| 00:15.50 | *** join/#asterisk-dev zopsi (~zopsi@2a01:4f8:201:94e5::2) |
| 01:58.55 | *** join/#asterisk-dev Juggie (~Juggie@unaffiliated/juggie) |
| 03:07.18 | *** join/#asterisk-dev nix8n82 (~AndChat56@2600:100e:b037:701d:1e5:666c:5dd5:2a73) |
| 04:26.31 | *** join/#asterisk-dev snuff-work (~blatus@203.87.61.22) |
| 04:26.31 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 05:01.45 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 05:09.23 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 05:12.08 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 05:32.08 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 06:31.45 | snuff-work | anyone know why sysctl might not be detected by configure ? |
| 06:44.17 | *** join/#asterisk-dev CELYA_ (~Thunderbi@LStLambert-657-1-90-73.w80-13.abo.wanadoo.fr) |
| 07:09.34 | wdoekes | snuff-work: manpage says "Glibc does not provide a wrapper for this system call; call it using syscall(2). Or rather... don't call it: use of this system call has long been discouraged". could be related? possibly icw macros not being defined before trying the conftest? |
| 07:10.31 | wdoekes | or "This system call is available only if the kernel was configured with the CONFIG_SYSCTL_SYSCALL option." |
| 07:11.49 | snuff-work | ok.. maybe it might be the latter.. but if i just define it.. asterisk still appears to 'do the right thing' |
| 07:12.22 | snuff-work | aka still shows 'swap space' in my case when doing core show sysinfo |
| 07:12.29 | snuff-work | and doesn't die horribly |
| 07:14.51 | snuff-work | wdoekes: is there an easy option to know if that var was set in my kernel build? |
| 07:15.02 | *** join/#asterisk-dev pchero_work (~pchero@109.70.54.56) |
| 07:17.59 | *** join/#asterisk-dev piling_ (~piling@unaffiliated/piling) |
| 07:18.41 | wdoekes | $ grep CONFIG_SYSCTL_SYSCALL /boot/config-`uname -r` |
| 07:18.41 | wdoekes | CONFIG_SYSCTL_SYSCALL=y |
| 07:18.58 | wdoekes | I suppose? |
| 07:19.23 | snuff-work | mm.. i see for my debian that it says its 'not configured' |
| 07:19.33 | snuff-work | but it still got autoconf fine |
| 07:22.13 | wdoekes | but does it use sysctl? or does it use the preferred sysinfo? |
| 07:26.54 | snuff-work | it has sysctl.. |
| 07:27.08 | wdoekes | ah.. |
| 07:27.15 | wdoekes | I think you have sysinfo too |
| 07:27.16 | snuff-work | i think its more the 'ifdef' around swap |
| 07:27.20 | wdoekes | yes |
| 07:27.25 | wdoekes | you don't have HAVE_SWAPCTL |
| 07:27.32 | wdoekes | so it needs the SYSCTL to show the values |
| 07:27.32 | snuff-work | yes.. |
| 07:28.00 | snuff-work | but it doesn't.. because those values are calc either with SYSCTL OR SYSINFO |
| 07:28.08 | wdoekes | correct |
| 07:28.30 | wdoekes | should be SYSCTL&&SWAPCTL || SYSINFO |
| 07:28.32 | snuff-work | so probably.. the ifdef in the show sysinfo should be changed |
| 07:30.06 | snuff-work | but even so.. the ifdef.. sysctl section doesn't mention swapctl either |
| 07:30.22 | snuff-work | in calling the 'swapmode' cmd |
| 07:36.06 | wdoekes | yes, but if there is no swapctl, then it returns 0 for the values, so they're not useful. but perhaps they should be checked for 0 instead of testing the macros |
| 08:03.44 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 08:04.06 | *** join/#asterisk-dev tzafrir (~tzafrir@137.158.23.63) |
| 08:04.06 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 08:18.27 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 08:20.28 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 08:26.11 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 08:43.49 | *** join/#asterisk-dev jkroon (~jkroon@uls-154-73-35-201.wall.uls.co.za) |
| 09:02.44 | *** join/#asterisk-dev oej (~oej@195-67-63-194.customer.telia.com) |
| 09:02.44 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 09:49.09 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 10:54.06 | *** join/#asterisk-dev oej (~oej@195-67-63-194.customer.telia.com) |
| 10:54.06 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 11:30.27 | *** join/#asterisk-dev oej (~oej@195-67-63-194.customer.telia.com) |
| 11:30.27 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 11:35.10 | *** join/#asterisk-dev oej (~oej@195.67.63.194) |
