Re: [Sip] draft-jennings-sip-dtls

On Feb 17, 2005, at 2:00 PM, Christian Stredicke wrote:

> 16 bit set an upper limit of 64 K connections. My feeling is that TCP 
> was not designed for this purpose.
>  
> Sorry for being so pragmatic. Maybe someone can show me an 
> implementation that can handle 100 K connections on one IP address.
>  

Actually, we fairly thoroughly flogged this topic a few years ago.

The net-net was that the 64k limit is a kernel limit, induced by the 
use of 16-bit indices in file descriptor tables and other places.

Some kernels (or embedded IP limitations that are not kernel specific) 
don't seem to have this limitation. I'm told by some of the bootserver 
engineers I talk with occasionally that they can get several hundred 
thousand TCP connections on Solaris 9, at a cost of about 4kB per 
connection. Of course, this eats a boatload of RAM.

But DTLS does have some nice properties, and its ability to scale 
larger on a larger set of OS platforms is quite possibly one of them.

--
Dean

_______________________________________________
Sip mailing list  https://www1.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sipping@ietf.org for new developments on the application of sip