-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, I just want to let you know, that I started some work on Q.931 TE state timers in current openzap ozmod_isdn stack. The current stack has some problems with freeing ressources when far end doesn't follow Q931 state machine cleanly. On my side I mainly miss a RELEASE after sending DISCONNECT and so m more and more channels are wasted over time. Openzap Q931 stack has some preworked timer and state handling (in form of Q931_UX states). Unfortunately it is disabled in source. I enabled it, added some Q931_UX states to Q931StateTE.c, added a timeout handler to ozmod_isdn.c and tested it with my special problem. In lab missed RELEASEs are now detected and the corresponding channel is freed cleanly - - hurray :) I would like to put my work in openzap trunk as soon as it works stable in production and FS board allows me to do so. I can't promise that I will implement all timers for Q931 TE nor I plan to work on Q931 NT timers. As I said: Just for your information. Have a nice weekend! Helmut -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32) iD8DBQFKMmbl4tZeNddg3dwRAvWbAJ43Vex8J3LiVMu9mvs4US23eif19gCeO37l 4CGbk8CA+n+/dMRo5y5A+i0= =ZtvM -----END PGP SIGNATURE-----
please make sure you stay tuned into #openzap and coordinate with stkn and the other guys doing work on the stack. That way we can make sure we get the best out of the code. On Fri, Jun 12, 2009 at 9:32 AM, Helmut Kuper <helmut.kuper@ewetel.de>wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hello,
I just want to let you know, that I started some work on Q.931 TE state timers in current openzap ozmod_isdn stack. The current stack has some problems with freeing ressources when far end doesn't follow Q931 state machine cleanly. On my side I mainly miss a RELEASE after sending DISCONNECT and so m more and more channels are wasted over time.
Openzap Q931 stack has some preworked timer and state handling (in form of Q931_UX states). Unfortunately it is disabled in source. I enabled it, added some Q931_UX states to Q931StateTE.c, added a timeout handler to ozmod_isdn.c and tested it with my special problem. In lab missed RELEASEs are now detected and the corresponding channel is freed cleanly - - hurray :)
I would like to put my work in openzap trunk as soon as it works stable in production and FS board allows me to do so.
I can't promise that I will implement all timers for Q931 TE nor I plan to work on Q931 NT timers.
As I said: Just for your information.
Have a nice weekend!
Helmut -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32)
iD8DBQFKMmbl4tZeNddg3dwRAvWbAJ43Vex8J3LiVMu9mvs4US23eif19gCeO37l 4CGbk8CA+n+/dMRo5y5A+i0= =ZtvM -----END PGP SIGNATURE-----
_______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ AIM: anthm MSN:anthony_minessale@hotmail.com <MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org <sip%3A888@conference.freeswitch.org> iax:guest@conference.freeswitch.org/888 googletalk:conf+888@conference.freeswitch.org<googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:213-799-1400
Helmut Kuper wrote:
Hello,
I just want to let you know, that I started some work on Q.931 TE state timers in current openzap ozmod_isdn stack. The current stack has some problems with freeing ressources when far end doesn't follow Q931 state machine cleanly. On my side I mainly miss a RELEASE after sending DISCONNECT and so m more and more channels are wasted over time.
Openzap Q931 stack has some preworked timer and state handling (in form of Q931_UX states). Unfortunately it is disabled in source. I enabled it, added some Q931_UX states to Q931StateTE.c, added a timeout handler to ozmod_isdn.c and tested it with my special problem. In lab missed RELEASEs are now detected and the corresponding channel is freed cleanly - hurray :)
I would like to put my work in openzap trunk as soon as it works stable in production and FS board allows me to do so.
I can't promise that I will implement all timers for Q931 TE nor I plan to work on Q931 NT timers.
As I said: Just for your information.
Have a nice weekend!
Helmut
Umm, you've been doing duplicate work then. The version of ozmod_isdn i have been working on is completely stateful and has a couple of timers already implemented. And i remember giving you the location of the git repository on IRC, earlier this year. (But never got any feedback) stkn _______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi Stefan, On 12.06.2009 19:21, Stefan Knoblich wrote:
Umm, you've been doing duplicate work then.
:( Well, I implemented just one timer by now. So not much time has been wasted ...
The version of ozmod_isdn i have been working on is completely stateful and has a couple of timers already implemented.
