Server dies with 100% RAM and swap file use.

Miguel Koren O'Brien de Lacy miguelk at KONSULTEX.COM.BR
Thu Sep 4 18:49:27 IST 2003


Stephen;

This comment is based on the assumprion that you have the satandard 
kernel that came with it. My previous server (2 months ago) was running 
RH 7.1 and all was fine as long as I did not bother it too much with new 
software, more processes, etc. Early this year I upgraded php, apache, 
snort, Perl and some other things (not Mailscanner, though) which I 
can't remember right now. Since then it also had lapses when it started 
just dying. Mine would run form about 1 or 2 weeks. The basic problem 
with that kernel is that the virtual memory process killer (the OOM 
killer) sometimes makes a mistake and kills the kernel. That happened in 
my case because the new processes ate up more ram than before and made 
the OOM act more often. Your case, though may be different. I would 
suggest a kernel update o a distribution update, if you can do that.

Miguel

Julian Field wrote:

> Does the RAM in use gradually creep up, or does it suddenly go to 100%?
>
> At 17:48 04/09/2003, you wrote:
>
>> I'm running MailScanner from an RPM install on Red Hat 7.1 on a 800mghz
>> Pentium 3 with 256MB ram and a 256MB swap. The server will run for
>> 24hours then die. The problem only started after installing MailScanner.
>> I have mailscanner configured not to stop spam and with 3 child procs
>> running, MailScanner v 4.23-11 RPM install. I have been using ClamAV for
>> virus scanning. Any ideas why MailScanner seems to be dying? I set the
>> auto restart config value to 10480 seconds also.
>>
>> -- 
>> Stephen Breen
>> c:Spot InterWorks
>
>
> -- 
> Julian Field
> www.MailScanner.info
> Professional Support Services at www.MailScanner.biz
> MailScanner thanks transtec Computers for their support
>



-- 
Esta mensagem foi verificada pelo sistema de antivírus e
 acredita-se estar livre de perigo.




More information about the MailScanner mailing list