![]() |
Re: VoIP / Asterisk / rtcomm
Actually maximCT on #maemo did try the ssh -w i suggested with rtcomm and it seems to work too (he was also the one testing the PPPoverSSH solution).
So it seems possible, i'll tell you more when i can gather some more info ... EDIT: a little off-topic, but in the mean time, is anyone able to use the nat modules here: https://garage.maemo.org/frs/?group_...release_id=649 I can't seem to insmod them, i get: insmod: cannot insert 'iptable_nat.ko': Unknown symbol in module (-1): No such file or directory REF: https://garage.maemo.org/projects/ipt-nat/ EDIT2: forget this, in the right order it works, but it still can't accept rules such as -j MASQ or SNAT :( |
Re: VoIP / Asterisk / rtcomm
Quote:
I'm trying to see what can be done with iproute2 but no success so far, but i'm learning ;-) If we can get the source IP to the tun0's IP adress, i'm pretty sure it would work now. Any quick fix from an iproute2 guru or, if possible, a working nat module poackage? :P |
Re: VoIP / Asterisk / rtcomm
Just a guess here, but has anyone tried forcing a default route addition with openVPN ? (like the "redirect-gateway" directive) If it's just a matter of the system not "seeing" the tun interface to use, maybe this would just force it's hand. Of course I could be totally wrong :)
|
Re: VoIP / Asterisk / rtcomm
Moonshine:
I tried with the -redirect-gateway def1 option, to no avail. Seems strange that the rtcomm worked with ppp and won't work with tun. |
Re: VoIP / Asterisk / rtcomm
Uhm, i didn't play with my SIP again until recently to prepare for a BT headset i just got...
First i was playing with my OS2007 and it would randomly work or die on answer from the other end etc... So i thought why not try OS2008, maybe i had bloated everything and destroyed rtcomm's install in the process... Same thing with 2008! Asterisk version is 1.4.13. All my other phones/softphones work but the n800 doesn't, here is my sip.conf user: [n800] ; fixme type=friend username=n800 quality=yes md5secret=570fc7518b4e5c8d3faca3fcf605132c dtmf=rfc2833 host=dynamic context=home When i dial in the home context another phone which i mapped to 1234, it rings and most of the time when that phone picks up, it just dies there. Sometimes it goes through. We're talking about all local phones here which is real weird! N800 on wifi LAN -> gateway with asterisk -> phone/ATA on wired LAN... Anyone ever had this random problem happen? Example debug i get in asterisk: WHEN IT WORKS: -- Executing [1234@home:1] Dial("SIP/n800-081e68c8", "SIP/spa2102b") in new stack -- Called spa2102b -- SIP/spa2102b-081f06f8 is ringing -- SIP/spa2102b-081f06f8 answered SIP/n800-081e68c8 -- Native bridging SIP/n800-081e68c8 and SIP/spa2102b-081f06f8 == Spawn extension (home, 1234, 1) exited non-zero on 'SIP/n800-081e68c8' (LAST LINE PROBABLY THE SAME AS WHEN IT DOESN'T WORK BUT ONLY AT THE END WHEN I HANG UP BY CHOICE NOT BY FORCE ;) WHEN IT DOESN'T WORK: -- Executing [1234@home:1] Dial("SIP/n800-081e68c8", "SIP/spa2102b") in new stack -- Called spa2102b -- SIP/spa2102b-081f40a0 is ringing -- SIP/spa2102b-081f40a0 answered SIP/n800-081e68c8 -- Native bridging SIP/n800-081e68c8 and SIP/spa2102b-081f40a0 == Spawn extension (home, 1234, 1) exited non-zero on 'SIP/n800-081e68c8' (SOME DELAY) [Nov 29 01:19:08] WARNING[24565]: chan_sip.c:12038 handle_response_invite: Re-invite to non-existing call leg on other UA. SIP dialog 'a182faa7-18bb-122b-32b2-00194fd4ead7'. Giving up. (I SOMETIME SEE THIS TOO) [Nov 29 01:21:57] WARNING[24565]: chan_sip.c:1938 retrans_pkt: Maximum retries exceeded on transmission f966787f-18bb-122b-5a8f-00194fd4ead7 for seqno 103 (Critical Request) |
Re: VoIP / Asterisk / rtcomm
Quote:
|
Re: VoIP / Asterisk / rtcomm
Quote:
But even with that fixed i still have the same symptoms. in: tcpdump -i eth1 n800-IP -A, from the gateway with asterisk, i can see: 11:31:08.732719 IP 192.168.1.1.11981 > 192.168.1.178.7079: UDP, length 44 11:31:08.738876 IP 192.168.1.178 > 192.168.1.1: ICMP 192.168.1.178 udp port 7079 unreachable, length 80 I got lucky and got a call to go through and had this in asterisk -rvvvvvv just around where it got disconnected after 12 minutes: [Nov 29 11:34:43] NOTICE[28729]: chan_sip.c:15460 sip_poke_noanswer: Peer 'n800' is now UNREACHABLE! Last qualify: 6 was still working a bit after the previous warning... [Nov 29 11:35:50] NOTICE[28729]: chan_sip.c:12336 handle_response_peerpoke: Peer 'n800' is now Reachable. (1428ms / 2000ms) then it died.... "Connection to one or more account lost" in rtcomm. == Spawn extension (home, 1234, 1) exited non-zero on 'SIP/n800-081edfb8' i probably hung up my home ATA/Phone here... [Nov 29 11:36:46] WARNING[28729]: chan_sip.c:12038 handle_response_invite: Re-invite to non-existing call leg on other UA. SIP dialog '347a5eeb-1910-122b-3595-00194fd4ead7'. Giving up. And after a while i see this. Same goes with qualify=no, nat=no ... :| |
Re: VoIP / Asterisk / rtcomm
Lots of variable here, but you might want to try adding:
canreinvite=no Then Asterisk won't allow the SIP clients to issue additional invites where they will try to talk directly with each other. (Which can create issues NAT/public addresses, rport, etc) |
Re: VoIP / Asterisk / rtcomm
Quote:
I'll keep testing and come back with results if i find something more... The weird thing is, my friend is using it with an Asterisk on OpenWRT without this canreinvite=no option and it works for him... same device/OS! |
Re: VoIP / Asterisk / rtcomm
Well, following the reported 1860 bug of rtcomm gave out a "hack" that finally enabled me to log in into my Asterisk over the OpenVPN link! It's not ideal, but it works for now.
First you have to establish the VPN link and take note of your assigned IP (10.x.x.x). Then , in a terminal window, log in as as root. Set up the SIP account (sip0 in my case) parameter "local-ip-address" as follows: mc-account set sip0 string:local-ip-address=10.x.x.x And voila! You can log into the Asterisk server. I am using my N800 with OS 2008, and the latest openvpn. |
All times are GMT. The time now is 08:10. |
vBulletin® Version 3.8.8