Woof!

On Wed, 07 Oct 2009 14:05:37 -0400, Dale Worley <[email protected]> wrote:

> The behavior you are seeing is the currently intended behavior.  There
> was extensive discussion about the retry strategy at the time it was
> revised, so it is probably more sensible than it first appears.  Off the
> top of my head, the fact that a sender hasn't seen a 100 for a request
> doesn't mean that the TCP connection has failed -- e.g., the other end
> might be a stateless proxy that is sending the request on via UDP.
>
> In any case, I would track down the design discussion of the current
> behavior before considering changing it.

What about the not closing the TCP/IP connection after a timeout, and then  
re-using it?  That's the part I'm concerned with.  A new connection could  
end up working (due to lots of reasons, such as NAT screwups, DNS round  
robin picking a different machine, etc.), while re-using one that has  
failed in the past seems like a good chance of failure in the future.

--Woof!
_______________________________________________
sipx-dev mailing list [email protected]
List Archive: http://list.sipfoundry.org/archive/sipx-dev
Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev
sipXecs IP PBX -- http://www.sipfoundry.org/

Reply via email to