CPU costs for single logging vs dual logging

Tom Glaser

CPU costs for single logging vs dual logging
We are currently performing dual logging with DB2 V7 z/OS to dasd, company
wide. We have a group who would like to move the test/development
subsystems (we have 40 test subsystems) to single logging, to save on CPU
and costs associated with migrating to tape.

Does anyone either know, or know where to look, or how to perform a test to
show the actual overhead associated with using dual logs vs single
logging? We would like to be able to say that by switching from dual
logging to single logging, we can save $nn.nn. However, if there is very
little overhead associated with dual logging, then we would like to keep
test the same as production, that is, using dual logs.

Thanks,
Tom Glaser
DB2 Systems Support
SBC Communications
(314) 340-0980
[login to unmask email]

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Roger Miller

Re: CPU costs for single logging vs dual logging
(in response to Tom Glaser)
You'd probably save more in tape costs than in cpu. I'd be very surprised
if you could measure a cpu difference for single versus dual logs - maybe
one tenth of one percent. The difference between two identical runs is
often different by 1%.

I'd push you toward archiving to disk to avoid the tapes. The other work
that would be required (and help in production too) would be to find and
fix the long-running units of recovery. Keep the amount of disk space for
archive logs small, maybe even deleting instead of archiving. For the
runaway situations, it would be necessary to have a backup and restore
setup. FlashCopy and the V8 subsystem backup and restore utilities might
be good for that.

Roger Miller

On Fri, 14 Jan 2005 09:02:14 -0600, Tom Glaser <[login to unmask email]> wrote:

>We are currently performing dual logging with DB2 V7 z/OS to dasd, company
>wide. We have a group who would like to move the test/development
>subsystems (we have 40 test subsystems) to single logging, to save on CPU
>and costs associated with migrating to tape.
>
>Does anyone either know, or know where to look, or how to perform a test
to
>show the actual overhead associated with using dual logs vs single
>logging? We would like to be able to say that by switching from dual
>logging to single logging, we can save $nn.nn. However, if there is very
>little overhead associated with dual logging, then we would like to keep
>test the same as production, that is, using dual logs.
>
>Thanks,
>Tom Glaser
>DB2 Systems Support
>SBC Communications
>(314) 340-0980
>[login to unmask email]
>

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm