Bluehost.com Web Hosting $6.95

Re: [AMaViS-user] Berkeley DB blocking during heavy load?

This is a discussion on Re: [AMaViS-user] Berkeley DB blocking during heavy load? within the Amavis User forums, part of the Anti-Spam and Anti-Virus Related Forums category; On Wed, Apr 06, 2005 at 04:43:41PM -0400, John Morrissey wrote: > On Wed, Apr 06, 2005 at ...


Go Back   Usenet Forums > Anti-Spam and Anti-Virus Related Forums > Amavis User

FAQ Members List Calendar Search Today's Posts Mark Forums Read
  #1 (permalink)  
Old 04-06-2005
Clifton Royston
 
Posts: n/a
Default Re: [AMaViS-user] Berkeley DB blocking during heavy load?

On Wed, Apr 06, 2005 at 04:43:41PM -0400, John Morrissey wrote:
> On Wed, Apr 06, 2005 at 01:39:37AM +0200, Mark Martinec wrote:
> > > I'm running amavisd-new 20040701-1

> >
> > I suggest 2.2.1 for production use.

>
> Ok, I'll start looking at upgrading.
>
> > > on some large-ish machines (dual 3GHz CPUs, 4GB RAM) doing virus
> > > filtering only. Under high load (more than a couple dozen amavisd
> > > processes active),

> >
> > Did you arrive at the optimum number of processes by measuring throughput?

>
> That's the gist of what we're trying to do, but we can't come up with an
> honest number when there is so much lock contention. They obviously have
> more horsepower available (plenty of RAM available, state is stored on a
> tmpfs so very little disk I/O is involved, and plenty of CPU remaining, just
> a lot of processes spinning on lock acquisition).


I suspect you might wind up with better CPU utilization with fewer
processes. Off the cuff, in postfix content filter mode, I think 10-12
concurrent processes should keep SA busy soaking up all the CPU it can
get.

BTW, unless you are running a large DB on the same machine, it seems
to me 4GB RAM is overkill. (Even with a substantial portion dedicated
to a RAM-based /var/amavis fs.)

[...]
> That begs the question of what happens if the remaining BDB use also causes
> excessive lock contention. The stats will be useful to us, and I'd like to
> have them. Is there any way of alleviating the contention? Do the stats
> modify a single state table? If so, a transactional BDB environment wouldn't
> help, since that code probably needs atomic increments.


(Sorry, can't help on this one.)

-- Clifton

--
Clifton Royston -- cliftonr@tikitechnologies.com
Tiki Technologies Lead Programmer/Software Architect
"I'm gonna tell my son to grow up pretty as the grass is green
And whip-smart as the English Channel's wide..."
-- 'Whip-Smart', Liz Phair


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
AMaViS-user mailing list
AMaViS-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/...fo/amavis-user
AMaViS-FAQ:http://www.amavis.org/amavis-faq.php3
AMaViS-HowTos:http://www.amavis.org/howto/
Reply With Quote
Reply
Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are Off
[IMG] code is Off
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On




All times are GMT +1. The time now is 12:40 PM.


Powered by vBulletin® Version 3.7.3
Copyright ©2000 - 2008, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO 3.0.0