incubator-wadi-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From lichtner <licht...@bway.net>
Subject Re: Replication using totem protocol
Date Fri, 13 Jan 2006 17:43:24 GMT

On Fri, 13 Jan 2006, James Strachan wrote:

> Yes, ActiveSpace currently has a distributed JCache using optimistic
> transactions which relies on total ordering; it could well be a good use
> case for using totem. I'd be interested in a performance comparison with
> Totem v ActiveMQ (indeed it'd be trivial to integrate Totem into ActiveMQ as
> a transport; we already have multicast, UDP, TCP, SSL, HTTP et al).

I need to impress upon people that I only have one box at home, and
therefore I can not do real tests. You can run totem on a single box, but
the results are slightly different because totem is cpu bound.

If somebody wants to see how fast EVS4J can go they need to download it
from bway.net/~lichtner/evs4j.html and read the instructions on that page
for running the benchmark class included in the library.

The benchmark consists of a N processors forming a ring and sending
receiving 1500-byte messages at will. With half the cpu on my machine I
get 6000 messages per second.

Pfister reports a study made that 90% of enterprise messages are 200 bytes
or less. With that size you get a lot more than 6000 per sec even though
I did not implement scatter/gather functionality.

> FWIW when I get chance I've been meaning to refactor the ActiveSpace code to
> make use of Spring Remoting / Lingo (probably moving it into the Lingo
> project) so it'd be much easier to decouple the remoting code easily.

I don't understand what this means here, but I have a feeling I probably
don't need to worry about it right now.

Guglielmo

Mime
View raw message