MIDRANGE dot COM Mailing List Archive



Home » MIDRANGE-L » January 2013

Re: Journaling without commitment control (WAS iSeries and SSD drives)



fixed

Y'all the benefits of journal caching can be used (for free) to some extent by something Rick Turner speaks of - it MIGHT be just turning on commitment control in a process without actually journaling - I forget the details. Whatever it is, there is a certain amount of caching going on.

Read that link I posted, as well as Charles' items, and this one.

Vern

On 1/18/2013 12:12 PM, Jack Kingsley wrote:
Rob, I am not licensed to option 42, I found this link, maybe the fact of
just doing bundle cache to main storage does the magic.

http://www-01.ibm.com/support/docview.wss?uid=nas1443fa6ceeb42ecc2862578e2004c2ac2



On Fri, Jan 18, 2013 at 12:59 PM, <rob@xxxxxxxxx> wrote:

These times are valid. Let me restate them because the commas are
confusing. These are month end times for Infor.
The tables were all journaled.
January: No SSD's. No option 42.
February: With SSD's. No option 42.
March: With SSD's. With option 42 of SS1.
Process Jan Feb March
CST600C 90 min 89 min 56 minutes
INV900C 115 min 108 min 52 minutes
of which
INV900 98 min 87 min 35 minutes

SSD's gave us some time back. But not nearly as much time as what adding
option 42 gave us. And 42 was a much lower cost.

And I have dozens of disk arms. On a 9117-MMA. That was probably yawning
and begging for something to do as far as memory and processor at that
time.

Real world times. Real world data. Real world system.

I do not have the capability to rewrite all of my vendor applications to
start using commitment control. Too often another real world situation.
And frankly, slapping on option 42 would have been much cheaper than
rewriting the thousands upon thousands of programs to use commitment
control. Granted, I could have done some Pareto analysis to determine
which programs used the most time and just did those but, again, would I
have gained the performance I did for the cost that I did? Assuming that
my devopers are not slaves that I would have had to feed and care for
regardless of whether or not there was real work to do.


Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1
Group Dekko
Dept 1600
Mail to: 2505 Dekko Drive
Garrett, IN 46738
Ship to: Dock 108
6928N 400E
Kendallville, IN 46755
http://www.dekko.com





From: Charles Wilt <charles.wilt@xxxxxxxxx>
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>,
Date: 01/18/2013 12:04 PM
Subject: Re: Journaling without commitment control (WAS iSeries and
SSD drives)
Sent by: midrange-l-bounces@xxxxxxxxxxxx



Quite possibly...

But remember, at this point it's unlikely anybody is turning on journaling
for the first time...

So really, if you were to turn off journaling your 1hr batch job might run
in 20-30min.

But are you really going to turn it off to see? So either run the
tools...or turn it on and try it free for 70 days...

Rob was impressed :)
http://archive.midrange.com/midrange-l/201004/msg00016.html

Process Jan, pre anything Feb, w/SSD March, w/opt 42 + SSD
CST600C 90 minutes 89 minutes 56 minutes
INV900C 115 minutes 108 minutes 52 minutes
of which
INV900 98 minutes 87 minutes 35 minutes

Charles


On Fri, Jan 18, 2013 at 11:45 AM, D*B <dieter.bender@xxxxxxxxxxxx> wrote:

Charles,

you are really telling (and believing) a batch application running 1 h
without journal will take 3 h, if you would start journaling for the
files??? I've never seen any numbers in this range within 20 years!!!

D*B
--
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.








Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2014 by MIDRANGE dot COM and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available here. If you have questions about this, please contact