> - what do you mean by average. I disclose very big secret: CPU is always > 100% busy or 0% busy (*). Always - mean in every tact (tick?).
Depends on how many CPs you have. > Date: Fri, 15 Apr 2011 23:37:06 +0200 > From: [email protected] > Subject: Re: CPU utilization/forecasting > To: [email protected] > > W dniu 2011-04-15 22:57, Ward, Mike S pisze: > > Hello all, can someone tell me if it's better to use CPU peak or CPU > > average to project growth. My way of thinking is if you use peak then > > you're sure to show where you need extra horse power. In other words if > > you can't process during your peaks what good are the averages. > > The answer is quite obvious: IT DEPENDS. > It depends on the following: > - what the peaks are. If the peak is caused by night job, then you > probably accept longer job run. If the peak is casued by on-line > processing then your customers could complain about it. > - what is your business decision/model/philosophy. Maybe your business > have to be cheap and can accept some delays. > - what do you mean by average. I disclose very big secret: CPU is always > 100% busy or 0% busy (*). Always - mean in every tact (tick?). > So the shorter period you take for CPU measurement the higher peaks you > will get. If your periods are as short as CPU cycle then your CPU usage > is binary: 100% or 0%. > So, what average would you like to take? 4H, 24H or 15 min.? > > > (*) Of course the above is some simplification. It does not consider > multiprocessing and ability to execute >1 instructions in one cycle. > It also does not consider the fact by increasing CPU power you can > shorten CPU time but not I/O time. > > My €0.02 > -- > Radoslaw Skorupka > Lodz, Poland ---------------------------------------------------------------------- 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
