Daniel, sounds like you found it. For others, MONVU and a few other
friends are included in the CMS samples. Look for a MONVIEW readme
on that disk for more details.
Bill Bitner - z/VM Customer Focus and Care - IBM Endicott - 607-429-3286
Richard, best wishes for much fun in your retirement.
Thanks for keeping us on our toes all these years. :-)
Bit
like
class 0 as there is no reason to delay it freeing up
resources.
Bill Bitner - z/VM Customer Focus and Care - IBM Endicott - 607-429-3286
are wrong
for 99% of the environments.
Sorry, my failure to get changes made there.
Bill Bitner - z/VM Customer Focus and Care - IBM Endicott - 607-429-3286
Hi Berry, I'm out of ideas without looking at data. You
may want to open a PMR.
Bill Bitner - z/VM Customer Focus and Care - IBM Endicott - 607-429-3286
will collect
by issuing a diagnose x'DC' during intialization. It's almost like
multiple bufferes got created.
Bill Bitner - z/VM Customer Focus and Care - IBM Endicott - 607-429-3286
Bill Bitner
. This is interesting for systems on
larger n-ways that want to reduce z/VM and LPAR overhead.
(Sorry sent this to wrong list yesterday. Senior moment.)
Bill Bitner
IBM z/VM Customer Care and Focus
If they are brand new to the platform, then the
following presentation might help:
http://www.vm.ibm.com/devpages/bitner/presentations/virtualb.pdf
Regards,
Bill
I still don't really understand the problem/concern,
but hopefully this answers any questions remaining.
As the documentation describes, there are roughly 2050 pages of
non-pageable structures for 1GB of vdisk space. On older releases,
z/VM 5.3, these structures were restricted below 2GB. The
Richard, Are you concerned that the default value isn't being
computed 'correctly' or are you just curious or are you concerned
that the documentation isn't correct or are you concerned that
you'll need to set an explicit limit greater than the default?
Or all of the above?
Bill Bitner
cases you seldom need more than 2GB. There are
various rules of thumb for this.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
not eligible, but that restriction
was lifted many releases ago (in the code, not in the doc).
Besides the books, you can find more at:
http://www.vm.ibm.com/perf/tips/prgmdcar.html
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
Thanks for all your contributions over the years Jeff, and wishing
you the best in the next chapter.
Bit
It depends on how you're going to use it. See the
System Limits presentation, chart 23.
Bit
Tom,
Can you send an example of your QUERY Frames to me? I can try to work
through the example with you and then post back here.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
Marty did an excellent job of summarizing the effects of mixing sizes.
Unpleasant in extreme cases yet infintely better than zero performance.
That's what Brian Wade described in the case studies at the zExpo.
/520pav.html
Bill Bitner
.
Wandschneider, Scott [EMAIL PROTECTED] writes:
I know that z/VM 5.3 supports 256GB of real storage and 8TB of virtual =
storage and expanded storage is limited to 128GB.
What does z/VM 5.2 and 4.4 support?
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
Stor Util % is a bit of a misnomer. The computation really comes
down to the sum of the estimated working set sizes of all active
virtual machines over the amount of DPA the scheduler thinks it
has available. Since estimated working set sizes can be greater than
the resident page count, values
Fred,
There are actually two flavors of CMM:
Cooperative Memory Management which is a set of APIs in Linux that
controls putting pages of a Linux guest out of use. This is often
described as a balloon technique where the balloon is inflated/deflated
to push out unneeded memory. z/VM's VMRM (VM
The idea is not unreasonable. We have discussed plan canidates in
the past that would be more graceful about page space filling up.
Not much agreement on what to do. This particular way of causing
havoc wasn't discussed. I'll pass it on to others to think about.
I don't recall having seen any
Not sure what you mean by official, but it's very likely that
you'll see more on this from IBM and through other sources.
Is there any chance IBM could possibly come out with an official
announcement or publication indicating their $250M savings by
consolidating to zSeries Linux? That might be a
netstat gate | drop 6 | count lines | cons
Thanks in advance.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
. Watch this space.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
Jim,
Rick had asked me about this and I didn't think it was normal
behaviour. I'm on the road, but have asked someone to look into
this. Could you please open a PMR so we can track this? Thanks, Bit
thought is to add something to the exit that
grabs perhaps QUERY TIME or something before at entering and
leaving the exit.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
, but all I have to offer. And would give you ballpark numbers.
Not sure you need any significant precision.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
.
Configuring such that the backup programs have
SET MDC INSERT OFF or NOMDC minidisks is generally a
good idea.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
What kind of VM system dump did you get? I looked, but didn't find
a problem record for this. Was any monitor or Performance Toolkit
history files available for the production system?
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
in xstore and act as a sort of shock absorber for
paging. In the measurements and customer data so far, 2GB has
shown to be acceptable. I'm sure there are other exceptions that
I haven't thought of yet or seen.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
allows pages to be moved out
to DASD based on 2 levels of LRU, the xstore one being time stamped.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
)
was happening in general on write I/Os. I knew there were
exceptions, but this seems broader. I'll put it on my list to
investigate. Do you recall if that was using diagnose I/O or SSCH?
If diagnose, was it recordlevel or default? Thanks.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607
blocks are involved with the write I/O
and handle appropriately. There are exceptions where
complicated write channel programs confuse MDC and for
integrity reasons, it will purge more of the cache then
it needs.
Those are my few random thoughts on this. Emphasis on
random. :-)
Bill Bitner - VM
See http://www.vm.ibm.com/perf/tips/z890.html for a little more
background.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
below 2GB would
often be suggested that one disable the xtore MDC to create
a greater chance of access register mode being used.
I tend not to just look at the hit ratio, but at the
I/Os avoided. 50% hit ratio on 1000 I/Os per second is
better than 90% hit ratio on 100 I/Os per second.
Bill Bitner
handles it with more grace (or is it mercy).
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
We are experiencing some problems with the VM Home Page.
With the flooding and vacations, we are having trouble locating
people to correct the problems. I apologize for the
inconvenience.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
filepool. The presentations:
http://www.vm.ibm.com/devpages/bitner/presentations/sfsperf.html
http://www.vm.ibm.com/devpages/bitner/presentations/sfsperf2.html
have some things that might help.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
are running on an IFL that works
everywhere at this point.
I am curious as to why you care about whether you
are running on an IFL or not?
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
estimate
the space you need for this and then also create a process to
clean up the disk full condition. The current MONWRITE does not
automatically clean up after itself.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
would
be preferable.
Bill Bitner - VM Performance Evaluation - IBM Endicott - 607-429-3286
traditional reminder: make as few changes
as possible at one time.
Bill Bitner
more interesting.
Your planning this for this weekend? (in case I think of anything
else)
Bill Bitner
44 matches
Mail list logo
