[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: swappiness of 2.6 kernel



On Tue, 5 Aug 2008, thveillon.debian wrote:

On a Desktop workstation (multimedia edit, 4Gb ram, timer freq being set @ 1000HZ) I've been using 20 for a long time, no problem whatsoever and no swapping at all. From my understanding for a server you need something closer or equal to 100, for a workstation anything between 0 and 20 will do, anyway if swapping must occur it will and the settings will be overridden, but no "provisional" swapping (that's my word, not sure it makes sens...) will occur, resulting in a more responsive experience especially when coming back to an application left idle for a while.

That's my experience, and therefore my 2 cents...

Tom

Tom,

That is about what I was thinking. Since this is a mail server it would seem like a disk cache is not a big deal. This being all the disk access is mostly mail files coming in. But you do need a bunch free memory for programs, very memory hungry program, involved in processing incoming mail.

So far the 80 setting has freed up some free memory for use. What I can't understand is why this 76k of swap memory keeps getting used. I'll clear it out and then couple of hours later there it is again. Anyway here is a snap shot of the current top:

Ken

top - 06:27:16 up 1 day, 20:15,  4 users,  load average: 1.22, 1.03, 0.90
Tasks: 118 total,   1 running, 117 sleeping,   0 stopped,   0 zombie
Cpu0  :  0.7%us,  1.0%sy,  6.3%ni, 92.1%id,  0.0%wa,  0.0%hi,  0.0%si,  0.0%st
Cpu1  :  1.6%us,  3.3%sy, 24.3%ni, 57.2%id, 12.8%wa,  0.3%hi,  0.3%si,  0.0%st
Mem:   3927348k total,  3520456k used,   406892k free,   277516k buffers
Swap:  5076532k total,       76k used,  5076456k free,  2126116k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
28287 clamav    20   0  437m 260m 1232 S    0  6.8  14:15.51 clamav-milter
28066 clamav    20   0  102m  86m  720 S    0  2.3   0:31.62 clamd
20863 root      37  17  106m  74m 2432 S   13  1.9   0:04.76 spamd
21252 smithj    37  17  103m  71m 2408 S   17  1.9   0:01.86 spamd
21395 root      37  17  102m  70m 2320 S    0  1.8   0:00.32 spamd
 4970 root      37  17  100m  68m 2180 S    0  1.8   1:28.18 spamd
 2505 root      20   0  246m  68m  804 S    1  1.8  14:30.21 milter-greylist
 2500 dcc       20   0 67132 9280 1288 S    0  0.2   1:08.84 dccifd
  442 root      20   0 39296 4180 1984 S    0  0.1   0:00.00 sendmail-mta
28692 root      20   0 39272 4124 1928 S    0  0.1   0:00.00 sendmail-mta
  865 root      20   0 39272 4092 1928 S    0  0.1   0:00.00 sendmail-mta
11537 root      20   0 39272 4092 1928 S    0  0.1   0:00.00 sendmail-mta
25395 root      20   0 39272 4092 1928 S    0  0.1   0:00.00 sendmail-mta
19006 root      20   0 39272 4080 1940 S    0  0.1   0:00.00 sendmail-mta
20825 root      20   0 39272 4080 1928 S    0  0.1   0:00.00 sendmail-mta
21420 root      20   0 39272 4072 1944 S    0  0.1   0:00.00 sendmail-mta
19220 root      20   0 39004 4008 1876 S    0  0.1   0:00.00 sendmail-mta

etc.....


Reply to: