One if my two Solaris servers has very high CPU load Average and the Milter-graylist process takes high CPU usage. We use BigBrother to monitor and report CPU Avage Load. (Each server takes 50% work load from the DNS distribution balancer) It shows LOAD AVG on hestia is above 250 a couple of times a day (high for 5 minutes then back to around 60 to 90) while LOAD AVG on demeter is 0.66 in the last 4 days Each server proccesses 30 thousand emails daily and it runs spamassassin, mine-defang software. Both servers has two 900MHZ CPUs with 4GB of RAM. We run mimedefang version 2.51 and SpamAssassin version 3.0.3 running on Perl version 5.8.0

I think that there is an application problem on Hestia but do not know what's wrong. It looks like milter-greylist proccess put the a high load on the cpu. Is there a problem here?
Any help will be very appreciated.
Thanks,
An H. Nguyen

Instant snapshots from 'top' command:

Problem host: 'hestia'

load averages: 1.20, 1.33, 1.42 hestia 15:19:09
69 processes:  66 sleeping, 1 stopped, 2 on cpu
CPU states: 41.2% idle, 55.9% user,  2.9% kernel,  0.0% iowait,  0.0% swap
Memory: 4.0G real, 3.3G free, 214M swap in use, 7.9G swap free

  PID USERNAME THR PR NCE  SIZE   RES STATE   TIME FLTS    CPU COMMAND
1958 smmsp 49 48 0 46.9M 45.8M cpu00 36:48 0 22.59% milter-greylist 776 defang 4 23 0 35.3M 34.5M sleep 0:09 0 5.99% mimedefang.pl 965 defang 4 20 0 32.1M 31.2M sleep 0:02 0 4.09% mimedefang.pl
  701 root       1 58   0 2560K 1536K cpu01   0:01    0  0.63% top
 1912 root      58 58   0 39.2M 37.6M sleep   2:10    0  0.62% ismilter
 1956 defang    74 58   0 4128K 3152K sleep   1:07    0  0.49% mimedefang
22038 root      14 58   0 3128K 1832K sleep   1:43    0  0.23% syslogd
 1961 root       1 43   0 2888K 1680K sleep   0:19    0  0.11% sendmail
  978 root       1 48   0 3312K 2648K sleep   0:00    0  0.11% sendmail
1942 defang 3 58 0 3576K 1592K sleep 0:13 0 0.10% mimedefang-mult
  987 root       1 58   0 3328K 2648K sleep   0:00    0  0.09% sendmail
  986 root       1 58   0 3336K 2656K sleep   0:00    0  0.09% sendmail
  988 root       1 58   0 3336K 2664K sleep   0:00    0  0.07% sendmail
  993 root       1 53   0 3304K 2512K sleep   0:00    0  0.02% sendmail
  979 root       1 55   0 3312K 2528K sleep   0:00    0  0.02% sendmail




Normal  host 'demeter'
load averages: 0.86, 1.18, 1.34 demeter 15:18:13
55 processes:  52 sleeping, 1 running, 1 stopped, 1 on cpu
CPU states: 27.8% idle, 69.4% user,  2.8% kernel,  0.0% iowait,  0.0% swap
Memory: 5.0G real, 4.3G free, 190M swap in use, 4.3G swap free

  PID USERNAME THR PR NCE  SIZE   RES STATE   TIME FLTS    CPU COMMAND
17104 smmsp 36 58 0 44.8M 43.7M sleep 36:01 0 12.21% milter-greylist 16025 defang 1 50 0 31.7M 31.0M run 0:10 0 6.91% mimedefang.pl
16210 root       1 58   0 2560K 1544K cpu00   0:00    0  1.31% top
17058 root      40 59   0 37.0M 35.3M sleep   2:12    0  0.39% ismilter
17101 defang    55 58   0 3648K 2680K sleep   1:05    0  0.34% mimedefang
17107 root       1 38   0 2888K 1672K sleep   0:20    0  0.11% sendmail
16260 root       1 58   0 3336K 2656K sleep   0:00    0  0.11% sendmail
16239 root       1 58   0 3328K 2656K sleep   0:00    0  0.10% sendmail
 4456 root       1 58   0 6568K 2720K sleep   0:03    0  0.10% sshd
10562 root      14 58   0 3144K 1856K sleep   1:40    0  0.09% syslogd
16052 defang 1 40 0 28.8M 28.1M sleep 0:03 0 0.07% mimedefang.pl
16245 root       1 48   0 3336K 2664K sleep   0:00    0  0.06% sendmail
16258 root       1 48   0 3320K 2648K sleep   0:00    0  0.06% sendmail
16251 root       1 56   0 3336K 2640K sleep   0:00    0  0.05% sendmail
16265 root       1 58   0 3320K 2528K sleep   0:00    0  0.05% sendmail

_______________________________________________
NOTE: If there is a disclaimer or other legal boilerplate in the above
message, it is NULL AND VOID.  You may ignore it.

Visit http://www.mimedefang.org and http://www.roaringpenguin.com
MIMEDefang mailing list [email protected]
http://lists.roaringpenguin.com/mailman/listinfo/mimedefang

Reply via email to