| 00:18.53 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-ezvxpewmfucfjmgt) |
| 00:28.02 | *** join/#asterisk-dev hitech95 (~hitech95@151.49.88.217) |
| 00:29.59 | hitech95 | Hi guys I have a question about the new pjsip, i have seen a strange behavior. Looks like that if I specify a outbound proxy on the registrar the Authorization is not sent. Is this correct or what? |
| 00:46.01 | *** join/#asterisk-dev snuff-work (~snuff-wor@210.9.148.102) |
| 00:46.02 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 00:47.21 | gtjoseph | hitech95: is the proxy returning a 401 or 407 requesting it? |
| 00:50.02 | hitech95 | gtjoseph, look at this: http://pastebin.freepbx.org/view/3ac95497#L14 in the second one I have added the outbound_proxy parameter on the configuration. in the first one no. |
| 00:51.24 | hitech95 | I know that the setup is done via freepbx but I double checked the config of asterisk. I would try to understand if is a problem of freepbx or asterisk. |
| 00:51.54 | gtjoseph | we need to see the entire exchange. on the first one, we must have sent a register without the authorization header to which the remote end sent a 401 so we sent the register again with the Authorization header. |
| 00:52.14 | gtjoseph | what was the response from the proxy on the second one? |
| 00:53.25 | gtjoseph | also, are you actually using freepbx or just their pastebin service? |
| 00:53.42 | gtjoseph | oh nevermind. i just saw the user agent. |
| 00:53.46 | gtjoseph | FPBX |
| 00:57.27 | hitech95 | No, the 401 is not sent by the proxy... (I have some trouble on debugging the register process). But I have a log. Give me a sec. This is the generated configuration: http://pastebin.freepbx.org/view/bbba8ccb |
| 00:58.09 | hitech95 | Right now I have solved with a custom dns rule so sip-infostrada.it is resolved as proxy.libero.it and all works but this is a hack! |
| 01:00.28 | gtjoseph | so it looks like you named all your objects 0415649102 which is fine but when you reference them you use 39XXXXXXX102. Was that just a cut and paste error? |
| 01:00.58 | gtjoseph | For instance, in your register, outbound_auth should be 0415649102 not 39XXXXXXX102 |
| 01:01.40 | gtjoseph | and in the endpoint, the aor is specified as 0415649102-pjsip bit it's name is actually 0415649102 |
| 01:02.12 | gtjoseph | so the first thing would be to make sure all the references are correct. |
| 01:03.06 | hitech95 | "that just a cut and paste error?" yup. it was a old cut & paste. now all is marked as "0415649102" |
| 01:03.36 | hitech95 | This is after the register send (forced by cli) http://pastebin.freepbx.org/view/521d4c2a |
| 01:04.08 | gtjoseph | ok so in this one, the response is 404 not found |
| 01:04.16 | gtjoseph | we can't do much with that |
| 01:04.59 | gtjoseph | i'd check that the URIs specified in the register are actually correct |
| 01:06.31 | hitech95 | but if you see there is a missing parameter in the request.... is that normal? |
| 01:07.14 | gtjoseph | what do you think is missing, the Authorization header? |
| 01:08.37 | hitech95 | yep |
| 01:08.51 | hitech95 | is I dont specify the proxy that is sent |
| 01:09.13 | hitech95 | this is the full config: http://pastebin.freepbx.org/view/cfcb1f85 |
| 01:09.23 | gtjoseph | We can only send an Authorization header if we get a 401 with a WWW-Authenticate header or a 407 with a WWW-Proxy-Authenticate header back from the remote end. |
| 01:09.55 | hitech95 | gtjoseph, oh ok. So the problem is the proxy that send a 404. |
| 01:10.01 | gtjoseph | exactly |
| 01:10.16 | hitech95 | sho the proxy is not standard. |
| 01:10.19 | hitech95 | *so |
