Re: [Ekiga-list] Coexistence of ekiga with an ATA

[ael <law_ence dev ntlworld com>]

Julien Puydt wrote:
> ael a écrit :
> 
>>Initially there are 14 packets exchanged. This is while ekiga is
>>starting up. At this point attempted calls to, say 500 ekiga net, fail
>>usually with abnormal call termination.
> 
> 
> You mean ekiga wasn't registered yet to ekiga.net ?

No. Normally there is no problem with registration. This was explained
earlier in this thread. I suspect that the system is behaving in some
respects at least as a symmetric NAT when the STUN protocol is
nondeterministic, although I have not yet tried to analyse the STUN
traffic: I am a newbie and need to read the relevant rfcs. I hope to
find time to do that soon.

Meanwhile, things work with this odd fix of "reactivating" the STUN
setting. But it would be nice to understand what is happening and to get
total reliability.

I did say that the captured data was the STUN traffic: it was filtered
so that only packets exchanged with the stun servers were collected. So
the registration phase was not captured.

A E Lawrence