| 00:21.20 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-ppbddvebnangkiqn) |
| 01:18.16 | *** join/#asterisk-dev DivideBy0 (~DivideBy0@unaffiliated/divideby0x0) |
| 01:35.17 | *** join/#asterisk-dev Bordr (~quassel@23-24-156-49-static.hfc.comcastbusiness.net) |
| 01:51.58 | *** join/#asterisk-dev Bordr (~Bordr@23-24-156-49-static.hfc.comcastbusiness.net) |
| 01:52.11 | *** part/#asterisk-dev Bordr (~Bordr@23-24-156-49-static.hfc.comcastbusiness.net) |
| 04:36.48 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 04:48.37 | *** join/#asterisk-dev craigify (~craigify@ip184-186-18-123.no.no.cox.net) |
| 04:52.52 | *** join/#asterisk-dev luckman212 (~luckman21@unaffiliated/luckman212) |
| 05:59.59 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 06:58.26 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 07:46.36 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 07:46.36 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 07:53.15 | *** join/#asterisk-dev luckman212 (~luckman21@unaffiliated/luckman212) |
| 08:00.57 | *** join/#asterisk-dev pchero_work (~pchero@109.70.54.56) |
| 08:03.22 | *** join/#asterisk-dev lumasepa (~gestoip@193.145.124.30.local.ull.es) |
| 08:14.11 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 10:42.00 | *** join/#asterisk-dev luckman212 (~luckman21@unaffiliated/luckman212) |
| 11:36.08 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 12:57.08 | *** join/#asterisk-dev aurs (~aurs@84.48.191.51) |
| 13:43.58 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@216-191-106-163.dedicated.allstream.net) |
| 13:55.16 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 13:56.33 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 14:08.22 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 14:12.37 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 14:14.04 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 14:43.54 | *** join/#asterisk-dev gtjoseph (~gtj0@unaffiliated/gtj) |
| 15:06.40 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-wjbssorcxzbpsgdb) |
| 15:29.21 | *** join/#asterisk-dev putnopvut (putnopvut@asterisk/master-of-queues/mmichelson) |
| 15:29.21 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 15:41.22 | gtjoseph | BillyCrook: ping |
| 15:41.54 | file | gtjoseph, are you going to retransmit your ping? |
| 15:42.12 | gtjoseph | this is my second retransmit. |
| 15:42.42 | file | ah, you retransmit as-is and don't add a counter or anything? |
| 15:43.38 | gtjoseph | :) |
| 15:44.09 | gtjoseph | it's a 2 day timeout |
| 15:44.21 | file | good to know |
| 15:50.36 | coreyfarrell | is git really slow for anyone else? been waiting 2 minutes for 'git review' to post a 6 line change (not done yet). |
| 15:55.22 | file | wonders if java needs a kick |
| 15:55.37 | gtjoseph | seems it's always slower in the mornings |
| 15:55.42 | file | a git pull was fine |
| 16:03.34 | coreyfarrell | not working for me, github is fine though. 'git review' timed out fetching gerrit - "ssh_exchange_identification: read: Connection reset by peer" |
| 16:04.15 | file | rawr |
| 16:07.17 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 16:08.58 | dan_j | file: do you recall the conversation we had about multi-homed ip with PJSIP using the wrong source ip? |
| 16:09.06 | file | yes |
| 16:09.14 | file | although I'm in a meeting, so my responses may be delayed |
| 16:10.10 | dan_j | You mentioned that asterisk lets the system decide the source ip. I've manually changed the source IP and confirmed it using 'ip route', however asterisk continues to use the original ip. Are you sure it lets the system decide? |