Very good :)
And i remember giving you the location of the git repository on IRC, earlier this year. (But never got any feedback)
Sorry, at that time we talked about q931 to pcap. By now I thought state timers are still not done, so there wasn't a reason to test it regarding state timers .... buuuut today things changed and I will download your openzap. regards Helmut -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32) iD8DBQFKNfDw4tZeNddg3dwRAkThAJ4iPiZ4ZOAKKWpmKdCbjM8oM5mH6QCeMo5m pV4Y1/hpO7osV8cuInYJd2o= =zoZu -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello community, to keep you informed about this. I got an early copy of the new openzap ISDN stack. I did some changes to get it running with my uplink provider. Most were dedicated to my uplink provider (AVAYA PBX, e.g. needed IEs in CALL_PROC, missed CON ACK, debug messages, ...) and a few were more general (like auto recover after a long term E1 link disconnect, not handled PROGRESS messages in cetrain call states, addind Timerhandlinf or call tear down) Since last friday I use it my prod environment (E1, Q931-TE, enBlock-Dialing). I have currently no problems. Timers work as expected and seems to keep the stack clean, although not all timers are implemented, yet ... So far Stefan and Tony did a very good job so far. Here is a snippet from my log showing the handling of my old originate problem of a missed RELEASE message from uplink after sending a DISCONNECT by the NEW stack: 2009-07-20 11:05:40.208698 [DEBUG] ozmod_isdn.c:1999 WRITE 9 - -------------------------------------------------------------------------------- [08 02 80 1e 45 08 02 81 90] 2009-07-20 11:05:40.208698 [NOTICE] Span:1 Starting timer 6 (timeout: 30000 ms) for call 30 [0x1e] 2009-07-20 11:05:40.208698 [DEBUG] Span:1 Call going from state 8 -> 11 2009-07-20 11:05:40.208698 [DEBUG] Span:1 Q931Rx43 return code: 0 [Waiting for RELEASE] 2009-07-20 11:06:10.279950 [NOTICE] Span:1 Stopping timer 6 for call 30 [0x1e] 2009-07-20 11:06:10.279950 [NOTICE] Span:1 Timer T305 timed out for call 30 2009-07-20 11:06:10.279950 [DEBUG] Span:1 Receiving message from Layer4 (size: 109, type: 77 [0x4d], CRV: 30, CRVFlag: 0) 2009-07-20 11:06:10.279950 [NOTICE] Span:1 Handling message from Layer4 in call state: 11 2009-07-20 11:06:10.279950 [DEBUG] Span:1 Processing RELEASE message from Local for CRV: 30 (0x1e) 2009-07-20 11:06:10.279950 [DEBUG] Span:1 Sending message to Q.921 (size: 109) 2009-07-20 11:06:10.279950 [DEBUG] Span:1 Creating Q.931 Message Header: ProtDisc 8 (0x8), CRV 30 (0x1e), CRVflag: 1 (0x1), MesType: 77 (0x4d) 2009-07-20 11:06:10.279950 [DEBUG] Span:1 XXX Adding IE 0x8 2009-07-20 11:06:10.279950 [DEBUG] Span:1 XXX Adding IE 0x28 2009-07-20 11:06:10.279950 [DEBUG] Span:1 XXX Adding IE 0x34 2009-07-20 11:06:10.279950 [DEBUG] Span:1 XXX Adding IE 0x1c 2009-07-20 11:06:10.279950 [NOTICE] Span:1 Unable to get reference for IE 28 (0x1c) 2009-07-20 11:06:10.279950 [DEBUG] Span:1 XXX Adding IE 0xd 2009-07-20 11:06:10.279950 [NOTICE] Span:1 Unable to get reference for IE 13 (0xd) 2009-07-20 11:06:10.279950 [DEBUG] ozmod_isdn.c:1999 WRITE 9 - -------------------------------------------------------------------------------- [08 02 80 1e 4d 08 02 81 e6] 2009-07-20 11:06:10.279950 [NOTICE] Span:1 Starting timer 9 (timeout: 4000 ms) for call 30 [0x1e] 2009-07-20 11:06:10.279950 [DEBUG] Span:1 Call going from state 11 -> 19 2009-07-20 11:06:10.279950 [DEBUG] Span:1 Q931Rx43 return code: 0 2009-07-20 11:06:10.419951 [DEBUG] ozmod_isdn.c:1181 READ 9 - -------------------------------------------------------------------------------- [08 02 00 1e 5a 08 02 81 d1] 2009-07-20 11:06:10.419951 [DEBUG] Span:1 Call is in state 19 [Release request] 2009-07-20 11:06:10.419951 [DEBUG] Span:1 Received message from Q.921 (ind 4, tei 0, size 13) MesType: 90, CRVFlag 0 (Originator), CRV 30 (Dialect: Q.931 TE) 2009-07-20 11:06:10.419951 [DEBUG] Span:1 Processing RELEASE COMPLETE message from Remote for CRV: 30 (0x1e) 2009-07-20 11:06:10.419951 [NOTICE] Span:1 Stopping timer 9 for call 30 [0x1e] 2009-07-20 11:06:10.419951 [DEBUG] Span:1 Sending message to Layer4 (size: 110) 2009-07-20 11:06:10.419951 [DEBUG] ozmod_isdn.c:696 Yay I got an event! Type:[5a] Size:[110] CRV: 30 (0x1e, CTX: Originator) 2009-07-20 11:06:10.419951 [DEBUG] ozmod_isdn.c:728 zchan 8124e08 (1:25) source isdn_data->channels_remote_crv[0x1e] 2009-07-20 11:06:10.419951 [DEBUG] ozmod_isdn.c:799 Received