| 01:10.57 | gtjoseph | well it is a little weird. usually the server will respond with the 401 or 407 and make you pass authentication before giving you the 404. |
| 01:11.59 | gtjoseph | try specifying "contact_user" in the registration and see if that makes a difference. |
| 01:12.27 | hitech95 | Thx for the help. I'm pretty new on SIP/VOIP. The SIP protocol seems to me so messed up, it's me. surely there is a reason for everything, but is just me. |
| 01:12.53 | gtjoseph | no it's not you. it IS messed up. :) It's just that we're used to the mess. |
| 01:14.25 | hitech95 | The proxy is a bit odd. the domain in fake. it can be resolved only by the isp dns server. The voip parameter are sent by the tr069 server to my isp router that I replaced and hacked to extract the password. So maybe it is to stop standard software... |
| 01:14.55 | gtjoseph | wow, could be. |
| 01:15.20 | hitech95 | let's see what does the contact_user do, let me check the config to see how to configure that... |
| 01:18.28 | *** join/#asterisk-dev infobot (ibot@rikers.org) |
| 01:18.28 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Check out our blog! blogs.asterisk.org -=- Follow on Twitter at @AsteriskDev |
| 01:19.24 | hitech95 | my isp is hell. They want you tu use their crappy stuff |
| 01:20.02 | hitech95 | asterisk is restarting the first gen Pi is quite slow :( |
| 01:24.17 | hitech95 | nop, nothing change: http://pastebin.freepbx.org/view/e24e0dec |
| 01:24.33 | hitech95 | at least now I know that the problem is that odd proxy |
| 01:27.16 | hitech95 | On my second trunk the ISP don't send the BYE when the caller hangup so the call is not ended. I'm not luky... |
| 01:27.27 | gtjoseph | :) |
| 01:28.30 | hitech95 | Anyway. Thanks again for the help. Here is 2:20 AM as you can see from logs. Now is time to sleep. Keep on the good work, asterisk looks amazing! |
| 01:29.47 | gtjoseph | Good night! |
| 02:52.38 | *** join/#asterisk-dev leedm777 (~textual@2604:2d80:8404:ffae:5539:907d:cc27:8595) |
| 02:52.38 | *** mode/#asterisk-dev [+o leedm777] by ChanServ |
| 06:21.45 | *** join/#asterisk-dev coreyfarrell (~coreyfarr@24-177-86-98.dhcp.nwtn.ct.charter.com) |
| 06:21.45 | *** mode/#asterisk-dev [+o coreyfarrell] by ChanServ |
| 07:13.22 | snuff-work | mm.. if i try to do like make -j3 |
| 07:13.32 | snuff-work | doesn't seem to do any in parallel atm |
| 07:13.45 | snuff-work | (git head of master) |
| 07:30.42 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 07:30.42 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 07:45.05 | *** join/#asterisk-dev tsearle (~tsearle@37.19.10.33) |
| 07:45.05 | *** mode/#asterisk-dev [+o tsearle] by ChanServ |
| 07:52.27 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:2c34:8fd6:525b:36d0) |
| 07:52.46 | *** join/#asterisk-dev snuff-work (~snuff-wor@210.9.148.102) |
| 07:52.46 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 08:12.00 | *** join/#asterisk-dev pchero_work (~pchero@109.70.54.56) |
| 08:13.23 | snuff-work | nm my noise.. this other * box has only 1 core assigned =) |
| 08:13.41 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:56ee:75ff:fed4:f268) |
| 08:19.01 | snuff-work | mm maybe i am right.. anyway.. if any1 else wants to confirm for me.. that doing a make -j3 doesn't actually build anything 'simultaneously' anymore |
| 08:24.13 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:56ee:75ff:fed4:f268) |
| 08:32.32 | *** join/#asterisk-dev mvanbaak (~mvanbaak@asterisk/contributor-and-bug-marshal/mvanbaak) |
| 08:32.32 | *** mode/#asterisk-dev [+o mvanbaak] by ChanServ |
| 08:45.24 | snuff-work | mm.. seems someone put the NOTPARALLEL in the main make file.. which seems to have inhertied all the way through the stack |