| 16:10.28 | file | yes, it is bound to 0.0.0.0 wildcard |
| 16:11.13 | dan_j | hmm. wonder why it's not changing then. Ok. will keep looking at it. Trying to work out a way to fix the system since i'm not too familiar with the asterisk source code. |
| 16:13.46 | gtjoseph | file: dan_j: hmmm this sounds like a problem i had last week where i had 2 transports set up, one bound to the internal address and the other to the external address but calls going out externally had the internal ip address. |
| 16:14.09 | gtjoseph | dan_j: is that your situation? |
| 16:14.11 | dan_j | gtjoseph: is that pjsip? |
| 16:14.13 | gtjoseph | yes |
| 16:14.36 | dan_j | To be honest, I'm not sure yet. I've only tested the public IP address. |
| 16:15.05 | gtjoseph | do you have a jira issues open for ithis? |
| 16:15.06 | dan_j | Heres the bug |
| 16:15.08 | dan_j | https://issues.asterisk.org/jira/browse/ASTERISK-25637 |
| 16:15.11 | gtjoseph | :) |
| 16:15.15 | gtjoseph | i'll take a look |
| 16:15.16 | dan_j | https://issues.asterisk.org/jira/browse/ASTERISK-25632 |
| 16:16.07 | dan_j | I'm just checking that my source ip settings are working so that I can be sure it's asterisk thats doing it. |
| 16:18.41 | dan_j | file: Confirmed. When I do ping tests, the source IP is the one i've specified and listed by 'ip route'. It appears that Asterisk is not letting the system determine the IP. I've restarted asterisk to ensure it's not bound to an old ip or something. |
| 16:20.44 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 16:22.27 | file | from an Asterisk perspective we bind a UDP port to 0.0.0.0 |
| 16:25.23 | dan_j | Yes. I saw that line in the code. However it doesnt explain why the system is using the original source ip for asterisk, and the correct one for everything else. |
| 16:31.06 | *** join/#asterisk-dev elguero (cf9a22a3@gateway/web/freenode/ip.207.154.34.163) |
| 16:37.45 | file | the other stuff explicitly binds to force it |
| 16:38.14 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-ynzgfaaviuvrbqyh) |
| 16:40.46 | file | coreyfarrell, did it come back to life? |
| 16:41.20 | coreyfarrell | file: not really. the review is posted for master, but the 'git review' client said it failed. now the web interface is giving me a 502 error trying to cherry-pick to 13 |
| 16:41.26 | file | argh |
| 16:42.45 | file | let's see |
| 16:43.03 | coreyfarrell | and about that patch, what do I do to propose that it apply to 13.7.0? it's a simple fix and the leak did not exist in 13.6.0. |
| 16:44.00 | file | you can cherry pick it to that branch |
| 16:44.26 | file | well |
| 16:44.31 | file | are we feeling lucky? |
| 16:45.00 | file | java.lang.OutOfMemoryError: Java heap space |
| 16:45.02 | file | SURPRISE |
| 16:45.10 | coreyfarrell | I'm not even sure I've cherry picked to 13.. I got the mailing list message but it's not showing up on the web ui |
| 16:45.47 | file | I'm going to restart gerrit |
| 16:46.03 | file | er |
| 16:46.04 | file | or not |
| 16:47.50 | file | will it even stop? |
| 16:48.19 | coreyfarrell | it's been stopped for 50 minutes :) |
| 16:48.29 | file | okay, restarted |
| 16:49.05 | file | better, or worse? |
| 16:50.22 | coreyfarrell | functional |
| 16:50.44 | coreyfarrell | thank you |
| 16:51.13 | file | discovering you have access to infrastructure: priceless |
| 17:01.14 | tm1000 | Anyone know a josh elson |
| 17:01.48 | tm1000 | Specifically: http://community.freepbx.org/t/fpbx13-rc-g722-and-1-way-audio-even-to-voicemail/31042/84?u=tm1000 |
