daveh@marob.MASA.COM (Dave Hammond) (05/17/89)
We have a few 286 machines running SCO 2.2.1, all of which exhibit amazingly high TIME values for 'swapper' on a ps table. On average, the time seems to increment by 5,000 minutes per day, or so (ie the TIME value, for a machine running 1 week without reboot, is ~35000:00). These machines have 4+ MB of RAM, run from 1 to 4 users, and run average load programs (vi, cc, make, cu, compress ...). Are these machines actually "swapping", or does the swapper just use a hunk of cpu time? If they are swapping, I can't believe its from lack of core -- so what [if any] tuning can be done to improve performance? -- Dave Hammond daveh@marob.masa.com
timsb@adspp.UUCP (Tim S. Boshart) (05/18/89)
In article <662@marob.MASA.COM> daveh@marob.masa.com (Dave Hammond) writes: }We have a few 286 machines running SCO 2.2.1, all of which exhibit }amazingly high TIME values for 'swapper' on a ps table. On average, }the time seems to increment by 5,000 minutes per day, or so (ie the }TIME value, for a machine running 1 week without reboot, is ~35000:00). }These machines have 4+ MB of RAM, run from 1 to 4 users, and run }average load programs (vi, cc, make, cu, compress ...). } }Are these machines actually "swapping", or does the swapper just use }a hunk of cpu time? If they are swapping, I can't believe its from }lack of core -- so what [if any] tuning can be done to improve performance? } }-- }Dave Hammond }daveh@marob.masa.com If I remember right, the swapper gets all of the idle time. So whenever nothing else is running, the swapper process gets the time added to it. -- Tim S. Boshart |uunet!adspp!timsb |"When it is possible for Abacus Data Systems |timsb@adspp.UUCP | programmers to program in English, 2707 Middlebury St. |1-219-295-4290 (voice)| we will find the programmers Elkhart, IN 46516 |1-219-522-2964 (fax) | cannot write in English"
trebor@biar.UUCP (Robert J Woodhead) (05/19/89)
In article <228@adspp.UUCP> timsb@adspp.UUCP (Tim S. Boshart) writes: >In article <662@marob.MASA.COM> daveh@marob.masa.com (Dave Hammond) writes: >}... On average, the time seems to increment by 5,000 minutes per day... >If I remember right, the swapper gets all of the idle time. So whenever >nothing else is running, the swapper process gets the time added to it. Even so, you have to admit it's impressive, especially when there are only 24 * 60 or 1,440 minutes in a day! I for one, when on deadline, would really appreciate one of these 5,000 minute days! Where can I get one? -- Robert J Woodhead, Biar Games, Inc. !uunet!biar!trebor | trebor@biar.UUCP "The lamb will lie down with the lion, but the lamb won't get much sleep." -- Woody Allen.
jbayer@ispi.UUCP (Jonathan Bayer) (05/20/89)
In article <662@marob.MASA.COM> daveh@marob.masa.com (Dave Hammond) writes: >We have a few 286 machines running SCO 2.2.1, all of which exhibit >amazingly high TIME values for 'swapper' on a ps table. On average, >the time seems to increment by 5,000 minutes per day, or so (ie the [ deleted ] >Are these machines actually "swapping", or does the swapper just use >a hunk of cpu time? If they are swapping, I can't believe its from >lack of core -- so what [if any] tuning can be done to improve performance? As it was explained to me by a rep at SCO, the swapper will eat up any excess time that is available. I don't understand the 5000 minutes a day, since there is only 1440 minutes in 24 hours. Possibly there is something wrong in the clock, but you don't have to worry about the time for the swapper. JB -- Jonathan Bayer Beware: The light at the end of the Intelligent Software Products, Inc. tunnel may be an oncoming dragon 500 Oakwood Ave. ...uunet!ispi!root Roselle Park, NJ 07204 (201) 245-5922 jbayer@ispi.UUCP