Re: Re: [certi-dev] CERTI performance issue

[Eric Noulard]

2009/6/16 Michael Raab <address@hidden>:
> Hi all,
>
> got the information on our internal tick usage.
> Internally we've used the tick function without parameters, which resulted
> in the discussed performance issues.
>
> I did a test using tick(min, max) and got a great performance boost.
> Execution time went down from 1200s to 51s,

Look good indeed :-)
What are the min and max value used?
How is the observed load of the machine now? Is it high?

> while the number of sent null
> messages and tick requests stayed nearly the same.

A silly question but,
Are the simulation result the same or at least similar
within your accepted tolerance?

> Some ideas how this is possible?

I would say that should be related to the way windows handle
syscalls and may be scheduling.
However I'm no windows expert so that's just a guess.

However as pointed out in another mail any "sleep" statement in your
code should be responsible of the boost because:

   1) You've said us that when your test has poor perf; the CPU load
       was low, which looks like a sleep

   2) The number of message did not change sgnificantly meaning
        that this was not the primary perf. issue.

> If this perfomance stays the same in further tests, i think we can delay the
> implementation of second generation time management techniques.

Such a shame :-)
May be we can drop you runtime to 5 sec :-)


-- 
Erk