| 08:48.47 | snuff-work | which according to blame was coreyfarrell =) |
| 08:53.29 | snuff-work | according to the docs.. it shouldn't go down the tree.. so that's interesting.. |
| 08:58.29 | snuff-work | wonders off... |
| 09:20.08 | *** join/#asterisk-dev skrusty (~skrusty@88.150.145.104) |
| 09:20.08 | *** mode/#asterisk-dev [+o skrusty] by ChanServ |
| 09:31.39 | *** join/#asterisk-dev jkroon (~jkroon@uls-154-73-35-201.wall.uls.co.za) |
| 10:25.04 | *** join/#asterisk-dev hitech95 (~hitech95@151.49.88.217) |
| 10:42.12 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 13:12.05 | *** join/#asterisk-dev Dovid (~dovid@ool-45738ae3.dyn.optonline.net) |
| 13:58.40 | *** join/#asterisk-dev leedm777 (textual@nat/digium/x-szqpnwqxtunaqvyz) |
| 13:58.41 | *** mode/#asterisk-dev [+o leedm777] by ChanServ |
| 15:03.07 | *** join/#asterisk-dev bford (d8cff501@gateway/web/freenode/ip.216.207.245.1) |
| 15:03.07 | *** mode/#asterisk-dev [+o bford] by ChanServ |
| 15:09.06 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-yzpttkrnxhmiebgl) |
| 15:09.06 | *** mode/#asterisk-dev [+o kharwell] by ChanServ |
| 15:28.06 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-chazipzyhkqzsuzq) |
| 15:28.06 | *** mode/#asterisk-dev [+o rmudgett] by ChanServ |
| 16:18.34 | *** join/#asterisk-dev Worldexe (~Worldexe@95-107-33-134.dsl.orel.ru) |
| 17:34.54 | *** join/#asterisk-dev jkroon (~jkroon@165.16.204.171) |
| 18:35.09 | coreyfarrell | snuff-work: yeah the parallel build was a problem in the root makefile, I can confirm in current versions of GNU make it does nothing to submake's. |
| 19:15.43 | coreyfarrell | rmudgett: looks like I'm getting some push-back about giving jansson the ability to call a function pointer which emulates __sync_add_and_fetch. :( |
| 19:22.19 | coreyfarrell | does anyone know the oldest version of GCC which can compile asterisk? I know the README says 3.0 but I doubt anyone tests this so who knows if that's still true. what other compilers are considered "core supported". |
| 19:24.16 | gtjoseph | well, gcc is the only "core supported" compiler. clang is best effort. |
| 19:33.04 | coreyfarrell | gtjoseph: obviously I'd need to post on the mailing list but if were to make master require gcc 4.1.2 or another compiler with __sync or __atomic built-ins, doesn't sound like this would be a big issue. |
| 19:33.29 | gtjoseph | i would think that's fair in 2018. |
| 19:34.36 | coreyfarrell | yeah 4.1.2 was released 11 years ago (according to the timestamp on ftp.gnu.org) |
| 19:35.36 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01cb0583d31d00b0d933d5e300f28f.ipv6.abo.wanadoo.fr) |
| 20:40.21 | *** join/#asterisk-dev danielyk (~danielyk@p4FC7D6E7.dip0.t-ipconnect.de) |
| 21:01.06 | rmudgett | gtjoseph: Does the current asterisk keep alive work with UDP to keep NAT pinholes open? |
| 21:03.21 | gtjoseph | nope |
| 21:06.11 | file | strictly connection oriented |
| 22:03.59 | *** join/#asterisk-dev gtjoseph (~gtjoseph@unaffiliated/gtj) |
| 22:03.59 | *** mode/#asterisk-dev [+o gtjoseph] by ChanServ |
| 22:06.21 | coreyfarrell | gtjoseph: what was the use case for spinlock.h? |
| 22:12.56 | *** join/#asterisk-dev tzafrir (~tzafrir@62-90-199-247.barak.net.il) |
| 22:12.57 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 22:26.55 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01cb0583d31d00b0d933d5e300f28f.ipv6.abo.wanadoo.fr) |
| 22:27.44 | kharwell | For those interested. I kinda changed up what I was going to do with the testsuite branching stuff. Turns out there are not a ton of min/max differences between the branches to make it worthwhile to push a review and wait. |
