You are viewing a plain text version of this content. The canonical link for it is here.
Posted to server-user@james.apache.org by Jeremy Morony <sh...@solamente.org> on 2004/01/20 23:42:37 UTC
james load testing/performance stuff
Hello! I am curious if anyone else could chime in with what kind of
delivery rates they can achieve with James. I've posted before on this
subject, but I've only seen one other thread regarding this issue, and
it seemed like that person was getting much better rates.
With the set up I have I'm a bit shy of sending 8,000, 70k messages an
hour. What I really need is much closer to three times that. Currently
we use Lyris, which can achieve those rates given our setup. I'd prefer
to use James because of its flexibility and it being an open source
project. (Also I'm less than impressed with Lyris' customer service!)
The test I run is pretty simple. I have a mailet that essentially works
like a really dumb listserv -- mail is sent to a specific account, the
mail is then forwarded to a list of recipients (essentially a big loop
calling getMailetContext().sendMail(...)).
Here's my set up:
* 1 ghz processor, 1 gig ram, redhat 7.3
* some kind of T1 out... (I don't have the full details at the moment)
* james2.2
* nntp, fetchpop, and fetch mail commented out of assembly.xml and
config.xml
* dbfile spool
* 20 remote delivery threads
* 12 spool threads
* a fairly bare processor pipeline (root only forwards eve
* mysql4.1.something running on a separate host
* jvm running in server mode, heap set using -Xmx=380m
Given all of this, is this typical of James's performance?
Jeremy.
---------------------------------------------------------------------
To unsubscribe, e-mail: server-user-unsubscribe@james.apache.org
For additional commands, e-mail: server-user-help@james.apache.org
Re: james load testing/performance stuff
Posted by Serge Knystautas <se...@lokitech.com>.
Jeremy Morony wrote:
> With the set up I have I'm a bit shy of sending 8,000, 70k messages an
> hour.
Eh, that's using 155kbps, which is theoretically full T1 speed. Buy a
bigger pipe.
--
Serge Knystautas
President
Lokitech >>> software . strategy . design >> http://www.lokitech.com
p. 301.656.5501
e. sergek@lokitech.com
---------------------------------------------------------------------
To unsubscribe, e-mail: server-user-unsubscribe@james.apache.org
For additional commands, e-mail: server-user-help@james.apache.org