| 00:29.07 | *** join/#asterisk-dev pchero (~pchero@211.178.226.108) |
| 00:29.07 | *** mode/#asterisk-dev [+o pchero] by ChanServ |
| 06:34.24 | *** join/#asterisk-dev jkroon (~jkroon@41.183.141.207) |
| 06:34.24 | *** mode/#asterisk-dev [+o jkroon] by ChanServ |
| 07:04.09 | *** join/#asterisk-dev jkroon (~jkroon@41.183.141.207) |
| 07:04.09 | *** mode/#asterisk-dev [+o jkroon] by ChanServ |
| 07:40.39 | *** join/#asterisk-dev jkroon (~jkroon@105.0.3.111) |
| 07:40.39 | *** mode/#asterisk-dev [+o jkroon] by ChanServ |
| 08:02.53 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:c93c:d751:8bb3:bb87) |
| 08:05.46 | *** join/#asterisk-dev CELYA_1 (~Thunderbi@lstlambert-657-1-90-73.w80-13.abo.wanadoo.fr) |
| 08:10.44 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:4191:86f9:52b9:6e89) |
| 08:35.33 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01cb0583d31d00dd6c39e4e3d4e964.ipv6.abo.wanadoo.fr) |
| 08:56.06 | *** join/#asterisk-dev jkroon (~jkroon@105.0.3.111) |
| 08:56.06 | *** mode/#asterisk-dev [+o jkroon] by ChanServ |
| 12:03.06 | *** join/#asterisk-dev jkroon (~jkroon@41.183.141.218) |
| 12:03.06 | *** mode/#asterisk-dev [+o jkroon] by ChanServ |
| 13:02.43 | *** join/#asterisk-dev jkroon (~jkroon@41.183.141.197) |
| 13:02.43 | *** mode/#asterisk-dev [+o jkroon] by ChanServ |
| 13:57.49 | *** join/#asterisk-dev kharwell (uid358942@gateway/web/irccloud.com/x-pgqwpssvzfpojibk) |
| 13:57.50 | *** mode/#asterisk-dev [+o kharwell] by ChanServ |
| 13:59.45 | *** join/#asterisk-dev bford (uid283514@gateway/web/irccloud.com/x-ltvdnnymuncexeqv) |
| 13:59.45 | *** mode/#asterisk-dev [+o bford] by ChanServ |
| 14:52.41 | seanbright | kharwell: that new debug log is a disaster |
| 14:56.32 | file | oh no, I just fed into your ego on another issue |
| 14:56.50 | kharwell | seanbright: oh wow just not looking. indeed a mess |
| 14:56.58 | kharwell | *now looking that is |
| 15:01.23 | seanbright | file: you had me at "you are correct" |
| 15:01.45 | file | seanbright: you are correct. |
| 15:02.14 | seanbright | so the RED issue is really "when beating the living hell out of asterisk... and also using red... it crashes" |
| 15:05.04 | file | kaboom |
| 15:08.13 | kharwell | red dead with no redemption |
| 15:14.49 | kharwell | lots of fracks around the scheduler. seanbright makes me think your original theory about SCHED_DEL could be right |
| 16:02.19 | *** join/#asterisk-dev skrusty (~skrusty@88.150.145.104) |
| 16:02.20 | *** mode/#asterisk-dev [+o skrusty] by ChanServ |
| 16:11.25 | seanbright | i wonder if we should create a separate scheduler context just for RED |
| 16:11.50 | seanbright | maybe something else that is using the same context is effing things up and it's just manifesting in the RED code because it runs so often |
| 16:12.18 | seanbright | i could also be talking out of my ass |
| 16:12.24 | seanbright | all of these things are equally likely |
| 16:12.30 | file | it's possible, but I doubtz it |
| 16:13.03 | seanbright | the talking out of my ass part? |
| 16:13.12 | seanbright | or the scheduler context theory? |
| 16:13.16 | file | yes |
| 16:13.25 | seanbright | mmhmm |
| 16:13.30 | file | :D to the scheduler context theory |
| 16:13.57 | seanbright | well then i got nothin' |
| 16:16.45 | file | do you have water at least? |
| 16:16.51 | file | hydrate, hydrate, hydrate |
| 16:27.55 | kharwell | hydrates with hot water steeped in ground coffee beans |
| 16:28.49 | seanbright | kharwell: did you notice that there were 3M log lines in just about 1 minute |
| 16:29.12 | kharwell | I did not |
| 16:29.22 | seanbright | i guess AMI logging makes that number kinda bullshity |
| 16:29.23 | kharwell | and some were even missing |
| 16:29.24 | seanbright | but still |
| 16:43.58 | seanbright | last question - could gdb tell you what line is here: |
| 16:43.59 | seanbright | 4/asterisk/modules/res_rtp_asterisk.so(+0x24000) [0x7f5265bfd000] |
| 16:44.31 | seanbright | # 3: /usr/lib64/asterisk/modules/res_rtp_asterisk.so(+0x24000) [0x7f5265bfd000] |
| 16:45.52 | seanbright | either way, the problem is clear, the rtp instance is getting destroyed and then someone is still trying to use it |
| 16:46.04 | seanbright | you can send payment to sean.bright@gmail.com |
| 16:47.30 | file | do you accept monopoly money? |
| 16:58.37 | kharwell | I think we'd have to have the core dump for that run. Even then not sure |
| 17:00.52 | kharwell | was trying to look back at the core dump I do have and see if the sched shows anything as you mentioned. Prob not helpful but "deleted = 0" for the currently running red_write task. So either it failed to delete properly or hasn't been yet |
| 17:07.50 | kharwell | now this might be interesting. "sched_thread = 0x0" seems like that shouldn't be NULL? |
| 17:08.35 | file | within the context of chan_sip that is normal |
| 17:18.01 | seanbright | chan_sip is awful |
| 17:20.57 | *** join/#asterisk-dev HannaM (~quassel@p54849510.dip0.t-ipconnect.de) |
| 17:24.19 | kharwell | looking at the one rtp object that was able to be printed it appears everything is in an init state. Think the object is actually corrupted by the point of failure and just happened to be able to still access it somewhat? |
| 17:27.41 | seanbright | i am all of a sudden happy to not have access to the core files |
| 18:52.37 | kharwell | what's the unique_id on an rtp_engine instance? I found the channel object associated with that unique id and it seems to be pointing to a different rtp instance |
| 19:02.58 | *** join/#asterisk-dev pchero (~pchero@211.178.226.108) |
| 19:02.58 | *** mode/#asterisk-dev [+o pchero] by ChanServ |
| 19:13.30 | file | kharwell: should be the channel |