| 17:04.11 | dan_j | gtjoseph: regarding your bug comment, do you mean with one private ip and one public ip? |
| 17:04.34 | gtjoseph | nope i can reproduce that one. and i just reproduced yours. |
| 17:04.44 | gtjoseph | i'm investigating now |
| 17:04.56 | dan_j | take a look at this pastebin |
| 17:05.14 | dan_j | http://pastebin.com/XweX03YS |
| 17:05.17 | file | coreyfarrell, angry Jenkins is angry |
| 17:05.50 | dan_j | gtjoseph: Note that the source ip is set to .87, but the kernel at the bottom is using .86 |
| 17:06.11 | file | tm1000, the name does not sound familira |
| 17:06.13 | file | familiar |
| 17:06.31 | coreyfarrell | file: angry at me? |
| 17:06.36 | file | coreyfarrell, angry at reality |
| 17:06.36 | tm1000 | file: but but. He's an asterisk developer. |
| 17:06.43 | tm1000 | ;) |
| 17:06.48 | file | tm1000, I'll look deeper in a second - you are #4 in queue |
| 17:06.59 | file | :D |
| 17:07.17 | gtjoseph | dan_j: got it |
| 17:07.37 | file | coreyfarrell, nope - looks good actually! I thought restarting gerrit may have made it unhappy if it was in a git clone/pull operation but it seems to be trucking along fine |
| 17:09.57 | file | tm1000, he has not filed issues that I can tell |
| 17:10.29 | file | he may do stuff privately for companies/etc |
| 17:14.13 | coreyfarrell | dan_j: from the routing table in the pastebin I think .86 is the expected source address for destination 8.8.8.8 - only the default route matches that destination addr? |
| 17:14.28 | tm1000 | file: no commits either? |
| 17:14.36 | tm1000 | ehhh you really dont have to look at the commits lol |
| 17:14.39 | file | tm1000, that is much harder to tell |
| 17:14.53 | tm1000 | file: yeah dont waste your time :-) |
| 17:15.27 | file | from a SIP perspective the comment is correct - it's asymmetric |
| 17:16.19 | dan_j | coreyfarrell: interesting. so i need to set the source for the default route too. along with all other routes. |
| 17:16.49 | file | although, to their defense, some of the underlying DSP implementations are wacky and don't provide a good mechanism to properly handle the situation |
| 17:17.04 | dan_j | coreyfarrell: seems you're correct. updating the routes and retrying. |
| 17:51.57 | gtjoseph | dan_j: i don't think messing with the routes is going to help. it didn't for me. BUT. I'll have a fix you can test in a few minutes. |
| 18:09.22 | dan_j | gtjoseph: I think i've managed to do it with the routes. Messy but if you have a fix to try, i dont mind giving it a go. |
| 18:09.34 | gtjoseph | cool |
| 18:14.42 | file | gtjoseph, that changes what media_address is for - it's what goes in the SDP, not what you bind to or locally send from |
| 18:15.19 | gtjoseph | yeah, but |
| 18:15.48 | gtjoseph | if you don't set a media address, the sdp is still correct |
| 18:16.21 | file | sure |
| 18:16.33 | gtjoseph | is that a question or a comment? |
| 18:16.46 | file | the sure is an acknowledgement that it is what will happen |
| 18:16.54 | gtjoseph | so what's it point? |
| 18:17.26 | file | allows you to change the value in SDP so that RTP traffic can come over a different non-local IP address if need be |
| 18:17.55 | gtjoseph | bit isn't that what setting the transport fields are for |
| 18:18.04 | file | transport controls SIP signaling |
| 18:18.10 | *** join/#asterisk-dev luckman212 (~luckman21@unaffiliated/luckman212) |
| 18:18.19 | gtjoseph | external_media_address? |
| 18:18.22 | file | with a bit of media for NAT environments |
