RE: Tape backup strategy

[Vern Hamberg <vhamberg@xxxxxxxxxxxxxxxxxxxxxxxxx>]

CPC3722 message gives time the save started - message time would be end of 
save.

To augment my earlier post that had results from a run of 1, I have the 
following on a 9406-270 2248. I bracketed the SAVLIB commands with 
WRKSYSSTS RESET(*YES) before and RESET(*NO) after, which gives elapsed time 
and CPU% over the run. Also cleared the SAVFs first, to eliminate that 
processing-probably not too important, but maybe. CPU% in mid-90's for all 
compressed runs, around 30% for non-compressed.

The first library comprises mostly PFs and DTAARAs and is 67.6 MB

SAVF                 Size     Elapsed
DTACPRNO       70,270,976    00:00:08
DTACPRYES      32,530,432    00:00:18    46%
DTACPRLOW      32,530,432    00:00:18    46%
DTACPRMED      11,554,816    00:01:05    16%   3-1/2 times as long
DTACPRHIGH      6,311,936    00:02:22     9%   2     times as long

The second library comprises PGMs, source PFs, etc., and is 2,401.2 MB

SAVF                 Size     Elapsed
DTACPRNO    2,502,967,296    00:09:30
DTACPRYES   1,524,711,424    00:11:55    60%
DTACPRLOW   1,524,711,424    00:11:57    60%
DTACPRMED     741,359,616    00:42:43    30%   4     times as long
DTACPRHIGH    664,027,136    04:01:07    27%   6     times as long

Mileage varies because different types of objects compress differently - 
programs are not going to compress as well as text data, e.g. 7 to 24 times 
as long for *HIGH over *LOW, 2 to 6 times as long for *HIGH over *MEDIUM.

I guess it all depends, whether to go all the way to *HIGH. 6 times as long 
to save 3% does not seem worth it. Jeff's case, OTOH, seems viable.

You pays your money...
Vern

At 11:24 AM 10/4/2004, you wrote:
> Hey, anything in your joblogs to indicate the start/stop time before and
> after the data compression was changed?  Want to see if the new
> compression slowed it down, so's as you would notice.  Just curious, not
> trying to compare to anything here.  Just always hearing the saw about
> performance and I want to check.
>
> Rob Berendt
> --
> Group Dekko Services, LLC
> Dept 01.073
> PO Box 2000
> Dock 108
> 6928N 400E
> Kendallville, IN 46755
> http://www.dekko.com
>
>
>
>
>
> rob@xxxxxxxxx
> Sent by: midrange-l-bounces@xxxxxxxxxxxx
> 10/04/2004 11:12 AM
> Please respond to
> Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
>
>
> To
> Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
> cc
>
> Fax to
>
> Subject
> RE: Tape backup strategy
>
>
>
>
>
>
> Glad to hear that.  The new compression is IMPRESSIVE!
>
>
> Rob Berendt
> --
> Group Dekko Services, LLC
> Dept 01.073
> PO Box 2000
> Dock 108
> 6928N 400E
> Kendallville, IN 46755
> http://www.dekko.com
>
>
>
>
>
> "Jeff Crosby" <jlcrosby@xxxxxxxxxxxxxxxx>
> Sent by: midrange-l-bounces@xxxxxxxxxxxx
> 10/04/2004 10:59 AM
> Please respond to
> Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
>
>
> To
> "'Midrange Systems Technical Discussion'" <midrange-l@xxxxxxxxxxxx>
> cc
>
> Fax to
>
> Subject
> RE: Tape backup strategy
>
>
>
>
>
>
> > True, but it's rather tough to do hardware compression on a save file.
>
> There are various areas where I do saves to a save file just prior to some
> big batch update process.  I also do a SAVLIB twice a day of our main
> database library to 2 different save files, then put these 2 save files
> onto
> a single tape.
>
> After you posted your message last week about the new options on the
> DTACPR
> parm, I changed all these to *HIGH.  Since Friday, my disk usage has gone
> from ~70% to ~52%.  And no one has complained about performance. <g>
>
> --
> Jeff Crosby
> Dilgard Frozen Foods, Inc.
> P.O. Box 13369
> Ft. Wayne, IN 46868-3369
> 260-422-7531
>
> The opinions expressed are my own and not necessarily the opinion of my
> company.  Unless I say so.
>
>
>
> --
> This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing
> list
> To post a message email: MIDRANGE-L@xxxxxxxxxxxx
> To subscribe, unsubscribe, or change list options,
> visit: http://lists.midrange.com/mailman/listinfo/midrange-l
> or email: MIDRANGE-L-request@xxxxxxxxxxxx
> Before posting, please take a moment to review the archives
> at http://archive.midrange.com/midrange-l.
>
>
> --
> This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing
> list
> To post a message email: MIDRANGE-L@xxxxxxxxxxxx
> To subscribe, unsubscribe, or change list options,
> visit: http://lists.midrange.com/mailman/listinfo/midrange-l
> or email: MIDRANGE-L-request@xxxxxxxxxxxx
> Before posting, please take a moment to review the archives
> at http://archive.midrange.com/midrange-l.
>
>
> --
> This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
> To post a message email: MIDRANGE-L@xxxxxxxxxxxx
> To subscribe, unsubscribe, or change list options,
> visit: http://lists.midrange.com/mailman/listinfo/midrange-l
> or email: MIDRANGE-L-request@xxxxxxxxxxxx
> Before posting, please take a moment to review the archives
> at http://archive.midrange.com/midrange-l.