| 22:28.05 | kharwell | Some stuff in 14 was weird too as some tests should be in 13 and 15, but not 14 branch. |
| 22:28.16 | *** join/#asterisk-dev hitech95 (~hitech95@151.49.88.217) |
| 22:28.48 | kharwell | Also this makes it so everything was based off master originally, so if for some reason we need to revert something on a branch or go back this will make it easier |
| 22:57.52 | snuff-work | coreyfarrell: np.. i'm just trying to figure out why my build (granted it has 'dont optimize' turned on) does not seem to build files in parallel inside the directories it goes to |
| 22:59.42 | coreyfarrell | snuff-work: maybe see if disabling COMPILE_DOUBLE makes a difference? |
| 23:00.59 | coreyfarrell | kharwell: I think you must have those reviews as drafts? I see them in the mailing list but not in gerrit. |
| 23:01.50 | kharwell | I deleted them. I'm going to try something different to minimize the reviews for the yaml file changes themselves. |
| 23:02.21 | kharwell | I have to go through every test and remove the max/min version and also remove sections for those tests that have modules max/min version stuff |
| 23:02.46 | kharwell | if I do that all in the new master and then base the new branches off that it will only make one review for that giant change |
| 23:03.49 | coreyfarrell | not sure if this was just weirdness in the mailing list post but please don't call the new branch 'new master' (I'm complaining about the space if it was really there) |
| 23:04.17 | kharwell | Unfortunately I had written a script to remove delete the unneeded files from the branches based on the min/max. I'll have to find a way to save that data off prior to removing the version info from the yaml files |
| 23:04.45 | kharwell | I had called it new_master (no space) I can just call it newmaster so no confusion |
| 23:04.57 | snuff-work | coreyfarrell: false alarm.. i do see multiple cc's going in background |
| 23:05.04 | snuff-work | sorry for noise =) |
| 23:05.25 | coreyfarrell | snuff-work: no problem |
| 23:06.14 | coreyfarrell | kharwell: new_master is fine.. weird that gerrit would substitute like that for a branch name. |
| 23:06.26 | snuff-work | btw.. when was ast_module_info changed.. |
| 23:06.45 | snuff-work | i'm just playing with chan-sccpb and i had to remove .nonoptreq = "chan_local" from the call |
| 23:07.55 | coreyfarrell | snuff-work: in the past week or so. FYI .nonoptreq never did anything. for master you want to replace it with .requires or .optional_modules. |
| 23:08.55 | snuff-work | .nonoptreq = not required option.. or non optional required ? |
| 23:09.04 | snuff-work | aka .required i guess? |
| 23:09.28 | coreyfarrell | snuff-work: actually unless you have a custom chan_local.so you should just delete that.. chan_local is no longer a module in any active version. |
| 23:09.50 | snuff-work | k |
| 23:10.48 | coreyfarrell | how to update .nonoptreq for modules that do still exist really depends on how the two modules interact, but in this case putting .requires = "chan_local" would make the updated module loader refuse to load that module. |
| 23:18.21 | snuff-work | ok.. so it was a change after ast 15 branched? |
| 23:18.33 | snuff-work | erm before i mean |
| 23:22.59 | jsmith | Anybody know how to select which AOR gets the call when Visual Voicemail on a Digium phone tries to play the voicemail message? |
| 23:23.04 | coreyfarrell | snuff-work: it will be new to Asterisk 16. |
| 23:23.28 | jsmith | If I have multiple AORs attached to a particular endpoint, it seems to be trying to call one of my other AORs to play the message :-/ |
| 23:25.28 | snuff-work | np |
