying the CDS's (when
On Wed, 14 Dec 2011 17:35:26 -0600, Ed Gould wrote:
Way back when and continuing forward I always put on parm=size=1000K
on all IEBCOPY (especially in SMPE) and got great throughput usually
50 percent (or more) faster runtimes. I think I started doing this
in the 1970's. I know we had to use it
I honestly don't think there was a default as the region it ran in I
think was 256K. Giving it plenty of memory on the parm really did
help out. I know my SMPE runs ran faster than any elses in the group.
I also set up a larger size (4096K,1024K) on the link/binder runs and
SMPE just ran
16 matches
Mail list logo