| 18:18.30 | *** join/#asterisk-dev leedm777 (leedm777@nat/digium/x-bebthltjkikymayx) |
| 18:19.35 | dan_j | I was under the impression that external_media_address was for nat only. |
| 18:19.42 | file | it is |
| 18:19.59 | gtjoseph | but what is NAT from the transport's perspective |
| 18:20.05 | gtjoseph | it doesn't actually care |
| 18:20.29 | file | it has to care about where the target is, so that the signaling is changed |
| 18:20.39 | file | and as a result the SDP is also changed |
| 18:20.47 | gtjoseph | exactly |
| 18:20.57 | file | we could have not lumped it in with the SIP transport itself, but we didn't |
| 18:21.14 | gtjoseph | yeah. we have 2 things doing the same thing |
| 18:21.34 | file | but back to the change - media_address explicitly states what it does in the sample config file and the XML documentation |
| 18:21.35 | gtjoseph | which is really confusing |
| 18:21.41 | dan_j | external_media_address is the external public ip if the asterisk box is behind a router. Thats not the same as we are discussing here. |
| 18:21.54 | gtjoseph | nope |
| 18:22.05 | gtjoseph | asterisk doesn't know whether it's behing a router or not |
| 18:22.34 | gtjoseph | there's no 'if'. |
| 18:22.40 | gtjoseph | if you set it, it's used |
| 18:22.54 | file | external_media_address uses the local_net settings to know |
| 18:23.02 | file | media_address on an endpoint does not |
| 18:23.22 | file | both alter the SDP, neither change the bound address |
| 18:24.25 | gtjoseph | so how should i control the bound address, yet another option on endpoint (or transport, or both) |
| 18:25.01 | file | doing it based on transport is extremely hard, because on outgoing you will have to ask PJSIP and it will guess based on what it knows at the time |
| 18:25.04 | gtjoseph | media_bind_address in endpoint? |
| 18:25.12 | file | an explicit option on endpoint will guarantee the behavior |
| 18:25.23 | file | both is probably what an end-user would like |
| 18:25.52 | gtjoseph | nope. i'm not compounding the problem with more dups |
| 18:26.33 | file | if the guess is right enough, noone would notice |
| 18:26.36 | dan_j | gtjoseph: pjsip transport binds it's signaling using bind=. But when I ask discussed this with file, PJSIP doesn't tell asterisk which transport is being used, which is probably why asterisk was never bound to an IP for audio. |
| 18:26.52 | dan_j | ask =last |
| 18:27.04 | gtjoseph | true but we HAD media_address on endpoint |
| 18:27.24 | gtjoseph | it's just that it only changes the SDP |
| 18:27.54 | dan_j | if that's set, would that mean that an endpoint can't be brought onto the internal network without changing media_address? |
| 18:27.59 | gtjoseph | actually looking at my traces, some of the sdp is changed and others not |
| 18:31.01 | gtjoseph | If I only bind the transport to a specific address, the SIP header is correct and the SDP/Connection Information is correct but the Owner/Creator is wrong |
| 18:31.20 | file | in practice the owner/creator doesn't matter that much |
| 18:31.27 | gtjoseph | So the remote end sends rtp to the correct address fur the local system sends rtp from the incorrect one. |
| 18:36.14 | gtjoseph | ok file: media_bind_address in endpoint? |
| 18:36.20 | file | that's fine |
| 18:36.24 | gtjoseph | i didn't want to touch alembic |
| 18:36.28 | gtjoseph | but |
| 18:36.37 | file | noone wants to touch alembic |
| 18:37.09 | gtjoseph | dan_j: try the patch in https://gerrit.asterisk.org/#/c/1928/ |
| 18:37.23 | gtjoseph | it's not final but i just want to know if this solves the issue for you |