| 23:25.51 | snuff-work | i would have checked myself first if i could drive git properly myself =) |
| 23:26.56 | *** join/#asterisk-dev leedm777 (~textual@2607:fb90:998:8b03:4db5:48b3:b944:1467) |
| 23:26.56 | *** mode/#asterisk-dev [+o leedm777] by ChanServ |
| 23:30.14 | snuff-work | mm 1 way audio is evil.. works both ways if other party calls.. i like when both sides have same issue.. |
| 23:30.24 | coreyfarrell | snuff-work: apparently gerrit lets you search for reviews that touched a file.. https://gerrit.asterisk.org/#/q/file:main/loader.c |
| 23:31.03 | snuff-work | nice |
| 23:32.12 | snuff-work | oh i was playing with my asterisk earlier on.. if i didn't have a hep.conf/phoneprov.conf asterisk wouldn't start |
| 23:32.30 | snuff-work | rather than just not load those modules |
| 23:33.14 | coreyfarrell | snuff-work: that sounds like an old revision, https://gerrit.asterisk.org/5452 should have fixed that. |
| 23:33.58 | snuff-work | mm.. well it was broken as of yesterday if i didn't have those files in my /etc/asterisk dir |
| 23:34.12 | coreyfarrell | what version/revision? |
| 23:35.05 | snuff-work | git/master |
| 23:35.10 | snuff-work | [Jan 25 10:34:59] ERROR[500]: config_options.c:706 aco_process_config: Unable to load config file 'hep.conf' |
| 23:35.11 | snuff-work | [Jan 25 10:34:59] ERROR[500]: loader.c:1640 start_resource_list: Failed to load res_hep_rtcp due to unfilled dependencies. |
| 23:35.11 | snuff-work | [Jan 25 10:34:59] ERROR[500]: asterisk.c:4336 check_init: Module initialization failed. ASTERISK EXITING! |
| 23:35.11 | snuff-work | âDestroying musiconhold processes |
| 23:36.40 | coreyfarrell | ah right now. could you post a ticket to jira? I'm about to eat I don't want to forget to look into this. |
| 23:36.50 | snuff-work | np |
| 23:39.56 | coreyfarrell | snuff-work: please attach your modules.conf to the ticket if it is changed from 'make samples'. |
| 23:40.32 | snuff-work | it is very very basic modules.conf in terms of what it specs to load/no-load |
| 23:40.53 | snuff-work | but i think its close to the samples one |
| 23:41.00 | snuff-work | anyway i put into case |
| 23:51.50 | rmudgett | snuff-work: On master, if you don't have a hep.conf file then res_hep.so fails to load which is a dependency for res_hep_rtcp.so. |
| 23:52.25 | snuff-work | yes.. i get that.. but i dont say i care about res_hep_rtcp |
| 23:52.44 | snuff-work | in my modules load |
| 23:52.58 | snuff-work | i just have the standard 'autoload yes' or whatever |
| 23:53.37 | rmudgett | With autoload it will attempt to load all modules even if you don't care about them. |
| 23:54.17 | snuff-work | ok.. but if a module fails and we dont use it why kill asterisk startup ? |
| 23:54.22 | coreyfarrell | res_hep_rtcp failing should not prevent asterisk from starting (unless config says it's required). should accept that res_hep_rtcp cannot load and move on. |
| 23:56.00 | snuff-work | where in the config would i put that the module is req? |
| 23:56.22 | rmudgett | require=chan_pjsip.so |
| 23:56.39 | snuff-work | so pjsip requires that module always? |
| 23:56.44 | rmudgett | Just like load=res_monitor.so |
| 23:57.01 | rmudgett | was an example. |
| 23:57.52 | snuff-work | i understand i could put it in my module.conf as a load requirement.. in which case i understand it bailing out |
| 23:58.18 | rmudgett | It does sound like a bug in the new module loader code. |
| 23:58.38 | snuff-work | happens for 'phoneprov.conf and hep.conf' |
| 23:59.03 | snuff-work | i do make use of pjsip as my only SIP transport |
| 23:59.57 | snuff-work | have to use sccpb for silly cisco 7937 phones that are polycoms in disguise but wont let a polycom sip firmware on them =) |