| 11:35.10 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 11:41.26 | *** join/#asterisk-dev tzafrir (~tzafrir@137.158.23.63) |
| 11:41.26 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 11:56.52 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 12:12.34 | *** join/#asterisk-dev acidfu_ (~acidfoo@modemcable002.114-70-69.static.videotron.ca) |
| 12:13.00 | gtjoseph | file: when you got the error yourself on the goto_on_blindxfr test, what version of Ubuntu 64 were you on? |
| 12:13.08 | file | 15.10 |
| 12:13.19 | file | Jenkins stuff runs 14.04 |
| 12:13.53 | gtjoseph | yeah, i've been testing on 14 and hadn't been able to reproduce. gonna try 15-64 now. |
| 12:14.08 | file | it happened for awhile for me, and then it stopped |
| 12:14.13 | file | it's... odd |
| 12:14.58 | *** join/#asterisk-dev newtonr (~newtonr@173-21-146-94.client.mchsi.com) |
| 12:14.58 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 12:15.13 | file | A wild newtonr from home appears. |
| 12:15.25 | gtjoseph | yep. it does seem to happen every time it runs on the ubuntu14-32 node though. |
| 12:15.54 | newtonr | wobbles |
| 12:20.15 | file | gtjoseph, I can offline that VM |
| 12:21.07 | gtjoseph | that might help to see if it happens on any of the others. it didn't on the centos ones but it hadn't run on the other ubuntu nodes in recent history. |
| 12:23.15 | file | I think today is Fake Monday. |
| 14:27.21 | *** join/#asterisk-dev tzafrir (~tzafrir@137.158.23.63) |
| 14:27.21 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 14:32.45 | *** join/#asterisk-dev putnopvut (putnopvut@asterisk/master-of-queues/mmichelson) |
| 14:32.45 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:39.03 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-qauqvqtjpgsapgkr) |
| 14:41.44 | file | nods to putnopvut and rmudgett |
| 14:41.52 | putnopvut | Don't hurt yourself |
| 14:42.09 | file | too late |
| 15:11.05 | file | gtjoseph, I *think* it's the sending of the two INVITE requests at the same time which is the trigger |
| 15:11.32 | file | or at least close to each other |
| 15:11.55 | file | I fudged the testsuite last week to wait 5 seconds at startup before doing the originates and it still happened |
| 15:11.56 | gtjoseph | hmmm. why would THAT trigger the issue??? |
| 15:12.14 | file | if I knew that I'd be wiser |
| 15:13.04 | gtjoseph | the error was EACCES correct? |
| 15:14.11 | file | EPERM |
| 15:14.29 | gtjoseph | that's even worse |
| 15:14.38 | file | I KNOWWWWWWWWWW |
| 15:14.44 | file | >_> |
| 15:16.43 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 15:16.44 | gtjoseph | i wonder what wmem_default and wmem_max are set to, or maybe udm_mem. could we be trying to send a huge datagram? |
| 15:17.07 | gtjoseph | or several large ones that fill the buffers |
| 15:17.19 | file | shouldn't be abnormally big |
| 15:17.28 | file | ye ol' INVITE to ourselves |
| 15:17.43 | file | that's another point - it's an INVITE from chan_pjsip to chan_pjsip |
| 15:18.15 | gtjoseph | 2 instances or loopback? |
| 15:18.24 | file | loopback |
| 15:19.06 | gtjoseph | hmmmmm. |
| 15:19.40 | file | Indeed |
| 15:19.49 | file | The sun is hot. |
| 15:21.03 | gtjoseph | it's using only 1 transport. i winder what'd happen if it used 2? one for outbound and the other for inbound. |
| 15:21.42 | gtjoseph | not as a solution but just as a test |
| 15:22.06 | file | Hrm good point |
| 15:22.48 | gtjoseph | how many cores and how much memory does the ubuntu14-32 node have? |
| 15:23.08 | file | 2 cores I think 1GV |
| 15:23.11 | file | GB |
| 15:23.27 | gtjoseph | let me knock mine down and try again |
| 15:23.32 | file | It's a container so I can allocate whatever and I don't oversubscribe on memory |
| 15:24.16 | gtjoseph | i just want to see if i can reproduce it now. |
| 15:25.38 | file | The issue is so narrow and weird that if separate transports worked... I'd be down with that |
| 15:26.58 | gtjoseph | still passes for me |
| 15:27.37 | gtjoseph | did you take that node out of the rotation? |
| 15:28.07 | file | Yes |
| 15:28.20 | file | 32 is out, do what you want |
| 15:28.26 | file | 64 is in |
