Re: ORBit speed measure

I've just tested CORBA::ORBit calling a null method over IIOP on a TAO
server 100000 times.

Over 100Mbit ethernet, I get 2564 invocations per second.
Over the loopback I get 7142 invocations per second

I think that's pretty good, especially from perl :)

BTW COPE gets 588 and 1428 respectively, which is not really
surprising as it is 100% perl.


Xavier Ordoquy writes:
 > Such as every test that has been done, it's always unfair because of this
 > or that.
 > However, I did not optimize this because I used my distribution
 > (slackware) stuffs. This is because I wanted to use it for some projects
 > and wanted to get an idea of the overhead. Thus as every user will not
 > recompile ORBit, I just took the straight distribution package. Therefore,
 > the times were provided as it.
 > If one can find bench between dynamic loading and ORBit (compiled with
 > what approched the most the dynamic loading) and benchs of ORBit between
 > the differents mean of transport (sockets, rpc, ...) it would greatly
 > interest lots of people. I just hope I didn't missed it.
 > I know CORBA is not an implementation, but to myself it's a language
 > abuse but it's commonly used.
 > Xavier Ordoquy
 > _______________________________________________
 > orbit-list mailing list
 > orbit-list gnome org

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]