Exim Mail Queue Clogging Up

[kirtchristensen]

Hi!

I have LMP installed on my VDS at http://www.myriadnetwork.com

My list is about 400K in size, and although I have a few sites on this VDS, I thought that LMP would run fine, and I'd be able to send at least a message a day to the list...

However, even with the Exim tweaks, it clogs up like crazy.

Talking to the server admins, they just tell me that that's as fast as Exim runs...

Right now, it's taking about 5 DAYS to send to the list.

Should I just give up on myriadnetwork and move my list (leave my sites) to a 1and1 server on Plesk and just run it by itself?

Kirt

[DW]

I don't recommend changing servers just yet.  There are a few things we can try.

I have seen this on servers using Exim sending a lot of email but haven't figured out how to fix it.  One client sending to more than a million users upgraded to a more powerful server and has had no problems since.  I think it's due to hardware and/or Exim "retry length" settings.  Decreasing the length of time over which Exim 'retries' failing messages before returning a bounce will result in a lower overall queue size.  Plesk/qmail's default is 7 days and I now tweak this down to 48 hours, but I'm not sure how to do this on Exim.

This URL looks promising: [link]

Retry rule examples
===================

Here are some example retry rules suitable for use when
"wonderland.fict.book" is a local domain:

     alice@wonderland.fict.book quota_5d  F,7d,3h
     wonderland.fict.book       quota_5d
     wonderland.fict.book       *         F,1h,15m; G,2d,1h,2;
     lookingglass.fict.book     *         F,24h,30m;
     *                          refused_A F,2h,20m;
     *                          *         F,2h,15m; G,16h,1h,1.5; F,5d,8h

I think the very last line might be a 'global' setting.. let's see..

The final rule in a retry configuration should always have asterisks in
the first two fields so as to provide a general catch-all for any
addresses that do not have their own special handling. This example
tries every 15 minutes for 2 hours, then with intervals starting at one
hour and increasing by a factor of 1.5 up to 16 hours, then every 8
hours up to 5 days.

Yes, this looks promising.  Try setting this and tweaking it down to 2 or 3 days.

You'll probably have/want to clear the queue and start over so your messages go out more quickly...  This looks good: [link]

Code: [Select]

exim -bpru|awk {'print $3'}|xargs exim -Mrm
I recommend stopping Exim before and starting it after running the above command to clear the queue:

Code: [Select]

/etc/rc.d/init.d/exim stop

Code: [Select]

/etc/rc.d/init.d/exim start
Please let me know how you fare - information about this could benefit a large number of ListMail users!

Regards

[kirtchristensen]

Ok Dean, here's what I'm trying:

*                      *           F,1h,30m; G,12h,1h,1.5; F,2d,8h

Instead of:

*                      *           F,2h,15m; G,16h,1h,1.5; F,4d,8h

I'm using whm/cpanel, so I used the whm exim config editor and placed the new line BELOW the old line, should that work?

(My hosting company tells me that if I go in and edit the exim config file, it will just get overwritten by whm/cpanel anyway.)

Kirt

[DW]

Kirt,

I think (but don't know if) the lines are processed top-down, so I'm not sure if that will work.  Why not place a # in front of the old line so it is not interpreted at all?

Regards

[kirtchristensen]

Dean,

I would, but the main tech guy at myriadnetwork.com said not to... It would just be overwritten by whm/cpanel anyway.

I tested another mailing today, and so far, no visible improvement.

I'm leaning toward just moving my mailing to a 1and1 server and not messing with it anymore.  For $100 a month, it's well worth it.

I can get 1and1 to send about 750k a day with normal tweaking, right?

Thanks for your help!

Kirt

[DW]

Kirt,

I can get 1and1 to send about 750k a day with normal tweaking, right?

Barring deliverability problems such as temporary blacklisting from AOL, HotMail or Yahoo and depending on your message size, yes - you should be able to achieve this rate.

Regards

[dl1217]

Hi,

Does anyone else have these retry problems on EXIM?

I figured out that this is what clogs my Mail Queue up with certain domains, for example snet,gmail,netzero,aol all send through fine and fast, but then soem domains liek comcast for example i receive these retry time not reached and such......  :evil:

Maybe i have to sort through domains and sned certain ones on PLESK and others on cpanel?

TY,