mail routing performance issues, backing up in /var/spool/mqueue

[Ugo Bellavance]

lester lasad wrote:


>>This looks like a sendmail problem, since
>>MailScanner's job is finished
>>once messages are in /var/spool/mqueue.  How is your
>>outgoing sendmail
>>called?
>
>
> My MS conf settings have it set to queue.  If I'm
> answering this incorrectly please let me know.
>
So I guess you didn't change anything in the way sendmail is called right?

>>
>>Do you see anything else, swapping, high cpu usage?
>>
>>what is the output of free?
>>what is the output of
>>
>>vmstat 1
>
>
> This is output from top.  hardly any memory available.

That is not true, according to vmstat, you're not swapping, so you're
not lacking memory here.  The memory, btw, is partly used by the kernel
cache.

>
>  12:18:55  up 23:19,  2 users,  load average: 1.61,
> 2.03, 1.91
> 58 processes: 57 sleeping, 1 running, 0 zombie, 0
> stopped
> CPU states:  cpu    user    nice  system    irq
> softirq  iowait    idle
>            total   13.4%    0.0%    4.2%   0.0%
> 0.0%    0.0%  182.0%
>            cpu00    1.3%    0.0%    0.7%   0.0%
> 0.0%    0.0%   97.8%
>            cpu01   12.1%    0.0%    3.5%   0.0%
> 0.0%    0.0%   84.2%
> Mem:  1291356k av, 1236132k used,   55224k free,
> 0k shrd,   51216k buff
>
> sorry for all the output but this is the result from
> vmstat 1

You have to do ctrl-c after a few seconds to have only a few lines of
output.

>
> vmstat 1
> procs                      memory      swap
> io     system         cpu
>  r  b   swpd   free   buff  cache   si   so    bi
> bo   in    cs us sy wa id
>  0  0  29716  52092  51224  43008    0    0     2
> 211  109    84 18  5  0 77
>  0  0  29716  36940  51224  43012    0    0     0
> 0  171   104 18  8  0 73
>  0  0  29716  42432  51224  43016    0    0     0
> 0  170   123 11  4  0 85
>  1  0  29716  31508  51224  43032    0    0     0

>>>I am not finding anything in the maillog that
>>would
>>>lead me to believe that there is a problem.

Could you show us some of your logs?

And what do you have when you do

mailq

?

>> The
>>
>>only
>>
>>>steps I have taken so far was to try using
>>
>>different
>>
>>>dns servers and I also turned on skip_rbl_checks
>>
>>in
>>
>>>/etc/MailScanner/spam.spamassassin.prefs.conf
>>
>>do you have a local caching nameserver?
>
>
> No...

This is probably not related, but you should run a local caching
nameserver or nscd.

>
>>>If I restart MailScanner it seems to route quickly
>>
>>but
>>
>>>only for a couple of minutes then it starts
>>
>>backing up
>>
>>>again.  I usually have anywhere from 200 to 500
>>>messages in mqueue at one time since this problem
>>>started surfacing.
>>
>>And how many did you have before?
>
>
> Mail was routed as quickly as it came in, I would say
> no more that 10 - 20 in the queue at a time.
>
>
>>>Any other suggestions would be greatly
>>
>>appreciated.
>>
>>>My box routes roughly 40,000 - 50,000 messages per
>>
>>day.

-------------------------- MailScanner list ----------------------
To leave, send    leave mailscanner    to jiscmail at jiscmail.ac.uk
Before posting, please see the Most Asked Questions at
http://www.mailscanner.biz/maq/     and the archives at
http://www.jiscmail.ac.uk/lists/mailscanner.html