Find the answer to your Linux question:
Results 1 to 3 of 3
Hi, I'm seeing consistently high load on a dual core P4 (3GHz) with 2GB RAM and a single SCSI disk. The machine acts as a mail server (smtp, pop3, imap) ...
Enjoy an ad free experience by logging in. Not a member yet? Register.
  1. #1
    Just Joined!
    Join Date
    Dec 2006
    Posts
    24

    interpreting vmstat output/tracking down high load


    Hi,

    I'm seeing consistently high load on a dual core P4 (3GHz) with 2GB RAM and a single SCSI disk. The machine acts as a mail server (smtp, pop3, imap) and runs qmail + spamassassin.

    Uptime is showing the load average as around 15 to 20, and a typical snippet from vmstat is show below:

    Code:
    # vmstat 10
    procs -----------memory---------- ---swap-- -----io---- --system-- ----cpu----
     r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy id wa
    20  2     16 664332   3928 499732    0    0   301    64  227   226  5 20 69  6
     4  3     16 669348   3396 492648    0    0   461   486 1431  1175 14 29 13 43
     3  1     16 669916   2808 491740    0    0   454   606 1447  1270 16 47 11 26
     2  2     16 671316   2824 488800    0    0   553   475 1325   908 15 42 15 28
     0  5     16 672980   2840 488036    0    0   381   544 1268   792 26 51  7 16
     4  5     16 671988   2808 489292    0    0   604   321 1190   917  5 29 16 49
     1  3     16 670476   2784 491152    0    0   617   298 1245   876  7 20 18 56
     1  2     16 669696   2732 492292    0    0   556   266 1223   987 19 42  7 32
     0  4     16 669884   2724 491076    0    0   441   570 1294  1075 18 49  3 29
    Context switches seem rather high, but whether they are excessively high, I don't know.

    From what I know, blocked processes and high waits (but are those waits worryingly high, or just healthy?) usually indicates a disk bottleneck, but disk i/o doesn't seem particularly intensive. I did wonder if the disk was starting to die, but I'm not seeing any errors in the logs.

    As you might expect for a mail server, it gets bombarded with a lot of spam, and one of the first things I did was to implement a couple of DNSBLs at the SMTP-level, figuring that this would dramatically cut the amount of spam being processesed by qmail and spamassassin. The volume of spam certainly seems to have gone down, but the load hasn't.

    Finally, it seems that qmail's outbound queue is fine - it's the inbound which is being affected.

    I did wonder about mounting /var/qmail/ as a ram disk, but I'm not sure it'd make a noticeable difference. I'm sure there must be something wrong somewhere else

  2. #2
    Just Joined!
    Join Date
    Apr 2009
    Posts
    1
    Hi,

    If you need to draw charts from your vmstat log, this
    may help you :
    Vmstax | Michenux


  3. #3
    Linux Newbie
    Join Date
    Apr 2008
    Location
    India
    Posts
    170
    did you check the dmesg ?
    Regards
    David Anand
    -->Success is the list of failures ...!!!

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •