High load average mail server 5.3-RELEASE

From: mariano benedettini (marianobe_at_gmx.net)
Date: 09/13/05

  • Next message: Xin LI: "Combine more operation within one system call: to do it, or not to do it?" 
    Date: Tue, 13 Sep 2005 20:06:58 +0200 (MEST)
To: freebsd-performance@freebsd.org

    I have a mail server running FreeBSD 5.3-RELEASE. It's a dual Xeon with 4gb
    of ram.
    The server is running Apache (serving Horde) , Postfix, Courier imapd w/SSL,
    Amavisd. Also a Postgresql as Horde's storage.
    The mail storage is accessed via NFS.
    The problem is that it's experiencing a really high load average (see the
    top -S ) while the CPU's are relatively idle. I think it's a IO problem
    (nfs?), but I cant figure out what exactly is causing this.

    top -S:

    last pid: 39883; load averages: 32.40, 15.81, 13.515 up
    20+22:51:19 14:13:43
    1084 processes:5 running, 1006 sleeping, 73 waiting
    CPU states: 6.0% user, 0.0% nice, 2.6% system, 0.1% interrupt, 91.3%
    idle
    Mem: 1599M Active, 1704M Inact, 311M Wired, 189M Cache, 112M Buf, 14M Free
    Swap: 2023M Total, 184K Used, 2023M Free

      PID USERNAME PRI NICE SIZE RES STATE C TIME WCPU CPU
    COMMAND
       11 root 132 0 0K 12K CPU3 3 468.7H 81.30% 81.30%
    idle: cpu3
       12 root 132 0 0K 12K CPU2 2 464.0H 80.42% 80.42%
    idle: cpu2
       14 root 132 0 0K 12K RUN 0 464.0H 80.27% 80.27%
    idle: cpu0
       13 root 132 0 0K 12K CPU1 1 442.6H 70.85% 70.85%
    idle: cpu1
    32019 www 4 0 33460K 22228K sbwait 1 0:04 1.61% 1.61%
    httpd
    29620 www 4 0 33900K 22648K kqread 2 0:06 1.56% 1.56%
    httpd
    39219 root 5 0 5288K 4064K ttyin 2 0:05 1.57% 1.56% top
       78 root -44 -163 0K 12K WAIT 0 337:32 1.46% 1.46%
    swi1: net
    17746 icn 4 0 3128K 2032K select 2 0:25 1.32% 1.32%
    imapd
       24 root -64 -183 0K 12K WAIT 0 136:15 0.83% 0.83%
    irq11: ohci0
    29170 www 20 0 31384K 20144K lockf 3 0:04 0.68% 0.68%
    httpd
    30476 www 20 0 31392K 20148K lockf 3 0:05 0.63% 0.63%
    httpd
    31091 www 4 0 32868K 21624K sbwait 3 0:04 0.59% 0.59%
    httpd
    39712 bnz 4 0 2628K 1668K select 1 0:01 0.69% 0.59%
    imapd
    32951 www 20 0 27996K 16716K lockf 2 0:03 0.54% 0.54%
    httpd
    39434 vscan 20 0 19820K 17896K lockf 2 0:00 0.55% 0.54%
    perl
       39 root -68 -187 0K 12K WAIT 2 140:23 0.34% 0.34%
    irq26: bge2
    36443 vscan 20 0 20492K 18540K lockf 2 0:02 0.34% 0.34%
    perl
       79 root -28 -147 0K 12K WAIT 2 157:14 0.29% 0.29%
    swi5: clock sio
      103 root 20 0 0K 12K syncer 3 90:12 0.29% 0.29%
    syncer
        4 root -8 0 0K 12K - 2 64:52 0.29% 0.29%
    g_down
    37590 vscan 20 0 20620K 18664K lockf 0 0:01 0.29% 0.29%
    perl
    32942 www 20 0 31612K 20360K lockf 3 0:04 0.24% 0.24%
    httpd
    33769 pgsql 4 0 20136K 13356K sbwait 2 0:00 0.24% 0.24%
    postgres

    -- Trafshow bge2:

    >From Address To Address Prot Bytes
    CPS
    =====================================================================================
    192.168.1.6..978 192.168.1.5..nfsd udp 5956916
    91306
    192.168.1.5..nfsd 192.168.1.6..978 udp 1511540
    118858
    192.168.1.5 192.168.1.6 udp 2272172
    6259
    192.168.1.6 192.168.1.5 udp 190084
    16715

    After searching the lists and documentation, I'm not sure what's
    causing the problem.
    Suggestions ?

    Mariano. 

    -- 
GMX DSL = Maximale Leistung zum minimalen Preis!
2000 MB nur 2,99, Flatrate ab 4,99 Euro/Monat: http://www.gmx.net/de/go/dsl
_______________________________________________
freebsd-performance@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "freebsd-performance-unsubscribe@freebsd.org"
  • Next message: Xin LI: "Combine more operation within one system call: to do it, or not to do it?"