Hello! Using OpenSER 1.0.0 in my path I encounter it rather often, that (exactly the same) requests (no append_branch()) are resent by the proxy (!) after 2 to 40 ms. Any idea what triggers that? /walter
Hi Walter,
are you sure is not about retransmission due the lack of a reply?
regards, bogdan
Walter Schober wrote:
Hello! Using OpenSER 1.0.0 in my path I encounter it rather often, that (exactly the same) requests (no append_branch()) are resent by the proxy (!) after 2 to 40 ms. Any idea what triggers that? /walter
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
On Tue, February 14, 2006 19:56, Bogdan-Andrei Iancu said:
Hi Walter,
are you sure is not about retransmission due the lack of a reply?
Hi Bogdan! I also often see retransmissions which are transmitted too fast (<100 ms instead of 500ms).
If I remember correctly, someone told me this is due to the timer implementation in (open)ser which is not very accurate.
regards klaus
regards, bogdan
Walter Schober wrote:
Hello! Using OpenSER 1.0.0 in my path I encounter it rather often, that (exactly the same) requests (no append_branch()) are resent by the proxy (!) after 2 to 40 ms. Any idea what triggers that? /walter
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
Hi Klaus,
that's true - I guess I was the one telling that :). But if that's the case, you will see only one retransmission between within the *first* second. Walter says something about "2 to 40 ms"
regards, bogdan
Klaus Darilion wrote:
On Tue, February 14, 2006 19:56, Bogdan-Andrei Iancu said:
Hi Walter,
are you sure is not about retransmission due the lack of a reply?
Hi Bogdan! I also often see retransmissions which are transmitted too fast (<100 ms instead of 500ms).
If I remember correctly, someone told me this is due to the timer implementation in (open)ser which is not very accurate.
regards klaus
regards, bogdan
Walter Schober wrote:
Hello! Using OpenSER 1.0.0 in my path I encounter it rather often, that (exactly the same) requests (no append_branch()) are resent by the proxy (!) after 2 to 40 ms. Any idea what triggers that? /walter
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
On Wed, February 15, 2006 11:01, Bogdan-Andrei Iancu said:
Hi Klaus,
that's true - I guess I was the one telling that :). But if that's the case, you will see only one retransmission between within the *first* second. Walter says something about "2 to 40 ms"
It's is hard to tell if the "fast" retransmission is the only one, or if there is also a retransmission after 500ms as usually the SIP phones answers to fast (<500ms)
regards klaus
regards, bogdan
Klaus Darilion wrote:
On Tue, February 14, 2006 19:56, Bogdan-Andrei Iancu said:
Hi Walter,
are you sure is not about retransmission due the lack of a reply?
Hi Bogdan! I also often see retransmissions which are transmitted too fast (<100 ms instead of 500ms).
If I remember correctly, someone told me this is due to the timer implementation in (open)ser which is not very accurate.
regards klaus
regards, bogdan
Walter Schober wrote:
Hello! Using OpenSER 1.0.0 in my path I encounter it rather often, that (exactly the same) requests (no append_branch()) are resent by the proxy (!) after 2 to 40 ms. Any idea what triggers that? /walter
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
Klaus Darilion wrote:
On Wed, February 15, 2006 11:01, Bogdan-Andrei Iancu said:
Hi Klaus,
that's true - I guess I was the one telling that :). But if that's the case, you will see only one retransmission between within the *first* second. Walter says something about "2 to 40 ms"
It's is hard to tell if the "fast" retransmission is the only one, or if there is also a retransmission after 500ms as usually the SIP phones answers to fast (<500ms)
you are right - my mistake : I read "2 to 40 secs" instead of ms.
regards, Bogdan
Klaus Darilion wrote:
On Tue, February 14, 2006 19:56, Bogdan-Andrei Iancu said:
Hi Walter,
are you sure is not about retransmission due the lack of a reply?
Hi Bogdan! I also often see retransmissions which are transmitted too fast (<100 ms instead of 500ms).
If I remember correctly, someone told me this is due to the timer implementation in (open)ser which is not very accurate.
regards klaus
regards, bogdan
Walter Schober wrote:
Hello! Using OpenSER 1.0.0 in my path I encounter it rather often, that (exactly the same) requests (no append_branch()) are resent by the proxy (!) after 2 to 40 ms. Any idea what triggers that? /walter
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
Hi!
Pls correct me, if I'm wrong, but I always thought, that retransmissions should be done by the UA, not the proxy?
Attached is (a rather complex call scenario - but currently I don't have an easier one). 002.002.002.25 is the OpenSER. Look at Line 118 and 124 (unfortunately Ethereal can't display filter that scenario and save it to a separate file... Fragmented frames ... infos lost). They are just sent by the proxy 0.033s afterwards.