Release Complete message for channel 0 2009-07-20 11:06:10.419951 [DEBUG] ozmod_isdn.c:804 Changing state on 1:25 from HANGUP to DOWN :D A known bug is the handling of incomming calls aiming for non registered, but known targets. Currently caller isn't informed about that and so its call stays open. Will try to fix that. regards Helmut -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32) iD8DBQFKZDx94tZeNddg3dwRApjxAJ4tynu+xrzM3uMWlAhgnp2fdjmi0wCfZvgU 1wxVCgyskanESU0UxHQ4Ars= =4xqu -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, I fixed the bug that openzap doesn't send a RELEASE on incomming calls for not registered but existing extensions. Day 3 running the new Q931-TE stack is still successful :) It seems I have still no open calls left after 1000+ of total calls ... regards helmut -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32) iD8DBQFKZY/a4tZeNddg3dwRAjYRAJ9fV1BJiJyyyrG2A5BWEUbhJZQ1bgCgkazo G8lw40GrRvfynDmYDZrLUU0= =pZ2r -----END PGP SIGNATURE-----
Helmut, This is wonderful news. Please keep up the good work. -MC On Tue, Jul 21, 2009 at 2:52 AM, Helmut Kuper <helmut.kuper@ewetel.de>wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hello,
I fixed the bug that openzap doesn't send a RELEASE on incomming calls for not registered but existing extensions.
Day 3 running the new Q931-TE stack is still successful :) It seems I have still no open calls left after 1000+ of total calls ...
regards helmut
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32)
iD8DBQFKZY/a4tZeNddg3dwRAjYRAJ9fV1BJiJyyyrG2A5BWEUbhJZQ1bgCgkazo G8lw40GrRvfynDmYDZrLUU0= =pZ2r -----END PGP SIGNATURE-----
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, just to close this thread, I would like to say that after 21000 pstn-calls and 45 days up time (old stack: max 10 days) the ISDN call resource management based on Q931 timers delivered by the *new* Q931 stack is still clean. Each done call was torn down to the DOWN state cleanly. This doesn't mean that the new stack is done - but for simple calls to and from PSTN it works very good compared to the old Q931 stack. regards Helmut -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32) iD8DBQFKnnak4tZeNddg3dwRAgPEAJsH+wg4CKV2MKQPAUbO69iijR/r4ACgug+a zLeZX1bmYNF4UlirU0XJHfc= =B/LZ -----END PGP SIGNATURE-----
w00t! On Wed, Sep 2, 2009 at 6:44 AM, Helmut Kuper <helmut.kuper@ewetel.de> wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hello,
just to close this thread, I would like to say that after 21000 pstn-calls and 45 days up time (old stack: max 10 days) the ISDN call resource management based on Q931 timers delivered by the *new* Q931 stack is still clean. Each done call was torn down to the DOWN state cleanly.
This doesn't mean that the new stack is done - but for simple calls to and from PSTN it works very good compared to the old Q931 stack.
regards Helmut -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32)
iD8DBQFKnnak4tZeNddg3dwRAgPEAJsH+wg4CKV2MKQPAUbO69iijR/r4ACgug+a zLeZX1bmYNF4UlirU0XJHfc= =B/LZ -----END PGP SIGNATURE-----
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
Great stuff, thanks for your hard work :). Keep it up. Regards, Diego On Wed, Sep 2, 2009 at 1:44 PM, Helmut Kuper <helmut.kuper@ewetel.de> wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hello,
just to close this thread, I would like to say that after 21000 pstn-calls and 45 days up time (old stack: max 10 days) the ISDN call resource management based on Q931 timers delivered by the *new* Q931 stack is still clean. Each done call was torn down to the DOWN state cleanly.
This doesn't mean that the new stack is done - but for simple calls to and from PSTN it works very good compared to the old Q931 stack.
regards Helmut -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32)
iD8DBQFKnnak4tZeNddg3dwRAgPEAJsH+wg4CKV2MKQPAUbO69iijR/r4ACgug+a zLeZX1bmYNF4UlirU0XJHfc= =B/LZ -----END PGP SIGNATURE-----
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi Diege, the flowers has to go to Stefan Knoblich not to me ;) He did the coding! regards Helmut On 02.09.2009 17:28, Diego Viola wrote:
Great stuff, thanks for your hard work :).
Keep it up.
Regards,
Diego -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32)
iD8DBQFKn4ZG4tZeNddg3dwRAt63AJ4kPRXMzCVWEZhQ+DLHePytAunGLgCeOY45 f56lUcvOv5q4jsc/Gjljt9I= =860q -----END PGP SIGNATURE-----
participants (5)
-
Anthony Minessale
-
Diego Viola
-
Helmut Kuper
-
Michael Collins
-
Stefan Knoblich