Mailscanner and server load

kfliong kfliong at WOFS.COM
Fri Aug 27 07:42:44 IST 2004


<x-flowed>
Thanks for the replies guys. I can learn a lot from all your comments.

Anyway, here are more info on my system.

Around 50,000 mails per day of which 95% are SPAMS. I am currently using
list.dsbl.org on my sendmail.cf to which is helping a little to kill off
mails before they can come into my server. But it have false identification
which is causing some users unable to send mail using SMTP, that's why I
need to rely on SURBL and remove dsbl on MTA.

I was using only bigevil previously and since SURBL will replace bigevil, I
have removed bigevil.

A recap of my system specs :

Celeron 1.3GHz, 512mb RAM, 60gb hdd, redhat 7.3, mailscanner+SA+clamav (all
latest stable version)

And I do not run any DNS server. But will do that once I upgraded my box to
fedora core 1 and ensim pro 4.0.1.

For now, I want to know if I should upgrade to 1GB RAM.

Here are my loads :

    procs                      memory    swap          io     system
  cpu
  r  b  w   swpd   free   buff  cache  si  so    bi    bo   in    cs  us
sy  id
  4  0  1 318036  29432  22700
116600  48  86     6    85   18    98  31  15  54
  2  0  0 317564  37516  22784
116796   0  48    84   664  395   309  54  46   0
  8  0  0 317560  41960  22964
117188  16   0   256  1066  494   539  60  37   2
  8  0  0 317560  10104  22996
117324  20  40    84   296  342   283  72  28   0
  1 10  1 317632   5400  23032 115168  22
1180    72  1520  514   219  43  15  42
  0 24  2 318244   5304  22768 113568  88
1558   144  1766  531   215   6   5  89
14  8  2 319088   5304  22824 111548  38
1668   136  1878  619   261   4   5  91
  0 17  2 319364   5304  22824 108176 178
1226   208  1368  462   242  10  10  80
21  5  2 319808   5308  22820 106608 176
1114   232  1270  481   335  13   8  79
  0 13  1 320048  23256  22816 104052 446
864   508  1064  450   356  12  11  77
10  6  2 320220   5304  22984 103500 550 624   622  1456  556   563  58  15  27
  9  5  2 320792   5400  22996 101436 166
1000   186  1160  363   239  36  11  53
  4 12  2 321136   5296  22972  98536 348
1276   358  1466  428   311  49  13  38
  6 10  1 320968  13900  22988  97416 288
766   304  1112  353   280  62   9  29
  3 13  1 321464   7692  23072  97532 292
310   382   722  282   339  38  12  50
  2 11  1 320416  25700  23132  97708 344
162   440   442  363   415  30  14  56


And here is the top command :

   2:16pm  up 3 days, 22:53,  1 user,  load average: 6.34, 6.06, 4.96
181 processes: 178 sleeping, 3 running, 0 zombie, 0 stopped
CPU states:  6.9% user,  6.9% system, 60.8% nice, 25.2% idle
Mem:   506048K av,  458944K used,   47104K free,       0K shrd,   24508K buff
Swap: 1020116K av,  315784K used,  704332K free                  124892K cached

<SORT BY MEMORY>
   PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME COMMAND
  1140 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:29 mysqld
  1160 mysql     11  10 19040  13M  1524 S N   0.4  2.7   0:36 mysqld
  1161 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
  1162 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
  2487 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
  3120 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
  6240 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
  6460 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
  6465 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
  6468 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
  6480 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
  6539 mysql     15  10 19040  13M  1524 S N   0.0  2.7   0:00 mysqld
31864 root       9   0 42128  10M  1692 S     0.0  2.1   0:31 MailScanner
32413 root       8   0 42184  10M  1656 S     0.0  2.1   0:30 MailScanner
31892 root       8   0 42312   9M  1648 S     0.0  2.0   0:32 MailScanner
31846 root       9   0 42196 9928  1684 S     0.0  1.9   0:29 MailScanner
31932 root       8   0 42172 9844  1656 S     0.0  1.9   0:32 MailScanner

<SORT BY CPU>
   PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME COMMAND
  6647 root      20   0 42332  19M  1816 R     5.9  3.9   0:00 MailScanner
  4653 apache     9   0 10924 5708  3216 R     4.4  1.1   0:01 httpd
  5264 apache     9   0 12080 7528  3836 S     4.4  1.4   0:00 httpd
  6433 apache    11   0 11840 7044  3756 S     2.4  1.3   0:00 httpd
31864 root      15   0 42132  10M  1692 S     1.9  2.1   0:32 MailScanner
  6046 apache     8   0 11128 6352  3632 S     1.4  1.2   0:00 httpd
    13 root       9   0     0    0     0 RW    0.4  0.0   6:53 kjournald
21042 root       9   0  2004 1656  1252 R     0.4  0.3   0:20 sshd
31932 root      12   0 42180 9104  1656 S     0.4  1.7   0:32 MailScanner
  6383 apache     9   0 11552 6748  3788 S     0.4  1.3   0:00 httpd
  6582 root      10   0  1156 1156   836 R     0.4  0.2   0:00 top
  6645 mysql     15  10 18876  13M  1524 S N   0.4  2.7   0:00 mysqld
  6649 root      15   0 42132  10M  1692 R     0.4  2.1   0:00 MailScanner
     1 root       8   0   456  412   392 S     0.0  0.0   0:08 init
     2 root       9   0     0    0     0 SW    0.0  0.0   0:00 keventd
     3 root       9   0     0    0     0 SW    0.0  0.0   0:00 kapmd
     4 root      19  19     0    0     0 SWN   0.0  0.0   0:00 ksoftirqd_CPU0

Please bear in mind that I have turned off 3 out of the 4 filtering in
spamcop_uri.cf.
Also, bayes is turned on. Max children is 5.

Thanks in advance.


At 04:41 PM 26/8/2004, you wrote: Hi!

 > I have just implemented SURBL. Server load shot up and stays around 30 >
making websites unaccessable. I have now edited SURBL to only check on >
SPAMCOP_URI_RBL. I have disabled the other 3. Now my load averages around >
12. This is still quite high. > > BTW, after some monitoring using "top", I
notice that my system is quite > RAM intensive. But "top" can't tell for
sure. What other tools can I use to > see whether the highload is due to
lots of disk accesses (due to not having > enough RAM).

This question would be more in its place on the SURBL mailinglists i think.
But, do you have local caching servers for the DNS zones ? And, whats the
mailvolume and machine specs you are using ?

Bye, raymond.

------------------------ MailScanner list ------------------------ To
unsubscribe, email jiscmail at jiscmail.ac.uk with the words: 'leave
mailscanner' in the body of the email. Before posting, read the MAQ
(http://www.mailscanner.biz/maq/) and the archives
(http://www.jiscmail.ac.uk/lists/mailscanner.html).

------------------------ MailScanner list ------------------------
To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
'leave mailscanner' in the body of the email.
Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).
</x-flowed>



More information about the MailScanner mailing list