Sendmail Thrashing CPU Endlessly

Sendmail Thrashing CPU Endlessly

[Daniel W]

Extract from "top":

===============================
12:10pm  up 4 days, 27 min,  1 user,  load average: 1.24, 1.37, 1.23
82 processes: 76 sleeping, 5 running, 1 zombie, 0 stopped
CPU states:  4.4% user,  0.5% system,  0.0% nice,  8.3% idle
Mem:   262144K av,  258592K used,    3552K free,   17440K shrd,       0K
buff
Swap:  262144K av,    3704K used,  258440K free                  146500K
cache

PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME COMMAND
20144 root      13   0  2656    2656  1904      R         94.5       1.0
28:34     sendmail
23448 root       3   0   1096    1096   832       R         5.6         0.4
0:00       top
===============================

You can see from the above that the sendmail process is hogging virtually
the entire CPU. This goes on permanently (this extract is actually soon
after i killed the sendmail process - if I didn't it would go on forever).

I am running a mail server with perhaps 1000 users but I still don't see why
this would cause sendmail to have to operate THIS hard.

The problem is that whilst sendmail is doing this the load average slowely
climbs. I've had load averages of 40+ and long before that services start to
die. All horribly messy.

If anyone could give me a few clues as to what could be causing this I'd be
most grateful!

Sendmail version is:  Sendmail 8.11.6/8.11.6
Red Hat Linux release 7.1 (Seawolf)

Thanks,
Daniel

Re: Sendmail Thrashing CPU Endlessly

[Richard Nairn]

What does your mailq and maillog report?  Is this all the time? Try and 
see if anything triggers it...



On Fri, 3 Oct 2003 12:32:35 +0100, Daniel W 
<daniel.watts@lincoln-oxford.com> wrote:

>
>
> Extract from "top":
>
> ===============================
> 12:10pm  up 4 days, 27 min,  1 user,  load average: 1.24, 1.37, 1.23
> 82 processes: 76 sleeping, 5 running, 1 zombie, 0 stopped
> CPU states:  4.4% user,  0.5% system,  0.0% nice,  8.3% idle
> Mem:   262144K av,  258592K used,    3552K free,   17440K shrd,       0K
> buff
> Swap:  262144K av,    3704K used,  258440K free                  146500K
> cache
>
> PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME COMMAND
> 20144 root      13   0  2656    2656  1904      R         94.5       1.0
> 28:34     sendmail
> 23448 root       3   0   1096    1096   832       R         5.6         
> 0.4
> 0:00       top
> ===============================
>
> You can see from the above that the sendmail process is hogging virtually
> the entire CPU. This goes on permanently (this extract is actually soon
> after i killed the sendmail process - if I didn't it would go on 
> forever).
>
> I am running a mail server with perhaps 1000 users but I still don't see 
> why
> this would cause sendmail to have to operate THIS hard.
>
> The problem is that whilst sendmail is doing this the load average 
> slowely
> climbs. I've had load averages of 40+ and long before that services 
> start to
> die. All horribly messy.
>
> If anyone could give me a few clues as to what could be causing this I'd 
> be
> most grateful!
>
> Sendmail version is:  Sendmail 8.11.6/8.11.6
> Red Hat Linux release 7.1 (Seawolf)
>
> Thanks,
> Daniel
>
>
>
>
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-admin" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html



-- 
|       Richard Nairn          Specializing in Linux
|     Nairn Consulting         Web / Database Solutions
|        Calgary, AB
| Richard@NairnConsulting.ca

Re: Sendmail Thrashing CPU Endlessly

[Daniel Watts]

mailq has only this:

----------------------------------------
                /var/spool/mqueue (1 request)
----Q-ID---- --Size-- -----Q-Time----- ------------Sender/Recipient---------
---
h92NEdi01184      263 Fri Oct  3 00:14 <MAILER-DAEMON@magd-oxford.com>
                 (Deferred: Connection timed out with
microdata.mdnet.com.br.)
                                       <jillballsn@mdnet.com.br>
----------------------------------------

Although I am running a virtual diectory system with over 50 domains. There
is no mailq program in /home/virtual/DOMAIN/usr/bin .


From the maillog I have these interesting lines:

--------------------
named[24515]: client 80.67.173.196#2498: zone transfer
'trinity-oxford.com/IN' denied

named[24515]: lame server resolving 'pidgeyou.com' (in 'pidgeyou.COM'?):
62.204.32.5#53
named[24515]: lame server resolving 'pidgeyou.com' (in 'pidgeyou.COM'?):
62.204.34.85#53

imapd[27258]: Command stream end of file, while reading line
user=shanthi.vignasen@oriel-oxford.com host=localhost.localdomain
[127.0.0.1]

imapd[5191]: Login failed user= auth= host=localhost.localdomain [127.0.0.1]
---------------------

I have an inkling that the last line might be the problem since I'm getting
quite a lot of these (about 5-8 times as many as everything else).

sendmail starts using as much of the CPU as it can as soon as it is started
up. I kill it and a few minutes later it is back up and trashing away.

Thanks for replying - your help is greatly appreciated!
Daniel