Kees

This is a simple application of Erlang M/M/C model. With 4 servers at 50% busy, the probability is 0.1739130435 that all servers will be busy.

Regards,

Don
******
Don Deese, Computer Management Sciences, Inc.
Voice: (703) 922-7027  Fax: (703) 922-7305
http://www.cpexpert.org
******



At 03:22 AM 6/14/2005, you wrote:
As far as I remember, you have to start digging into 'Eigen numbers' or
something like that

-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf
Of Vernooy, C.P. - SPLXM
Sent: 14 June 2005 08:05
To: [email protected]
Subject: Re: Disk I/O tuning still possible?

"Robert Hamilton" <[EMAIL PROTECTED]> wrote in message
news:<[EMAIL PROTECTED]>...
> >> So, I ask,  is I/O tuning still possible, or even necessary?
> >> ..
> >> I believe you can run them quite high, well over 50%.
> >> Since (the simple equation), the odds of all four being busy is:
> >>
> >> 5 ** 4 = .125
> >>
> >> Or 12.5% of the time there would be no channel busy.
> >> Of course, you have to calculate the 'Relative Path Busy', since your
own
> >I/O doesn't count.
> >> But, I did say 'simple'.
> >>
> >
> >Beware: .5 **4 = 0.0625, 6% of the time no path is busy.
> >It also means that (1-.5) ** 4 = 0.0625 6% if the time all paths are busy
> >and you will encounter delays and this sounds a little high to me.
> >
> >.5 ** 6 = 1.5% all paths busy
> >.6 ** 8 = 1.6% all paths busy
> >
>
> I/O tuning is still alive, especially in DASD replication projects.
> However, it takes time and practice to build I/O tuning skills.
>
> The formula (Utilization ** 4) just doesn't work. It's not even close. The
> math is much more complex, and we can't simplify some things. It could
make
> a huge difference. To make the story short, for 4 paths at 50%, "the
> probability of all busy" is approx 17%.
>

Could you elaborate this? I am not graduated in statistics, but I thought I
knew something about it.

Kees.







--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.323 / Virus Database: 267.7.1 - Release Date: 6/13/2005

----------------------------------------------------------------------
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