What I should have indicated was there no NO plans to exceed the number
of physical IFLs installed.

The number specified on the MACHINE statement would always be EQ or LT
the number of physical IFLs.

My question was more on the benefits of running with MORE available to
zVM with FEWER specified in the quest directory entry.

TIA 


Gerard 


-----Original Message-----
From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of
Mark Post
Sent: Thursday, February 14, 2008 2:17 PM
To: [email protected]
Subject: Re: Running Oracle on GT 2 IFLs

>>> On Thu, Feb 14, 2008 at  1:10 PM, in message
<[EMAIL PROTECTED]>, "Shockley,
Gerard C" <[EMAIL PROTECTED]> wrote: 
> Excuse the generally vague nature of this question.  Here goes ...
> 
> Anyone have any experience pro/con running Oracle under zVM/Linux with

> GT N+1 engines.
> 
> Are there benchmark numbers something like LSPR? Here is why.
> 
> We are discussing sizing's and I've heard that it may be advisable to 
> remove MACHINE statements in the zVM guest directory for Oracle.

The general rule is to not have more virtual CPUs defined to guest than
that guest can saturate at peak loads.  So, if you see your CPU pegging
at 100% or close to it, add another CPU.  Oracle can and will make use
of them.  Just don't define extra CPUs "just in case" *unless* you use
the maxcpus kernel parameter to only activate (at boot time) the ones
you know will get used.  Then, if needed, you can activate more than
that without taking the system down.


Mark Post

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions, send
email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or
visit http://www.marist.edu/htbin/wlvindex?LINUX-390

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Reply via email to