On Sat, 6 Sep 2008 12:09:46 -0700, Ron Hawkins
<[EMAIL PROTECTED]> wrote:

>Mark,
>
>I converted  SYSVTOC, SYSVVDS and SYSIGGV2 on a three way MAS back when we
>were still on parallel channels. The MIM file was on 3880-23 and really
>loved that cache, but the benefits were substantial, especially during the
>batch run hardware reserves on VTOCs and catalogs were a major pain.
>Converting SYSIGGV2 was to solve some soft embrace issues we were hitting.
>

Ron,

SYSIGGV2 is being converted.  You have to or you will run into deadly embraces
as you already saw.  This has been documented since MVS/ESA. 


>Did the same thing on a two way MAS a few years later at two sites that also
>had a lot of the same contention. One with GRS and the other with MIM using
>the Turbo option. This was on ESCON and the benefits were noticeable but not
>as substantial.
>
>I would think that converting these three Reserves would be one of the first
>things you do if you go MIM or GRS, Star or ring. While they are very high
>in the overall count of reserves, the rate is actually quite low. 1,000,000
>reserves in day is only 11 per second. I would think twice before doing this
>beyond three systems in a GRS ring though.
>
>Ron

As I said I have 8 systems involved.  I would never consider it with GRS
RING, but MIM works more like a star configuration the way data is passed
around the MIMplex.  And I agree that not all reserves are the same - some
are much more expensive than others.   I see about 50,000,000 for
SYSZVVDS on one of the busiest LPARs in this MIMplex and that system
was IPLed last weekend.   I indicated in a planning meeting that I suspect
we could live with converting those resources but I still want to be cautious
and only turn them on for 1 evening and see what happens. And then only
if I have to (we may not require the use of LDMF at all since this particular
environment has 1-2 scheduled outages a month).  

But just to illustrate that you can't just start changing reserves to ENQs at
will in an environment that is not GRS STAR (or MIM in a CF), we once
changed the STK resource name that controls access to their CDS for
HSC and VTCS.  Virtual tape processing didn't come to a halt but things
really backed up and batch fell hours behind before I figured out what
was going on.   This was in this same environment (8 systems in the
MIMplex across 2 sysplexes).   

Regards,

Mark
--
Mark Zelden
Sr. Software and Systems Architect - z/OS Team Lead
Zurich North America / Farmers Insurance Group - ZFUS G-ITO
mailto:[EMAIL PROTECTED]
z/OS Systems Programming expert at http://expertanswercenter.techtarget.com/
Mark's MVS Utilities: http://home.flash.net/~mzelden/mvsutil.html
  

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to