OSEC

Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com
Postfix Archives: Re: Slow downstream site wedged my Postfix...

Re: Slow downstream site wedged my Postfix...

Subject: Re: Slow downstream site wedged my Postfix...
From: Andrew McNamara (andrewmconnect.com.au)
Date: Wed Jan 26 2000 - 20:19:27 CST

>Mail for excess recipients would have to be deferred (leaving them
>in the incoming queue would make Postfix thrash over the same files
>as it is looking for deliverable recipients).

Okay. I'd come to this conclusion also.

>Code: in qmgr_message_assign(), when a destination has more than
>some number of (todo+busy) references, pass the recipient to
>qmgr_defer_recipient() instead of adding the recipient to the
>site-specific queue. Pick N large enough that it does not affect
>delivery to the local domain.

I was thinking the recommended setting for N would be around a quarter
of qmgr_message_active_limit (although I'd leave it to the site admin
to chose).

>However, this will make it difficult to do performance measurements
>or to do mass mailings that have many recipients at the same site.
>The problem is that the excess recipients will be delivered as a
>little burst, about 15 minutes later, and that the remainder of
>the deferred recipients for that site would be deferred again
>(causing much thrashing of the disk), with another little burst of
>deliveries about 15 minutes later, and so on, until the deferred
>queue has no more recipients for that site.

Would qmgr need to remove them from the deferred queue at all? I was
hoping it could just ignore jobs over the threshold.

I still don't know what to do about jobs that are over the threshold
for some recipients and not for others.

 ---
Andrew McNamara (System Architect)

connect.com.au Pty Ltd
Lvl 3, 213 Miller St, North Sydney, NSW 2060, Australia
Phone: +61 2 9409 2117, Fax: +61 2 9409 2111

This archive was generated by hypermail 2b27 : Wed Jan 26 2000 - 20:25:49 CST