Re: Fw: [Asterisk-Users] Possible Sip logic bug?
Sorry Clif, as a professional working with protocol analysis at corporations in more than 40 states, I should have known better. Never gave it a thought the issue could have been earlier in the call/session setup. I'll dig into that, and if still need help/suggestions will post the full debug trace. Rich - > It is very important (at least to me) to have the whole SIP call flow. > That is, I must see the initial > INVITE come from the originating phone all the way to the last message. > I can only speculate at > this point but it appears that the second leg (destination) may never > have ACK'd the call which > could have Asterisk in a bad state. I cannot be sure of this without > the entire flow but if this is the > case, not only do you have a config problem, Asterisk has an unhandled > error state.Did you > answer the destination? Did it have 2-way voice path? > > Rich Adamson wrote: > > >Clif and all... > > > >At the bottom of this post is the "sip show debug" for the problem. > >The underlying problem (again): when C7960 hangs up on working conversation, > >the C7960 sends a BYE to *, but * never sends a BYE to the sip gateway. > > > >Any suggestions would be greatly appreciated. > > > >Rich > > > > > > > >>Try it again after executing: "sip debug" and give us the actual SIP > >>messages. The devil > >>is usually in the details. > >> > >> > >> > >> > > > > > ___ > Asterisk-Users mailing list > [EMAIL PROTECTED] > http://lists.digium.com/mailman/listinfo/asterisk-users > To UNSUBSCRIBE or update options visit: >http://lists.digium.com/mailman/listinfo/asterisk-users ---End of Original Message- ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
Re: Fw: [Asterisk-Users] Possible Sip logic bug?
Rich, It is very important (at least to me) to have the whole SIP call flow. That is, I must see the initial INVITE come from the originating phone all the way to the last message. I can only speculate at this point but it appears that the second leg (destination) may never have ACK'd the call which could have Asterisk in a bad state. I cannot be sure of this without the entire flow but if this is the case, not only do you have a config problem, Asterisk has an unhandled error state.Did you answer the destination? Did it have 2-way voice path? Rich Adamson wrote: Clif and all... At the bottom of this post is the "sip show debug" for the problem. The underlying problem (again): when C7960 hangs up on working conversation, the C7960 sends a BYE to *, but * never sends a BYE to the sip gateway. Any suggestions would be greatly appreciated. Rich Try it again after executing: "sip debug" and give us the actual SIP messages. The devil is usually in the details. ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
Re: Fw: [Asterisk-Users] Possible Sip logic bug?
Clif and all...
At the bottom of this post is the "sip show debug" for the problem.
The underlying problem (again): when C7960 hangs up on working conversation,
the C7960 sends a BYE to *, but * never sends a BYE to the sip gateway.
Any suggestions would be greatly appreciated.
Rich
> Try it again after executing: "sip debug" and give us the actual SIP
> messages. The devil
> is usually in the details.
>
> Rich Adamson wrote:
>
> >Anyone have comments on this? Really could use some suggestions or ideas
> >why this is happening. Thanks.
> >Rich
> >
> >
> >
> >
> >>Anyone recognize this as a sip logic bug?
> >>
> >>Example Case:
> >> C7960 -> * -> sip gateway -> pstn
> >> (sip gateway config'ed with canreinvite=no, but shouldn't have an
> >> impact on this.)
> >>
> >>Outgoing call initiated from C7960. Call is completed and conversation
> >>is very much normal. All equipment on the same wire; no nat.
> >>
> >>The C7960 user hangs up the phone. Pkt flows (as observed by sniffer)
> >>are:
> >>
> >>C7960 sends sip BYE packet to *
> >> * returns 200 OK
> >>* sends INVITE to sip gateway<< where is BYE?
> >> sip gateway responds with 100 Trying
> >> sip gateway responds with 200 OK
> >> sip gateway responds with 200 OK
> >> sip gateway responds with 200 OK
> >>
> >>The end result, the sip gateway does not drop the pstn line until the
> >>"called" number hangs up.
> >>
> >>It would appear that asterisk has an issue dropping the call. When the
> >>C7960 issues the BYE, I would expect * to send a BYE to the sip g/w.
> >>Is this a * logic problem (or my logic problem)?
> >>
> >>(I'm actually running CVS-12/04/03-14:24:40 and has been very stable
> >>in this production environment. Is it time to update this one even
> >>though it is 99% sip hardphone based?)
> >>
> >>Rich
Note: Call is already established from C7960 (193.92) via * (193.101) to
the sip gateway (193.109) which called cell phone 444-1234. The "sip show
channels" was executed, followed by "sip debug", then hung up the C7960
watching the results below. Note the C7960 sends the BYE and * confirms,
but * never sends a BYE to the gateway. Sniffer on the wire confirms the
exact same thing.