| 18:39.01 | dan_j | Ok. Just removing the routes I've added. |
| 18:40.43 | dan_j | Am I correct in saying that media_bind_address in endpoint means that an endpoint cannot move from a transport with a public ip, to a transport with a private ip? |
| 18:40.43 | gtjoseph | file: CRAP. adding the bind_address to ast_sip_endpoint_media_configuration would break ABI. |
| 18:40.58 | gtjoseph | true |
| 18:41.09 | gtjoseph | problem? |
| 18:41.42 | dan_j | Not really for me. I use the private IP for VPN connections. But I wonder if someone else will report this as an issue. |
| 18:42.06 | gtjoseph | that's ok. |
| 18:43.03 | gtjoseph | file: how about a flag... bind_rtp_to_media_address? |
| 18:43.17 | file | uno sec, I just accidentally closed every terminal |
| 18:44.23 | file | TBH - if you only want to go one of the two ways, asking PJSIP would not require an alembic change and it should just work |
| 18:44.37 | file | if you don't want to do that then a flag is fine |
| 18:45.20 | gtjoseph | what do i ask it for? the bind address or the external_media_address? I'm thinking bind |
| 18:45.38 | file | SO.... |
| 18:45.41 | gtjoseph | since that's what i want to do. |
| 18:45.57 | file | unlike chan_sip which knows what transport it will use all the time... |
| 18:46.05 | file | PJSIP determines it when a message is sent |
| 18:46.22 | gtjoseph | so at the point we bind rtp, we don't know the transport |
| 18:46.29 | file | correct |
| 18:46.34 | file | that's the crux of the problem |
| 18:47.14 | gtjoseph | which is an issue since there are times when the endpoint can't specify an explicit one. |
| 18:47.19 | gtjoseph | TLS |
| 18:47.29 | file | right |
| 18:47.57 | gtjoseph | so back to the flag |
| 18:52.44 | dan_j | I'm currently on 13.5.0. Do I need to download the trunk version or is 13.6.0 ok to try out this patch? |
| 18:52.55 | gtjoseph | 13.6.0 |
| 18:53.23 | gtjoseph | it's probably apply just finr to 1`3.5.0 though |
| 18:55.44 | dan_j | I made another change to the source code, to add the regcontext support to pjsip. But didnt know how to read from the config files so it's hard coded in my source file. |
| 18:56.13 | dan_j | This is the error i get when compiling with 13.5.0 |
| 18:56.14 | dan_j | http://pastebin.com/zgWX90BY |
| 18:57.44 | gtjoseph | did you do a git review to pull the patch or did you just dowenload it and apply? |
| 18:57.50 | gtjoseph | download and apply should work |
| 18:58.01 | gtjoseph | i.e patch -p1 < ... |
| 18:58.35 | *** join/#asterisk-dev goddva (~goddva@2a02:fe0:c510:e300:ae9e:17ff:fe91:a8c9) |
| 18:59.01 | dan_j | just downloaded the file and placed it in the correct location |
| 19:02.19 | *** join/#asterisk-dev ipengineer (~zconkle@static-71-252-134-63.dllstx.fios.verizon.net) |
| 19:04.32 | dan_j | nuts. forgotten how to compile with pjsip |
| 19:22.33 | dan_j | compiling now |
| 19:23.32 | gtjoseph | file: we can just use the config.ini.sample as is for alembic correct? |
| 19:23.40 | file | yes |
| 19:23.50 | gtjoseph | cool |
| 19:28.00 | dan_j | gtjoseph: same problem |
| 19:28.02 | dan_j | http://pastebin.com/56hLteAY |
| 19:28.29 | gtjoseph | not sure what's going on with the compile |
| 19:28.36 | gtjoseph | i didn't touch that code. |
| 19:28.38 | dan_j | i need to git the whole thing to get an up to date version? |
| 19:28.53 | gtjoseph | yeah, that's probably best |
| 19:29.46 | dan_j | "git clone http://gerrit.asterisk.org/asterisk asterisk" includes your patch or is that separately patched? |
