More recently, I've reached 200+ msgs/second on my laptop (dual-core) with the Ode 1.1 branch. I'm expecting the trunk (with all the thread/transaction refactoring) to put us in overdrive once it's stabilized. But don't believe us, go benchmark it! :)
alex On 9/14/07, Matthieu Riou <[EMAIL PROTECTED]> wrote: > > Hi, > > To add on Paul's reply, Alex had done some performance measurements some > time ago. See: > > http://www.nabble.com/Ode-Performance%3A-Round-I-tf3880525.html#a10997631 > > We would probably fare better now as several of the improvements Alex > mentions have been implemented. Which doesn't mean there's not more to do, > there's actually still quite a bit of possible optimizations. An important > thing to note on that side is that ODE support both persistent and in > memory > processes living at the same time within the same engine. So you can where > you want to be fast and where you want to be reliable. > > Other than that, I believe the pi-calculus architecture is better. From a > user standpoint you don't directly care much, but it has important > implications on scalability and maintenance that you benefit from. > > Finally and most crucial: all committers are nice guys :-) > > Cheers, > Matthieu > > On 9/14/07, Paul Brown <[EMAIL PROTECTED]> wrote: > > > > Hi, Thilina -- > > > > Well, albeit that I'm certainly biased, here are a few things that > > folks won't argue with: > > > > - ODE is the only BPEL engine with a business-friendly, > > extender-friendly license. There are other engines, who shall remain > > nameless, that are available under a GPL or LGPL license. > > > > - ODE has a diverse and open community that is neither controlled nor > > dominated by any commercial entity. > > > > - ODE provides a good level of flexibility for integration into other > > middleware (AXIS2, ServiceMix, TBD). > > > > - Since the olden days, ODE (since before it was ODE) has won customer > > "bake-off" competitions against other BPEL engines. Typically, ODE > > was at least even on speed (pretty constant constraint of throughput) > > and better on ergonomics and compliance. > > > > -- Paul > > > > > > > > On 9/14/07, Thilina Gunarathne <[EMAIL PROTECTED]> wrote: > > > Hi all, > > > Is there any performance data or performance comparisons available for > > ODE. > > > > > > I would also like to know how ODE will be compared against other > > > available BPEL engines. I mean what are the most compelling features > > > of ODE which will rank it higher than the other engines.. > > > > > > thanks, > > > Thilina > > > > > > PS: Hope nobody will start a flame war. I just need the date :). > > > -- > > > Thilina Gunarathne - http://thilinag.blogspot.com > > > > > > > > > -- > > [EMAIL PROTECTED] > > http://mult.ifario.us/ > > >
