> I'll shuddup now ;-)
I won't shuddup now! I suspect that some "Blanket statements" can turn
out to be "Wet Blanket statements", ruining the party at some customer
sites, especially at new z/VM sites running POC's on hand-me-down machines
without a lot of real memory. As always: "**It depends**!!" {probably
(c) Bill Bitner}
Saying that all Linux swap disks should be defined as VDISKs does not take
into account different customer situations.
We happen to be blessed with lots of cheap(er), hand-me-down DASD from the
500lb Gorilla. OTOH, we _could_ be short of real memory (three large z/OS
guests in a virtual sysplex on that system).
Memory can be relatively expensive compared to relatively cheap DASD
(especially when one gets "hand-me-downs").
Those secondary 'warning flag' VDISKs eat away at real, shared memory. If
you have a lot of Linux guests and not a lot of spare memory, aside from
making a call to delight your IBM hardware rep, you might want to
consider/total the memory cost of hundreds of small 'warning flag' swap
disks vs cheap disk. After all, they those swap disks are not supposed to
be used regularly. They are a warning that something else needs to be
evaluated and improved - maybe even making that servers swap VDISK larger!
More likely: trimming the default virtual storage size of the Linux guest
to something a lot smaller than the x86 folks say it MUST BE to function
properly.
But if you let a lot of servers fail over onto the warning flag swap disks
without doing anything about it, then someone could complain that all your
Linux guest performance is bad and abandon the migration to the mainframe.
Of course, a good performance monitor **and the skills to understand how
to use it and how to fix the problems** can alleviate that situation. That
assumes a lot, as many new z/VM customers (and many old ones, too!) don't
have those skills.
Summary: if you do use small failover swap space (a very good idea),
ensure that the warning is to pop up properly when they are used (as
Barton mentioned), test that warning by causing swap to the failover disk
(VDISK -or- real disk, regardless), and then react promptly when something
sets off the alert in production.
Mike Walter
Hewitt Associates
Disclaimer: What I -don't- know about performance fills lots of books...
literally!
"Scott Rohling" <[email protected]>
Sent by: "The IBM z/VM Operating System" <[email protected]>
04/29/2009 08:51 PM
Please respond to
"The IBM z/VM Operating System" <[email protected]>
To
[email protected]
cc
Subject
Re: SWAPGEN
Right - I got that from Barton's post.
Consider that there are times when using a real disk might make sense ..
I'm just not big on blanket statements... Dedicating a resource is not
always a big waste and does not always benefit one to the detriment of the
rest. It's a balancing act and it takes all the tricks in the bag to keep
it from falling over. A real disk might make real sense with an
ill-behaved, unpredictable guest that is causing you swapping headaches.
That's what I was trying to point out..
Maybe I'd stay quiet if I saw some qualifiers like 'in general' when
people talk about BIG/REAL WASTE. Guidelines are nice, but I object to
them being presented as 'rules'.
I'll shuddup now ;-)
Scott
On Wed, Apr 29, 2009 at 6:16 PM, Rich Smrcina <[email protected]> wrote:
The biggest benefit of z/VM is to virtualize resources as much as
possible. Dedicating a resource to a specific guest is a big waste and
benefits one to the detriment of all the rest. It would be much better to
allocate all virtual disk swap for the Linux guests, then allocate the
disk that would have been dedicated to swap as page space. That way
everyone wins.
Scott Rohling wrote:
Maybe -- but having a real disk (and setting an alert for that) helps
isolate the issue to a single guest rather than affecting critical shared
resources (in this case memory/paging) when a guest starts swapping more
than normal or than it 'should'.
I like the idea of having a 'failover' swap area on real disk -- whether
you have a single VDISK or 2 prioritized ahead of it. I guess whether
it's waste of resource depends on your POV.. But your point is taken..
Scott
On Wed, Apr 29, 2009 at 4:42 PM, Barton Robinson <
[email protected]> wrote:
giving real disks to swap is a real waste of resource. It is much better
to take the "extra disk resource" that you allocate but never want to use,
and assign it to z/VM paging to enhance your paging subsystem. Then
define two vdisks for swap, prioritize them, and set an alert when the 2nd
disk is being used.
Martin, Terry R. (LOCKHEED MARTIN Performance Engineering/CTR) (CTR)
wrote:
Hi
I am using SWAPGEN to define by z/Linux VDISKS I also want to define a
real disk for swap. My question is can I use SWAPGEN to define a swap on
real DASD? If you have an example of the control card syntax to accomplish
this that would be great?
//Thank You,//
//Terry Martin//
//Lockheed Martin - Information Technology//
//z/OS & z/VM Systems - Performance and Tuning//
//Cell - 443 632-4191//
//Work - 410 786-0386//
//[email protected] <mailto:[email protected]>//
--
Rich Smrcina
The information contained in this e-mail and any accompanying documents may
contain information that is confidential or otherwise protected from
disclosure. If you are not the intended recipient of this message, or if this
message has been addressed to you in error, please immediately alert the sender
by reply e-mail and then delete this message, including any attachments. Any
dissemination, distribution or other use of the contents of this message by
anyone other than the intended recipient is strictly prohibited. All messages
sent to and from this e-mail address may be monitored as permitted by
applicable law and regulations to ensure compliance with our internal policies
and to protect our business. E-mails are not secure and cannot be guaranteed to
be error free as they can be intercepted, amended, lost or destroyed, or
contain viruses. You are deemed to have accepted these risks if you communicate
with us by e-mail.
