Re: Help with core dump

[Bowie Owens <bowie owens csiro au>]

mmacleod webmail co za wrote:

> > Are we talking CPU utilisation or latency? ORBit was designed to be lean
> > and fast so I don't think it would be the problem there. For the most
> > part orbitcpp is a thin wrapper around ORBit. Try profiling your code
> > with gprof or some other tool. If you find too much time is taken up by
> > orbitcpp, I can look at improving its efficiency.
> >    
>
> My server has an internal loop in its factory object (in a different thread) 
> that executes every n seconds and updates a whole lot of stuff to new values,
> if my client trys to execute anything using the factory in this period, it 
> sits and waits for a long time then has a COMM failure, it doesnt do this in 
> mico though.
>
> Im confused as to why it would be different, what poa threading policy does 
> orbit(cpp) default too?
Single threaded by default.

> could this maybe be the cause?
>  
I would imagine that MICO starts up by default in single threaded as well.

> Any help would be appreciated thanks.
>  
What thread library are you using? Is it possible that the thread that 
the POA is in is getting starved for CPU time?

If you could produce a simple (as simple as possible) test case that 
reproduces the bug, I might be able to find some time to look at it. 
Otherwise you could try posting to the ORBit list.

--
Bowie Owens

CSIRO Mathematical & Information Sciences
phone  : +61 3 9545 8055
fax    : +61 3 9545 8080
mobile : 0425 729 875
email  : Bowie Owens csiro au