Nevertheless there are some more problems in that call: RTPProxy assigns same port to voice and video on reinvite in opposite direction and repeated ACK is not terminated in transaction and ftag on previous proxy isn't translated correctly, since transaction already gone there...
Not that I have a problem now, it's just disturbing.
BTW: RTPProxy bug workaround by if looseroute if method == INVITE || REFER unforce_rtp_proxy force_rtp_proxy
Br Walter
-----Original Message----- From: Bogdan-Andrei Iancu [mailto:bogdan@voice-system.ro] Sent: Wednesday, February 15, 2006 12:06 PM To: Klaus Darilion Cc: Walter Schober; users@openser.org Subject: Re: [Users] Requests resent by proxy?
Klaus Darilion wrote:
On Wed, February 15, 2006 11:01, Bogdan-Andrei Iancu said:
Hi Klaus,
that's true - I guess I was the one telling that :). But if that's the case, you will see only one retransmission between within the *first* second. Walter says something about "2 to 40 ms"
Just learned, that on receiving 100 Trying, the UA can give that responsibility to the proxy :-)
-----Original Message----- From: users-bounces@openser.org [mailto:users-bounces@openser.org] On Behalf Of Walter Schober Sent: Wednesday, February 15, 2006 5:52 PM To: users@openser.org Subject: RE: [Users] Requests resent by proxy?
Hi!
Pls correct me, if I'm wrong, but I always thought, that retransmissions should be done by the UA, not the proxy?
Attached is (a rather complex call scenario - but currently I don't have an easier one). 002.002.002.25 is the OpenSER. Look at Line 118 and 124 (unfortunately Ethereal can't display filter that scenario and save it to a separate file... Fragmented frames ... infos lost). They are just sent by the proxy 0.033s afterwards.
Nevertheless there are some more problems in that call: RTPProxy assigns same port to voice and video on reinvite in opposite direction and repeated ACK is not terminated in transaction and ftag on previous proxy isn't translated correctly, since transaction already gone there...
Not that I have a problem now, it's just disturbing.
BTW: RTPProxy bug workaround by if looseroute if method == INVITE || REFER unforce_rtp_proxy force_rtp_proxy
Br Walter
-----Original Message----- From: Bogdan-Andrei Iancu [mailto:bogdan@voice-system.ro] Sent: Wednesday, February 15, 2006 12:06 PM To: Klaus Darilion Cc: Walter Schober; users@openser.org Subject: Re: [Users] Requests resent by proxy?
Klaus Darilion wrote:
On Wed, February 15, 2006 11:01, Bogdan-Andrei Iancu said:
Hi Klaus,
that's true - I guess I was the one telling that :). But if that's the case, you will see only one retransmission between within the *first* second. Walter says something about "2 to 40 ms"
On Wed, February 15, 2006 17:52, Walter Schober said:
Hi!
Pls correct me, if I'm wrong, but I always thought, that retransmissions should be done by the UA, not the proxy?
Attached is (a rather complex call scenario - but currently I don't have an easier one). 002.002.002.25 is the OpenSER. Look at Line 118 and 124 (unfortunately Ethereal can't display filter that scenario and save it to a separate file... Fragmented frames ... infos lost). They are just sent by the proxy 0.033s afterwards.
Nevertheless there are some more problems in that call: RTPProxy assigns same port to voice and video on reinvite in opposite direction and repeated ACK is not terminated in transaction and ftag on previous proxy isn't translated correctly, since transaction already gone there...
I do not know if this is the reason in your scenario, but: Some SIP clients (eyebam) do falsely use the record-route headers of reINVITEs to build a new route set. Than it may happen that the ftag in the Route header (which indicates the direction) gets swapped. This is a bug in the client which can be avoided by do not use record_route for indialog requests (loose_route processed requests).
regards klaus
Not that I have a problem now, it's just disturbing.
BTW: RTPProxy bug workaround by if looseroute if method == INVITE || REFER unforce_rtp_proxy force_rtp_proxy
Br Walter
-----Original Message----- From: Bogdan-Andrei Iancu [mailto:bogdan@voice-system.ro] Sent: Wednesday, February 15, 2006 12:06 PM To: Klaus Darilion Cc: Walter Schober; users@openser.org Subject: Re: [Users] Requests resent by proxy?
Klaus Darilion wrote:
On Wed, February 15, 2006 11:01, Bogdan-Andrei Iancu said:
Hi Klaus,
that's true - I guess I was the one telling that :). But if that's the case, you will see only one retransmission between within the *first* second. Walter says something about "2 to 40 ms"
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
-
Bogdan-Andrei Iancu -
Klaus Darilion -
Walter Schober