Comment #5 on issue 2347 by [email protected]: Diameter :
PeerFSMImpl.queueStart NPE
http://code.google.com/p/mobicents/issues/detail?id=2347
Hi,
Though this fix seems to solve the issue, I'm seeing a strange behaviour in
the trace :
22 Feb 10:47:53.511 INFO NetworkGuard:77 - Open incomming connection
Socket[addr=/my.cli.ent.ip,port=56269,localport=3868]
/** killing client **/
22 Feb 10:48:04.566 INFO TCPTransportClient:144 - Read thread is stopped
(unknown_host)
22 Feb 10:48:09.811 INFO NetworkGuard:77 - Open incomming connection
Socket[addr=/my.cli.ent.ip,port=45517,localport=3868]
Why TCPTransportClient doesn't recognize the lost connection ? without the
fix I have :
- NetworkGuard: Incoming from my.cli.ent.ip
/** restart client **/
- TCPTransportClient: Read thread stopped (unknown_host)
- NetworkGuard: Incoming from my.cli.ent.ip
/** issue 2347 behaviour **/
- TCPTransportClient: Read thread stopped (myclient host)
- NetworkGuard: Incoming from my.cli.ent.ip
- TCPTransportClient: Read thread stopped (myclient host)
etc.
I don't know if I'm really clear but I can't see why on the disconnect
triggered from the client the read thread is not identified while on the
next closes, this time triggered by the stack due to the issue, the read
thread identify successfully the connexion.
For what I know it doesn't seems to cause any problem, and it might even
not be related directly to this issue but it's good to know.
