--
* Paul <paul@???> [20040114 11:08]: wrote:
> Hi All,
>
> I am just doing some fine tuning for our mail system and wondering if our
> retry configuration is correct.
> Below is what is currently set to, which I am pretty sure is the default for
> exim. We run a very large site on exim-3.36
> Our /var/spool/mail/input is 16GB and doesn't fluctuate that much, only
> slighty. I have graphs of it's size from mrtg
>
> Also whats the best way to run exim and its queue daemon? Currently we're
> doing /usr/sbin/sendmail -bd -q1h
> Is it better to do /usr/sbin/sendmail -bd and then do a seperate
> /usr/sbin/sendmail -q1h or is it better to just combine them both.
> Should we be using a lower -qx (minute or hour) setting? Someone mentioned
> 15m, is too much for our site?
I believe unleashing several queue runners might help reduce the size of
your input. 16GB is so huge I wonder what is stored there. Does your
site have over 1 million users? ;(
exim -bd -q15m should be fine, but I believe there are other
suggestions, like unleasing multiple queue runners.
> I have also considering using another server for all outgoing mail, so in
> theory instead of mailing waiting in /var/spool/input it goes via another
> server and sits in the queue on that server.
Why can't it just go straight to the destination?
> Are there any associated nasties with doing this? No hosts or sites will
> refuse my emails or anything wierd because the sending host isn't listed as
> an MX?
fallback hosts should (I think) be used only when you cannot deliver
immediately. For example, what difference does it make when all the mail
go and stay on the queue of the other box? To me it sounds more like a
simple case of shifting goal posts, not tackling the problem.
> Grand total summary
> -------------------
> At least one address
> TOTAL Volume Messages Hosts Delayed Failed
> Received 20GB 543163 48651 59840 11.0% 71048 13.1%
> Delivered 19GB 487914 5522
You received 20GB and deliver 19GB. Well, that seems fast! How comes
then you have an input directory of 16GB? Doesn't make sense to me, but
I could be having a bad day.
> Deliveries by transport
> -----------------------
> Volume Messages
> local_delivery 3886KB 666
> remote_smtp 1046MB 40524
> spam_transport 394KB 34
> virtual_delivery 17GB 446690
This adds up, according to the above stats.
> ######################################################################
> # RETRY CONFIGURATION #
> ######################################################################
>
> # This single retry rule applies to all domains and all errors. It specifies
> # retries every 15 minutes for 2 hours, then increasing retry intervals,
> # starting at 1 hour and increasing each time by a factor of 1.5, up to 16
> # hours, then retries every 8 hours until 4 days have passed since the first
> # failed delivery.
>
> # Domain Error Retries
> # ------ ----- -------
> * quota_7d
> * quota F,2h,15m; F,3d,1h
> * quota_7d
> * quota F,2h,15m; F,3d,1h
Aren't those rules redundant?
I have only changed retry rules for some intermittently connected hosts
that fetch mail from my server using ETRN.
cheers
- wash
+----------------------------------+-----------------------------------------+
Odhiambo Washington . WANANCHI ONLINE LTD (Nairobi, KE) |
<wash at wananchi dot com> . 1ere Etage, Loita Hse, Loita St., |
GSM: (+254) 722 743 223 . # 10286, 00100 NAIROBI |
GSM: (+254) 733 744 121 . (+254) 020 313 985 - 9 |
+---------------------------------+------------------------------------------+
"Oh My God! They killed init! You Bastards!"
--from a /. post
--
[ Content of type application/x-pkcs7-signature deleted ]
--