phoenix*CLI>
phoenix*CLI> sip show channels
Peer User/ANRCall ID Seq (Tx/Rx) Lag Jitter Format
222.111.193.109 4441234 66841295427 00103/0 0ms ms ULAW
222.111.193.92 300000036bc3-8b 00102/00103 0ms ms ULAW
2 active SIP channel(s)
== Spawn extension (from-sip, 64441234, 2) exited non-zero on 'SIP/3000-375c'
-- Executing SetCIDNum("SIP/3000-ead2", "") in new stack
-- Executing Dial("SIP/3000-ead2", "SIP/[EMAIL PROTECTED]") in new sta
ck
-- Called [EMAIL PROTECTED]
-- SIP/222.111.193.109-fb6e answered SIP/3000-ead2
-- Attempting native bridge of SIP/3000-ead2 and SIP/222.111.193.109-fb6e
SIP Debugging Enabled
Sip read: I>
BYE sip:[EMAIL PROTECTED]:5060 SIP/2.0
Via: SIP/2.0/UDP 222.111.193.92:5060
From: "NPI-Rich" ;tag=00036bc38b88045b25941469-0a0c5ae
b
To: ;tag=as751f96fc
Call-ID: [EMAIL PROTECTED]
Date: Thu, 05 Feb 2004 15:13:50 GMT
CSeq: 103 BYE
User-Agent: CSCO/6
Content-Length: 0
Proxy-Authorization: Digest username="3000",realm="asterisk",uri="sip:222.111.19
3.101",response="bb01af8f1eac65d392b68147867e79e6",nonce="7660e36e",algorithm=md
5
10 headers, 0 lines
Sending to 222.111.193.92 : 5060 (non-NAT)
Transmitting (no NAT):
SIP/2.0 200 OK
Via: SIP/2.0/UDP 222.111.193.92:5060
From: "NPI-Rich" ;tag=00036bc38b88045b25941469-0a0c5ae
b
To: ;tag=as751f96fc
Call-ID: [EMAIL PROTECTED]
CSeq: 103 BYE
User-Agent: Asterisk PBX
Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER
Contact:
Content-Length: 0
to 222.111.193.92:5060
set_destination: Parsing for address/port to
send to
set_destination: set destination to 222.111.193.109, port 5060
We're at 222.111.193.101 port 14308
Answering with preferred capability 4
Answering with capability 8
Answering with non-codec capability 1
11 headers, 10 lines
Reliably Transmitting:
INVITE sip:[EMAIL PROTECTED]:5060 SIP/2.0
Via: SIP/2.0/UDP 222.111.193.101:5060;branch=z9hG4bK40ac3b17
From: "NPI-Rich" ;tag=as3310fadb
To: ;tag=8c44b610-98bad313
Contact:
Call-ID: [EMAIL PROTECTED]
CSeq: 104 INVITE
User-Agent: Asterisk PBX
Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER
Content-Type: application/sdp
Content-Length: 219
v=0
o=root 14743 14745 IN IP4 222.111.193.101
s=session
c=IN IP4 222.111.193.101
t=0 0
m=audio 14308 RTP/AVP 0 8 101
a=rtpmap:0 PCMU/8000
a=rtpmap:8 PCMA/8000
a=rtpmap:101 telephone-event/8000
a=fmtp:101 0-16
(no NAT) to 222.111.193.109:5060
== Spawn extension (from-sip, 64441234, 2) exited non-zero on 'SIP/3000-ead2'
Sip read: I>
SIP/2.0 100 Trying
Call-ID: [EMAIL PROTECTED]
CSeq: 104 INVITE
From: NPI-Rich ;tag=as3310fadb
To: ;tag=8c44b610-98bad313
Via: SIP/2.0/UDP 222.111.193.101:5060;branch=z9hG4bK40ac3b17
Conten
Re: Fw: [Asterisk-Users] Possible Sip logic bug?
Rich, Try it again after executing: "sip debug" and give us the actual SIP messages. The devil is usually in the details. Rich Adamson wrote: Anyone have comments on this? Really could use some suggestions or ideas why this is happening. Thanks. Rich Anyone recognize this as a sip logic bug? Example Case: C7960 -> * -> sip gateway -> pstn (sip gateway config'ed with canreinvite=no, but shouldn't have an impact on this.) Outgoing call initiated from C7960. Call is completed and conversation is very much normal. All equipment on the same wire; no nat. The C7960 user hangs up the phone. Pkt flows (as observed by sniffer) are: C7960 sends sip BYE packet to * * returns 200 OK * sends INVITE to sip gateway<< where is BYE? sip gateway responds with 100 Trying sip gateway responds with 200 OK sip gateway responds with 200 OK sip gateway responds with 200 OK The end result, the sip gateway does not drop the pstn line until the "called" number hangs up. It would appear that asterisk has an issue dropping the call. When the C7960 issues the BYE, I would expect * to send a BYE to the sip g/w. Is this a * logic problem (or my logic problem)? (I'm actually running CVS-12/04/03-14:24:40 and has been very stable in this production environment. Is it time to update this one even though it is 99% sip hardphone based?) Rich ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users ---End of Original Message- ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
Fw: [Asterisk-Users] Possible Sip logic bug?
Anyone have comments on this? Really could use some suggestions or ideas why this is happening. Thanks. Rich > Anyone recognize this as a sip logic bug? > > Example Case: > C7960 -> * -> sip gateway -> pstn > (sip gateway config'ed with canreinvite=no, but shouldn't have an > impact on this.) > > Outgoing call initiated from C7960. Call is completed and conversation > is very much normal. All equipment on the same wire; no nat. > > The C7960 user hangs up the phone. Pkt flows (as observed by sniffer) > are: > > C7960 sends sip BYE packet to * > * returns 200 OK > * sends INVITE to sip gateway<< where is BYE? > sip gateway responds with 100 Trying > sip gateway responds with 200 OK > sip gateway responds with 200 OK > sip gateway responds with 200 OK > > The end result, the sip gateway does not drop the pstn line until the > "called" number hangs up. > > It would appear that asterisk has an issue dropping the call. When the > C7960 issues the BYE, I would expect * to send a BYE to the sip g/w. > Is this a * logic problem (or my logic problem)? > > (I'm actually running CVS-12/04/03-14:24:40 and has been very stable > in this production environment. Is it time to update this one even > though it is 99% sip hardphone based?) > > Rich > > > ___ > Asterisk-Users mailing list > [EMAIL PROTECTED] > http://lists.digium.com/mailman/listinfo/asterisk-users > To UNSUBSCRIBE or update options visit: >http://lists.digium.com/mailman/listinfo/asterisk-users ---End of Original Message- ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