| 19:30.07 | dan_j | sorry. not familiar with git. |
| 19:32.28 | *** join/#asterisk-dev jameswf_ (uid27319@gateway/web/irccloud.com/x-tzwceqyyenoliqsu) |
| 19:33.01 | gtjoseph | that clone gets everything |
| 19:33.31 | gtjoseph | then you can do a "git review -d 1928" |
| 19:33.53 | dan_j | hmm |
| 19:33.54 | dan_j | http://pastebin.com/S0v0fPju |
| 19:35.20 | *** join/#asterisk-dev GameGamer43_ (sid5533@gateway/web/irccloud.com/x-eraybkuktrhnqxca) |
| 19:35.49 | gtjoseph | you're high maintenance aren't you. :) |
| 19:36.00 | gtjoseph | are you registered on gerrit? |
| 19:36.13 | dan_j | :) not usually |
| 19:36.14 | dan_j | never used it. |
| 19:36.34 | *** join/#asterisk-dev opticron (~opticron@75.76.201.152) |
| 19:36.34 | *** mode/#asterisk-dev [+o opticron] by ChanServ |
| 19:36.35 | gtjoseph | give me a minute, i forgot what the commands are for a new download. |
| 19:37.00 | dan_j | I'm just going off what it says on www.asterisk.org |
| 19:37.30 | gtjoseph | i don't trust that site :) |
| 19:41.08 | gtjoseph | dan_j: "git clone -b 13 http://gerrit.asterisk.org/asterisk asterisk-13" worked for me |
| 19:42.17 | dan_j | Looks like the password reset feature at 'yooden.digium.api:8095' is down |
| 19:43.58 | dan_j | Looks like the password reset feature at https://openid.asterisk.org/ is down |
| 19:44.52 | *** join/#asterisk-dev ipengineer (~zconkle@static-71-252-134-63.dllstx.fios.verizon.net) |
| 19:46.05 | *** join/#asterisk-dev Jacke (jacke@valhalla.bofh.org) |
| 19:46.39 | file | what's the username? |
| 19:47.14 | dan_j | journo |
| 19:47.23 | dan_j | i get the email but the reset link doesnt work. |
| 19:47.29 | file | you have two accounts |
| 19:47.30 | dan_j | tried it from my mobile too |
| 19:47.38 | file | (fun fact) |
| 19:47.55 | dan_j | how did that happen? |
| 19:48.02 | file | dunno! the other one is 'kesher' |
| 19:48.48 | file | if you replace yooden.digium.api:8095 in the link does it work? |
| 19:49.02 | file | if not I can take care of it |
| 19:49.35 | dan_j | replace it with what? it's already that |
| 19:49.40 | gtjoseph | is yooden.digium.api even a valid public hostname? |
| 19:49.56 | gtjoseph | er, no |
| 19:49.57 | file | replace it with openid.asterisk.org |
| 19:50.01 | dan_j | api must be one of the new TLDs |
| 19:50.08 | dan_j | or not |
| 19:50.27 | dan_j | The requested URL /crowd/console/resetpassword.action was not found on this server. |
| 19:50.35 | file | digium.api is for infrastructure located at API Digital in Huntsville, it is where much of the Asterisk stuff is held |
| 19:51.51 | gtjoseph | dan_j: i'm heading out to lunch for about an hour. i'll check with you when i get back |
| 19:52.11 | file | dan_j, I emailed you a temporary password - please please please change it upon login |
| 19:54.40 | dan_j | Which username was that for? |
| 19:54.46 | file | journo |
| 19:55.09 | dan_j | Doesnt appear to work at openid |
| 19:55.09 | file | once I find the right tab I'm going to see if I can grab the creation date for the other |
| 19:55.42 | dan_j | journo is the one i use most. kesher might have been related to a g729 purchase or something? |
| 19:55.57 | file | public infrastructure isn't connected to Digium for authentication |
| 19:56.31 | file | oh - do you have a license agreement signed? Gerrit won't allow access otherwise |
| 19:56.46 | dan_j | completed one about 5 minutes ago |
| 19:56.55 | file | it's not instantly done, legal has to approve |
