On 4/18/2013 10:40 AM, Simon Wilkinson wrote:
> Then, there is an additional question - is the hit you take from cache 
> invalidation worth terminating a call for? That's going to be a trade off, 
> but I suspect that the current behaviour just doesn't allow a server enough 
> time to tidy up.

This I believe is the critical question.   Is the additional cost of
invalidating the cache after receipt of the first RX_CALL_BUSY which
will force a new RXAFS_FetchStatus rpc to be issued before the call that
is canceled due to BUSY can be retried less expensive than the old
behavior of waiting for the call channel to become available or for a
timeout to be hit.

My gut reaction is that the additional expense both in terms of round
trip time and server load is not worth the optimization of immediately
retrying a call that received BUSY on another channel.

Jeffrey Altman


Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

Reply via email to