| 15:28.36 | gtjoseph | maybe run the test manually on 32 |
| 15:28.59 | gtjoseph | just to confirm that it can be reproduced at will on at least some machine |
| 15:31.42 | file | At will, I wish |
| 15:32.30 | gtjoseph | trying now |
| 15:34.50 | file | eats lunch |
| 15:52.23 | file | run test run |
| 15:53.07 | file | gtjoseph, having a transport per endpoint doesn't break anything so that's good |
| 15:53.53 | gtjoseph | wouldn't think so as long as all the contacts have the correct ports |
| 15:54.00 | file | loop de loop |
| 15:54.22 | gtjoseph | just running that one test? |
| 15:54.39 | file | yes |
| 15:54.55 | file | it failed in the same scenario on my dev box in the past, so I'll let it go go go |
| 15:54.58 | gtjoseph | has it passed at least once? |
| 15:55.08 | file | yes |
| 15:55.17 | file | it was consistently failing, then it consistently passed |
| 15:55.23 | gtjoseph | i ran it 5 times about 10 minutes ago and it failed every time |
| 15:55.31 | gtjoseph | cool |
| 15:55.39 | file | (on my dev box, not on a build agent) |
| 15:55.55 | file | where did it fail, the 32-bit agent? |
| 15:55.59 | gtjoseph | yep |
| 15:56.36 | file | is that genesis or gravity? |
| 15:56.47 | gtjoseph | whichever one is port 2022 |
| 15:56.59 | gtjoseph | the hostnames didn't resolve for me |
| 15:57.27 | file | yeah they are local to me |
| 15:58.02 | file | let's see what happens |
| 15:58.50 | file | passed |
| 15:59.03 | file | using the separate transports |
| 15:59.37 | file | passed again |
| 15:59.52 | gtjoseph | on the agent? |
| 15:59.57 | file | yes |
| 16:00.00 | gtjoseph | yay! |
| 16:00.09 | gtjoseph | doesn't explain the problem though. |
| 16:00.17 | file | indeed |
| 16:00.34 | file | I suspect it's along the lines of what you were saying |
| 16:00.38 | gtjoseph | effectively what kernel is in use? |
| 16:00.56 | file | 4.4.6 |
| 16:01.19 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 16:01.28 | gtjoseph | i wonder if it's an issue with libraries expecting an older kernel? |
| 16:01.47 | file | I'd expect weirder stuff |
| 16:01.50 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 16:02.06 | gtjoseph | true but it might explain why it was working on my VM |
| 16:02.15 | file | wouldn't explain why it failed on my dev box |
| 16:02.23 | gtjoseph | ah, no true. |
| 16:02.59 | file | I think it's the sending of both SIP INVITEs on the loopback near the same time causing something lower level to tell us to back off |
| 16:03.11 | gtjoseph | yeah, makes sense |
| 16:03.13 | file | separate transports, separate ports, separate sockets |
| 16:03.24 | file | so it doesn't happen there |
| 16:03.50 | file | I've run it 5 times on the agent, all passed |
| 16:04.02 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 16:04.45 | gtjoseph | it's odd though... /proc/sys/net/ipv4 doesn't look like a 4.x kernel. |
| 16:05.04 | file | LXC might expose something different |
| 16:05.12 | gtjoseph | yeah, i guess. |
| 16:07.51 | file | I'm going to get the transport changes up, we can get them in and see what happens in practice |
| 16:08.00 | gtjoseph | makes sense |
| 16:08.31 | gtjoseph | maybe take the 64 bit node out to send more to the 32 bit node |
| 16:08.54 | file | I'm not that bothered about it |
| 16:10.42 | file | on the plus side Jenkins is mighty stable on the periodics |
| 16:11.01 | gtjoseph | don't jinx it |
| 16:11.46 | file | I'd like to jinx it! would flush out flaky tests more |
| 16:32.49 | *** join/#asterisk-dev DivideBy0 (~DivideBy0@unaffiliated/divideby0x0) |
| 16:42.55 | *** join/#asterisk-dev newtonr (~newtonr@173-21-146-94.client.mchsi.com) |
| 16:42.55 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 16:55.47 | *** join/#asterisk-dev pchero (~pchero@109.70.54.56) |
| 16:59.27 | *** join/#asterisk-dev piling_ (~piling@unaffiliated/piling) |
| 17:17.42 | *** join/#asterisk-dev piling (~piling@unaffiliated/piling) |
| 17:21.26 | *** join/#asterisk-dev jameswf (uid27319@gateway/web/irccloud.com/x-zeqhqvpyxxtonfvz) |
| 17:34.49 | *** join/#asterisk-dev piling_ (~piling@unaffiliated/piling) |
| 17:41.23 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01cb0583d31d009cc0f9f2bb49e0b7.ipv6.abo.wanadoo.fr) |
