[omniORB] omniorb and omniorbpy speed

[Sai-Lai Lo]

To make a fair comparision, why not compile omniORB 3.0.2 with CC 4.2,
under the same optimisation?


>>>>> Timothy Docker writes:

> Hi,

> I'm (still) evaluating using omniorbpy in lieu of fnorb. As I has
> hoped, omniorbpy is fast, but not as quick as I might have hoped. I 
> have written a simple client that requests a large data structure from 
> a remote service. I have written the client 3 times:

> 	- Orbacus 3.3.2 (the orb we use for C++ development), compiled
>           with sun CC-4.2

> 	- omniORB 3.0.2 compiled with g++ 2.95

> 	- omniORBpy (on top of omniORB 3.0.2) as above.

> (all under solaris). When run, I get the following execution times....

>  | tcc2:any_test $ time ./obtest 
>  | 4961 rows retrieved
>  | 
>  | real    0m2.688s
>  | user    0m0.630s
>  | sys     0m0.040s
>  | tcc2:any_test $ time ./omnitest 
>  | 4961 rows retrieved
>  | 
>  | real    0m4.276s
>  | user    0m2.130s
>  | sys     0m0.170s
>  | tcc2:any_test $ time ./omnipy_test
>  | 4883 rows retrieved
>  | (with 12 cols)
>  | 
>  | real    0m11.878s
>  | user    0m9.410s
>  | sys     0m0.170s
>  | tcc2:any_test $ 


-- 
Sai-Lai Lo                                   S.Lo@uk.research.att.com
AT&T Laboratories Cambridge           WWW:   http://www.uk.research.att.com 
24a Trumpington Street                Tel:   +44 1223 343000
Cambridge CB2 1QA                     Fax:   +44 1223 313542
ENGLAND