| 19:57.47 | dan_j | unfortuately, i'm unable to download via http. will have to try the patch once the account gets approved and i can download via ssh |
| 19:58.05 | file | we also mirror to github |
| 19:58.23 | file | https://github.com/asterisk/asterisk |
| 19:59.09 | dan_j | will the mirror already have gtjoseph's patch? |
| 19:59.48 | dan_j | doesnt look like it |
| 20:00.11 | file | yeah, I don't actually know what it would try to do in that case... |
| 20:01.57 | dan_j | I'll get it from github and then copy in the patch by hand |
| 20:06.34 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-179-40-172.cust.bezeqint.net) |
| 20:06.35 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 20:19.06 | file | just a heads-up for y'all, there's an Asterisk blog at https://www.asterisk-blog.com/ now which is more focused on developer/project stuff |
| 20:27.03 | *** join/#asterisk-dev malcolmd (malcolmd@pdpc/sponsor/digium/malcolmd) |
| 20:27.03 | *** mode/#asterisk-dev [+o malcolmd] by ChanServ |
| 20:48.02 | dan_j | gtjoseph: Patch didnt seem to help |
| 20:48.17 | dan_j | Unless I have to do some config change too? |
| 20:48.21 | gtjoseph | the latest version of the patch needs a new config |
| 20:48.40 | gtjoseph | set bind_rtp_to_media_address to yes in your endpoint |
| 20:49.06 | gtjoseph | and set media_address accordingly |
| 20:50.32 | dan_j | Ok. Think im using the first version of the patch. will get the latest |
| 20:50.50 | gtjoseph | i uploaded the new patch just before i went to lunch |
| 20:52.51 | dan_j | no probs. didnt see it. and been cooking dinner for the last 20 minutes |
| 21:00.11 | dan_j | gtjoseph: I got the first patch to install, but the 3rd gives me this |
| 21:00.11 | dan_j | http://pastebin.com/KFmsVthU |
| 21:00.48 | gtjoseph | looks like you didn't apply the entire patch |
| 21:01.03 | gtjoseph | there are more files |
| 21:01.15 | gtjoseph | hold on, i'll pb the entire patch |
| 21:02.18 | dan_j | I couldnt git from asterisk. I had to git from github which i assume doesnt have your patch yet |
| 21:02.47 | gtjoseph | correct |
| 21:03.17 | gtjoseph | https://pastebin.com/X4Kp7tVG |
| 21:03.33 | gtjoseph | while in the asterisk directory... patch -p1 < patch |
| 21:03.45 | gtjoseph | might want to do --dry-run first |
| 21:14.12 | dan_j | bind_rtp_to_media_address = true or yes? |
| 21:18.36 | dan_j | Ok. Yes. That patch seems to work perfectly. |
| 21:18.45 | gtjoseph | yes? |
| 21:18.55 | gtjoseph | cool |
| 21:19.02 | dan_j | Yes. looks like it. coming from the right ip now. |
| 21:19.18 | dan_j | Thats great! Much better than messing with the routes. |
| 21:19.38 | dan_j | Thanks for your help! When will that appear in the main trunk? |
| 21:19.50 | *** join/#asterisk-dev Tim_Toady (~fuzzy@snf-33276.vm.okeanos.grnet.gr) |
| 21:21.18 | gtjoseph | it should go into 13.8 |
| 21:21.29 | gtjoseph | 13.7 is pretty much baked |
| 21:21.51 | dan_j | Out of interest, is there a helper class that deals with reading config files? I want to see if I can add regcontext to pjsip |
| 21:23.18 | gtjoseph | yep. to add a new option to endpoint, look at the bottom of res/res_pjsip/pjsip_configuration.c |
| 21:24.01 | gtjoseph | of course, you still have to modify the appropriate structures and logic to do something with it. |
| 21:26.15 | dan_j | It wouldnt be on the endpoint but on the whole pjsip |
| 21:26.56 | dan_j | any endpoint registering would cause a NoOp extension to be added. Identical to what regcontext does to chan_sip |
| 21:27.15 | gtjoseph | ah, that would go in res/res_pjsip/config_global.c |