| 18:29.51 | *** join/#asterisk-dev piling_ (~piling@unaffiliated/piling) |
| 19:21.43 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 19:34.15 | *** join/#asterisk-dev opticron (~opticron@75.76.51.46) |
| 19:34.15 | *** mode/#asterisk-dev [+o opticron] by ChanServ |
| 19:55.44 | *** join/#asterisk-dev F2Knight_ (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 20:24.04 | *** join/#asterisk-dev F2Knight (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 20:40.31 | *** join/#asterisk-dev F2Knight (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 20:49.02 | gtjoseph | putnopvut: working on missing ari events... does this ring any bells? If I create then dial a local channel, whether the dialplan calls hangup or I do a DELETE, the only event generated is a ChannelHangupRequest and the channel actually stays active. If I put the channel in a bridge before dialing, everything works as expected. |
| 20:49.31 | gtjoseph | something about the channel being in a bridge makes it work |
| 20:50.32 | putnopvut | That's an interesting discovery. I'm thinking about why that would happen. |
| 20:51.16 | gtjoseph | the other thing i discovered was that ari_cchannel_create wasn't subscribing the channels so that explains part of the missing events. |
| 20:52.29 | putnopvut | If the bridge is not being destroyed in your test, it's possible that its continued existence is what gets you channel events. Maybe the bridge acts as a subscriber for channel events. So if the bridge sticks around, you still get events for that channel? Except I'd expect that if you remove the channel from the bridge, the bridge would no longer be interested, so it doesn't really make a lot of sense. |
| 20:53.13 | gtjoseph | hmmm. it's not only events in this case though, the channel doesn't actually get hung up |
| 20:54.06 | gtjoseph | although that may still be stasis related if messages aren't being generated i guess. |
| 20:57.20 | *** join/#asterisk-dev klow (~textual@c-98-247-49-57.hsd1.wa.comcast.net) |
| 21:18.31 | *** join/#asterisk-dev newtonr_ (~newtonr@173-21-146-94.client.mchsi.com) |
| 21:18.32 | *** mode/#asterisk-dev [+o newtonr_] by ChanServ |
| 21:19.10 | *** join/#asterisk-dev F2Knight (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 21:26.22 | *** join/#asterisk-dev F2Knight (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 21:26.28 | gtjoseph | putnopvut: digging deeper, is there something that the dial bridge may not be doing that a 'real' bridge does? maybe because it's invisible? |
| 21:29.22 | *** join/#asterisk-dev putnopvut (putnopvut@asterisk/master-of-queues/mmichelson) |
| 21:29.22 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 21:30.48 | file | putnopvut, . |
| 21:30.49 | file | <gtjoseph> putnopvut: digging deeper, is there something that the dial bridge may not be doing that a 'real' bridge does? maybe because it's invisible? |
| 21:31.19 | *** join/#asterisk-dev F2Knight (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 21:32.55 | putnopvut | The invisible bridge does have bridge-related stasis publications stifled. But channel events shouldn't have any changes. The bridge's creation, aside from being invisible, is the same as any other Stasis bridge. |
| 21:33.30 | gtjoseph | yeah, i just changed it to mixing and visible and it didn't help |
| 21:34.33 | *** join/#asterisk-dev F2Knight (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 21:51.14 | *** join/#asterisk-dev F2Knight (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 22:00.27 | *** join/#asterisk-dev F2Knight (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 22:25.11 | *** join/#asterisk-dev F2Knight (~F2Knight@c-50-139-85-237.hsd1.or.comcast.net) |
| 22:35.02 | *** join/#asterisk-dev tzafrir (~tzafrir@137.158.23.63) |
| 22:35.02 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 23:09.42 | *** join/#asterisk-dev babak (uid19622@gateway/web/irccloud.com/x-ogeenzjuclukwspj) |
| 23:13.37 | *** join/#asterisk-dev kippi (97e42d58@gateway/web/freenode/ip.151.228.45.88) |
| 23:13.47 | kippi | I am trying to up the amount of pseudo channels, however /sys/module/dahdi/parameters/max_pseudo_channels is missing, how can I get these defaults created |