inconsistent performance

[Simon Jones]

Could be I guess, netstat just shows up the usual mass of smtp
connections - it's just really odd why it's ok on the most part but as
soon as somthing triggers a choke they all start backing the queue up
with messages.

I'll do some more checking around.

thanks Richard

SMJ

2008/7/7 Richard Frovarp <richard.frovarp at sendit.nodak.edu>:
> Simon Jones wrote:
>>
>> hello chaps,
>>
>> anyone have an idea why i'm seeing inconsistent performance on all 3
>> of my gateway servers?  Nothing shows up errors in the maillog and
>> MailScanner --Lint checks out ok but from time to time the machines
>> will choke and i'll stack 2k and rising messages up in the hold queue.
>>
>> one day they'll work fine and the hold queue will be normal, then all
>> of a sudden they'll start backing up.
>>
>> I have postfix reading from mysql for the relay_domains /
>> relay_recipients and transport maps as well as logging to a seperate
>> db on the same seperate db server for mailwatch.
>>
>> the mysql db has dual oppy's with 10gb ram so it's pretty beefy and
>> doesn't seem to be maxed at all.
>>
>> the gateways run fairly heavy but are by no means topping out.  I've
>> tried dropping max children and the batch processing settings but to
>> no avail.
>>
>> any ideas would be really appreciated.
>>
>> thanks
>>
>> SMJ
>>
>
> What about mail volume? Is it consistent across the days in question? Or,
> even if the numbers are the same, you might end up with bursty traffic on
> the days you backup, causing everything to fall behind. Botnets, mailing
> lists, all of that sort can drop traffic on you in one heck of a hurry at
> times.
> --
> MailScanner mailing list
> mailscanner at lists.mailscanner.info
> http://lists.mailscanner.info/mailman/listinfo/mailscanner
>
> Before posting, read http://wiki.mailscanner.info/posting
>
> Support MailScanner development - buy the book off the website!
>