how to properly set #of logical processors per LPAR?

2008-06-16 Thread Pawel Leszczynski
Hello everybody,

recently, as our production LPARs apetite for CPU has grown meaningful 
(we are in process connecting succesive branches of bank to new application),
our sysprogs decided to simply put online subsequent processors to this LPARs.
We ended in 12 processors assigned to each of them.
I would like to know if 12 processors/LPAR is not too many, i.e.
measure how much capacity it 'wasted' that way.
I realise that RMF monitor III shows overhead on let's say LPAR dispatching
but I guess that's a different story.

Where is this CPU 'wasted'?

That's my naive thinking (please correct me):

Serialization beetween processors within LPAR is done via locks.
Requests for spin lock (from this point of view) are 'CPU-wasting' - processor 
have to wait actively (if wait at all) for obtaining such a lock.
Where are other places???
Is there any tool which sums such 'CPU-wasted' cycles and shows how many 
percent of CPU is 'wasted' in such way?

Regards,
PAwel Leszczynski
PKO BP SA

--
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



Re: how to properly set #of logical processors per LPAR?

2008-06-16 Thread Martin Packer
Pawel, it's hard to advise on that unless you give us some more hints 
about your configuration...

e.g LPAR setup, software and hardware models etc.

Just for starters it's entirely reasonable to consider that a logical 
12-way might find inside one processor book of a z9 S54.

Martin

Martin Packer
Performance Consultant
IBM United Kingdom Ltd
+44-20-8832-5167
+44-7802-245-584
[EMAIL PROTECTED]








Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 
741598. 
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU






--
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



Re: how to properly set #of logical processors per LPAR?

2008-06-16 Thread Kelman, Tom
 From: Pawel Leszczynski
 Sent: Monday, June 16, 2008 3:18 PM
 
 Hello everybody,
 
 recently, as our production LPARs apetite for CPU has grown meaningful
 (we are in process connecting succesive branches of bank to new
 application),
 our sysprogs decided to simply put online subsequent processors to
this
 LPARs.
 We ended in 12 processors assigned to each of them.
 I would like to know if 12 processors/LPAR is not too many, i.e.
 measure how much capacity it 'wasted' that way.
 I realise that RMF monitor III shows overhead on let's say LPAR
 dispatching
 but I guess that's a different story.
 
 Where is this CPU 'wasted'?
 
 That's my naive thinking (please correct me):
 
 Serialization beetween processors within LPAR is done via locks.
 Requests for spin lock (from this point of view) are 'CPU-wasting' -
 processor
 have to wait actively (if wait at all) for obtaining such a lock.
 Where are other places???
 Is there any tool which sums such 'CPU-wasted' cycles and shows how
many
 percent of CPU is 'wasted' in such way?
 
 Regards,
 PAwel Leszczynski
 PKO BP SA
 
We need a little more information here.  For example, what are the total
number of physical processors?  One rule I've followed is that you don't
want to have the number of LPs to be more than 2.5 times the number of
PPs.  That might be a little outdated, I don't know, but it's what I
use.



*
If you wish to communicate securely with Commerce Bank and its
affiliates, you must log into your account under Online Services at 
http://www.commercebank.com or use the Commerce Bank Secure
Email Message Center at https://securemail.commercebank.com

NOTICE: This electronic mail message and any attached files are
confidential. The information is exclusively for the use of the
individual or entity intended as the recipient. If you are not
the intended recipient, any use, copying, printing, reviewing,
retention, disclosure, distribution or forwarding of the message
or any attached file is not authorized and is strictly prohibited.
If you have received this electronic mail message in error, please
advise the sender by reply electronic mail immediately and
permanently delete the original transmission, any attachments
and any copies of this message from your computer system.
*

--
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