[Live-devel] Benchmarking

Pinky Jiri.Pinkava at vscht.cz
Mon Apr 9 16:38:06 PDT 2007


Dne Mon, 09 Apr 2007 19:14:43 -0400
Vlad Seryakov <vlad at crystalballinc.com> napsal(a):

> 
> Ross Finlayson wrote:
> >> Even with completely cached file, it seems to me that delay while
> >> scanning all sockets
> > 
> > Sets of sockets aren't 'scanned' anywhere in the code.  Instead,
> > the event loop is implemented (by default, with the
> > "BasicTaskScheduler" class) using "select()", which should be
> > efficient even with large numbers of readable sockets.
> > 
> > It's possible, however, that with large numbers of open sockets,
> > you might be running into some OS-imposed limit that could be
> > hurting performance.  I think Linux has ways of configuring the OS
> > to support more open sockets, so that's something that you might
> > want to look into.
> 
> Yes, i meant select, i am writing scheduler using poll to see if 
> performance will increase.
> 
> 
> > (Also, re. 'benchmarking' in general: Remember, You Have Complete
> > Source Code.)
> 

I have it done already and do some testing now.



> That was more informative question what others have, i am doing my
> own tests but it is always good to know what is going on outside my
> environment. _______________________________________________
> live-devel mailing list
> live-devel at lists.live555.com
> http://lists.live555.com/mailman/listinfo/live-devel


More information about the live-devel mailing list