| 21:35.35 | dan_j | Thanks for everything. It's been really helpful! Been trying to move to pjsip for a few months now. |
| 21:37.20 | gtjoseph | dan_j: no problem! |
| 21:39.03 | gtjoseph | file: do you know of anything in the testsuite that can actually examine the source address of a packet? |
| 21:40.51 | gtjoseph | oooh... hep |
| 21:47.20 | file | can also use the UDP receiver in twisted |
| 21:47.51 | gtjoseph | ah. didn't think about twisted |
| 21:48.24 | gtjoseph | hep'd be easier though since there are already tests |
| 21:51.31 | *** join/#asterisk-dev opticron (~opticron@75.76.201.152) |
| 21:51.31 | *** join/#asterisk-dev jrrose (jon@nat/digium/x-atzgjfajtnrtykad) |
| 21:51.31 | *** mode/#asterisk-dev [+o opticron] by sendak.freenode.net |
| 21:54.00 | *** join/#asterisk-dev [TK]D-Fender (~joe@64.235.216.2) |
| 22:01.49 | gtjoseph | sorry file: one more thing to bug you with... the tests in tests/hep are all pjsip related so i'm not sure whether they're testing hep or pjsip. where should a test that uses hep to test pjsip go? |
| 22:02.16 | file | um, tests/channels/pjsip |
| 22:02.30 | gtjoseph | yeah, that's what i thought |
| 22:49.10 | gtjoseph | Ha! I just played tt-monkeys on my computer and my cat freaked out. |
| 22:50.08 | dan_j | gtjoseph: just thinking about the patch and it might be easier for me to use the routing option to fix things so that endpoints can move between transports. what do you think? |
| 22:50.44 | gtjoseph | well, you can. How can I make it work for you though? |
| 22:51.38 | dan_j | from what file was saying, i'm not sure you can, since pjsip doesnt advertise the transport that is being used. |
| 22:51.57 | gtjoseph | well, let me dig a little more |
| 22:56.24 | dan_j | ok, thanks. I tried last week but didnt know enough about pjsip to work out how to query for a transport. |
| 22:56.51 | gtjoseph | well, we can't actually query for it at this point but maybe there's another way |
| 23:00.43 | dan_j | ok. you know more than me. i've got to sign off for the night. getting late here. thanks for your help so far. |
| 23:47.56 | gtjoseph | file: you still around? |
| 23:54.35 | file | I was not, but now I am |
| 23:55.03 | DivideBy0 | does anyone technical go to asterisk world in 2 weeks? |
| 23:55.08 | gtjoseph | :) What would yo think of us calling pjsip_endpt_acquire_transport ourselves early on if there was no transport specified on the endpoint |
| 23:55.29 | file | DivideBy0, only in spirit |
| 23:55.48 | gtjoseph | then setting the transport so pjsip doesn't go through the selection process again |
| 23:55.52 | file | gtjoseph, you can try... |
| 23:55.56 | DivideBy0 | bah, ok thanks |
| 23:55.56 | gtjoseph | but? |
| 23:56.09 | file | SRV and NAPTR would get you in master |
| 23:56.44 | gtjoseph | It would have to be after that but before the rtp session is created |
| 23:57.00 | file | resolution is async |
| 23:57.46 | gtjoseph | who gets the results of the lookup, pjsip or us? |
| 23:58.09 | file | we get a callback and populate a result structure, and hand it back to PJSIP |
| 23:58.50 | gtjoseph | we get the callback from pjsip? |
| 23:58.56 | gtjoseph | oh i get it |
| 23:58.57 | file | we get the callback from our DNS layer |
| 23:59.02 | gtjoseph | it's asking us for resolution |
| 23:59.06 | file | yes |
| 23:59.24 | file | you may be able to insert yourself into the stack after that |
| 23:59.56 | file | it shouldddd be possible to take that and use a function PJSIP provides to determine what interface and IP address will be used for the traffic (the NAT does it) |