When doing an attended transfer via Asterisk (typically *2) I’ve noticed that several channels are opened and this causes severe lag on my old system. The scenario is as follows: SIP extension 7335 dials SCCP extension 6000. 6000 transfers (attended) 7335 into a MeetMe conference. Then 6000 hangs up and dials back into the same MeetMe conference. The conference is now unusable because of latency. A blind transfer does not produce this. This is the resulting channels:
asterisk1*CLI> show channels
Channel Location State Application(Data)
SCCP/6000-00000009 86000@from-internal: Up MeetMe(86000|asMp)
SIP/7335-644f s@macro-dial:10 Up Bridged Call(Local/86000@from-
Zap/pseudo-166815924 s@from-pstn:1 Rsrvd (None)
Local/86000@from-int 86000@from-internal: Up MeetMe(86000|asMp)
Local/86000@from-int s@from-internal:1 Up Transferred Call(SIP/7335-644f
5 active channels
2 active calls
Here’s what a blind transfer looks like for the same scenario:
asterisk1*CLI> show channels
Channel Location State Application(Data)
SCCP/6000-0000000b 86000@from-internal: Up MeetMe(86000|asMp)
Zap/pseudo-115101211 s@from-pstn:1 Rsrvd (None)
SIP/7335-3a8d 86000@from-internal: Up MeetMe(86000|sMp)
3 active channels
2 active calls
Why 2 extra channels for the attended transfer? After all, the extension that does the transfer DOES hang up.
(Lag & open channels exist even when only using SIP extensions, it’s not a SCCP issue)
Asterisk 1.2.5