| 00:38.28 | *** join/#asterisk-bugs mjordan (~mjordan@user-69-1-6-49.knology.net) |
| 00:38.29 | *** mode/#asterisk-bugs [+o mjordan] by ChanServ |
| 01:41.12 | *** join/#asterisk-bugs tzafrir (~tzafrir@212.179.75.202) |
| 01:52.52 | *** join/#asterisk-bugs mjordan (~mjordan@user-69-1-6-49.knology.net) |
| 01:52.52 | *** mode/#asterisk-bugs [+o mjordan] by ChanServ |
| 02:50.44 | *** join/#asterisk-bugs ChanServ (ChanServ@services.) |
| 02:50.44 | *** mode/#asterisk-bugs [+o ChanServ] by kornbluth.freenode.net |
| 05:15.07 | *** join/#asterisk-bugs Entomologist (~nobody@nat/digium/x-ngllbbmyliibaaus) |
| 05:15.49 | *** join/#asterisk-bugs jrose_atDigium (~jrose_atD@nat/digium/x-gijugozgxqmurbhz) |
| 05:16.31 | *** join/#asterisk-bugs malcolmd (~malcolmd@pdpc/sponsor/digium/malcolmd) |
| 05:16.31 | *** mode/#asterisk-bugs [+o malcolmd] by ChanServ |
| 09:36.20 | *** join/#asterisk-bugs tzafrir (~tzafrir@local.xorcom.com) |
| 11:07.10 | *** join/#asterisk-bugs seanbright (~sean@asterisk/contributor-and-bug-marshal/seanbright) |
| 11:07.10 | *** mode/#asterisk-bugs [+o seanbright] by ChanServ |
| 11:43.47 | *** join/#asterisk-bugs blitzrage (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 11:43.47 | *** mode/#asterisk-bugs [+o blitzrage] by ChanServ |
| 12:08.59 | *** join/#asterisk-bugs tzafrir (~tzafrir@local.xorcom.com) |
| 13:04.47 | *** join/#asterisk-bugs mjordan (~mjordan@nat/digium/x-vdfpeoilblcmxbxf) |
| 13:04.47 | *** mode/#asterisk-bugs [+o mjordan] by ChanServ |
| 13:26.30 | *** join/#asterisk-bugs serafie (~erin@nat/digium/x-dyqgkkksbfbiufci) |
| 13:51.55 | *** join/#asterisk-bugs putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 13:51.55 | *** mode/#asterisk-bugs [+o putnopvut] by ChanServ |
| 14:37.54 | *** join/#asterisk-bugs zerohalo (~zerohalo@67.103.230.22) |
| 15:17.22 | *** join/#asterisk-bugs newtonr (~newtonr@nat/digium/x-dctunntsowztidta) |
| 15:27.27 | *** join/#asterisk-bugs jrose_atDigium (~jrose_atD@nat/digium/x-bftbjfnkepdwinnh) |
| 20:01.43 | *** join/#asterisk-bugs jgowdy (jgowdy@basharteg.com) |
| 20:04.44 | mjordan | jgowdy: so the last check-in against res_timing_timerfd that affected file descriptors was on 08/17/11 |
| 20:04.59 | mjordan | so whatever behavior you're seeing, it was most likely prevalent in all of Asterisk 10 |
| 20:05.32 | jgowdy | Yeah, I've seen it off and on, but the patching was more to do with the usage |
| 20:05.37 | jgowdy | https://issues.asterisk.org/jira/browse/ASTERISK-18223 |
| 20:05.45 | jgowdy | there was a fix to manager.c |
| 20:05.53 | mjordan | right |
| 20:06.05 | mjordan | but if you're seeing that in res_timing_timerfd, I wouldn't expect that issue or patch to apply to it |
| 20:06.25 | mjordan | in which case, you might want to open a new issue in JIRA |
| 20:06.29 | jgowdy | unless the issue is similar to the manager.c issue, where the timer handle itself was being used after close etc |
| 20:06.35 | jgowdy | I'm still trying to trace it out |
| 20:06.56 | mjordan | that's possible, but the fix for that issue wouldn't have affected it. The root causes may be similar in nature. |
| 20:07.02 | jgowdy | yep |
| 20:07.09 | mjordan | and we'd still want a new issue in JIRA, since it is a different module |
| 20:07.20 | jgowdy | definitely |
| 20:08.06 | mjordan | I have to pop off to a meeting, but please do file an issue. It most likely isn't something that has been looked at yet. |
| 20:09.23 | jgowdy | I will |
| 20:09.29 | jgowdy | I have some externalivr.c fixes too |
| 20:09.34 | jgowdy | I've been meaning to make a case for |
| 20:21.40 | jgowdy | https://issues.asterisk.org/jira/browse/ASTERISK-20032 |
| 20:24.16 | mjordan | yick |
| 20:24.53 | mjordan | does it only happen in interactions with externalivr? |
| 20:25.41 | mjordan | I'll comment on the issue :-) |
| 20:31.55 | jgowdy | It's a big mesh of ExternalIVR, FastAGI, and AMI |
| 20:32.18 | jgowdy | I'm trying to isolate where in our code we are |
| 20:32.23 | jgowdy | so I could more accurately answer that |
| 20:32.42 | jgowdy | The problem is my current sample of this had a couple hundred concurrent calls going on |
| 20:33.22 | mjordan | what channel technologies? |
| 20:35.53 | jgowdy | chan_sip |
| 20:36.47 | jgowdy | 100% G.711 u-law voice, have spandsp compiled in for fax, but I haven't seen any information to show this related to faxing |
| 20:38.13 | mjordan | nah, just was thinking of local channels |
| 20:38.24 | mjordan | may want to tack a sanitized sip.conf up on there as well. |
| 20:38.33 | mjordan | Have you noticed any orphaned SIP channels when this happens? |
| 20:41.13 | jgowdy | I'll have to recreate the situation to find out, but as far as I've seen it's only the one channel acting up, and everything else looks fairly normal |
| 20:41.20 | jgowdy | I'm betting it's just one call scenario |
| 20:41.30 | jgowdy | there's a couple things I'm not liking |
| 20:41.45 | jgowdy | One, when we close the fd for the timer, we don't set to -1 |
| 20:42.38 | mjordan | jgowdy: yeah, I agree, that would be a good practice |
| 20:42.38 | jgowdy | and timerfd_timer_ack doesn't have a return value |
| 20:42.53 | jgowdy | so there's no way to know that we've gone to hell and we should probably murder this channel |
| 20:43.01 | jgowdy | which is why it just spins |
| 20:44.57 | jgowdy | so either way, I'm going to patch to -1 and check for -1 before ao2find and submit that |
| 20:45.05 | jgowdy | just for quality's sake |
| 20:45.20 | mjordan | okay |
| 20:45.57 | jgowdy | I know the bug is research light, it won't be for too long |
| 20:46.01 | jgowdy | just getting it in there |
| 21:22.58 | *** part/#asterisk-bugs mjordan (~mjordan@nat/digium/x-vdfpeoilblcmxbxf) |
| 21:33.14 | *** part/#asterisk-bugs newtonr (~newtonr@nat/digium/x-dctunntsowztidta) |
| 22:16.45 | *** join/#asterisk-bugs tzafrir (~tzafrir@212.179.40.134) |
| 23:37.46 | jgowdy | wow, did they remove W and w from senddtmf? |
| 23:40.58 | leifmadsen | define, "they" |
| 23:41.11 | jgowdy | don't know yet, I'm searching for the commit |