| 01:20.19 | *** join/#asterisk-dev erinspice (~erin@207.98.195.107) |
| 01:45.34 | *** join/#asterisk-dev wdbl (~not@ool-44c0619d.dyn.optonline.net) |
| 02:10.26 | *** join/#asterisk-dev philipp64|laptop (~chatzilla@192.55.64.175) |
| 03:08.28 | *** join/#asterisk-dev datachomper (~Adium@207.98.189.248) |
| 03:10.25 | *** part/#asterisk-dev datachomper (~Adium@207.98.189.248) |
| 04:05.09 | *** join/#asterisk-dev razu (~razu@razu.data.ee) |
| 04:11.50 | *** join/#asterisk-dev philipp64|laptop (~chatzilla@71-34-116-81.ptld.qwest.net) |
| 06:07.09 | *** join/#asterisk-dev tzafrir_laptop (~tzafrir@bzq-218-155-145.cablep.bezeqint.net) |
| 06:07.09 | *** mode/#asterisk-dev [+o tzafrir_laptop] by ChanServ |
| 07:17.35 | *** join/#asterisk-dev philipp64|laptop (~chatzilla@71-34-116-81.ptld.qwest.net) |
| 08:13.10 | *** join/#asterisk-dev tzafrir_laptop (~tzafrir@local.xorcom.com) |
| 08:13.10 | *** mode/#asterisk-dev [+o tzafrir_laptop] by ChanServ |
| 08:25.28 | *** join/#asterisk-dev tamiel (~tamiel@213.30.183.226) |
| 08:31.23 | *** join/#asterisk-dev schmidts (~schmidts@2001:858:4:0:66b9:e8ff:fec3:ae88) |
| 08:31.24 | schmidts | good morning |
| 08:40.14 | *** join/#asterisk-dev hehol (~hehol@2001:1438:1009:200:2c33:2562:f0bd:cf24) |
| 08:49.06 | schmidts | anybody awake right now? |
| 08:50.02 | schmidts | does someone of you have an idea why 1.8 needs 5 times more memory than 1.6? i got a statistic for sip calls which shows this difference and i dont have an idea what could cause this. |
| 08:50.59 | schmidts | only one thing comes to my mind, that 1.8 allocates a udptl struct for every sip pvt even if its not needed but 5 times more memory usage?? |
| 08:53.14 | *** join/#asterisk-dev rushowr (~rushowr@99-28-31-100.lightspeed.stlsmo.sbcglobal.net) |
| 08:57.02 | *** part/#asterisk-dev rushowr (~rushowr@99-28-31-100.lightspeed.stlsmo.sbcglobal.net) |
| 10:22.29 | *** join/#asterisk-dev ccesario (~ccesario@187.75.139.188) |
| 10:24.49 | *** join/#asterisk-dev sgimeno (~chatzilla@163.117.211.10) |
| 10:26.17 | *** join/#asterisk-dev guilhermebr (~Guilherme@200.103.96.98) |
| 10:43.22 | *** join/#asterisk-dev ccesario (~ccesario@187.75.139.188) |
| 12:42.08 | pabelanger | schmidts: I _think_ chan_iax2 is using more then before. Check reviewboard for a patch |
| 13:01.01 | Marquis42 | pabelanger: That was my first guess too, but are you loading chan_iax2 schmidts? |
| 13:01.13 | *** join/#asterisk-dev rrb3942 (~rbullock@208.34.105.161) |
| 13:02.58 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 13:02.58 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 13:18.09 | schmidts | yes its load but not used, this was only test with sip |
| 13:18.38 | schmidts | but to be honest, i didnt made this test i only helped to figure the testing parameters and got the results ;) |
| 13:19.41 | schmidts | the review i start today could cause a little bit of this memory consumption but i am not sure if this would be the only thing. What comes to my mind are the ao2 containers which are much more used in 1.8 than 1.6 but i dont think that its that big issue. |
| 13:19.45 | schmidts | atleast memory is cheap ;) |
| 13:21.23 | *** join/#asterisk-dev russellb (~russell@asterisk/digium-open-source-team-lead/russellb) |
| 13:21.23 | *** mode/#asterisk-dev [+o russellb] by ChanServ |
| 13:24.32 | *** join/#asterisk-dev Dovid (~Dovid@213.8.118.62) |
| 13:24.42 | Dovid | any "bug marshals" here ? |
| 13:29.29 | schmidts | dovid maybe we can help you too? |
| 13:29.56 | Dovid | ok. i seem to have found a bug in 1.8.X (tried 1.8.0 and 1.8.2.2) |
| 13:30.24 | Dovid | http://pastebin.com/7Btbrdxu |
| 13:31.23 | Dovid | if the called party picks up and does not enter any DTMF then the caller hears tt-monkeys, however if the called party presses DTMF, then the called party hangs up, the call coninues for the caller however he does not hear anything |
| 13:32.40 | Dovid | so there is no audio to the caller after the called hangs up (if the call was bridged) |
| 13:33.49 | schmidts | maquis42 and pabelanger: http://pastebin.com/UExGHju4 these are the test results i got and you can see a big difference in memory usage. |
| 13:34.07 | schmidts | the 8% more cpu usage of 1.8 may belong to ipv6 |
| 13:35.38 | Dovid | schmidts: Do you think I should open a ticket for it ? |
| 13:37.38 | pabelanger | Dovid: Did you try #asterisk first? |
| 13:38.12 | Dovid | pabelanger: No beause it looks more like a bug then functionality. |
| 13:38.29 | pabelanger | Start there, any post more information. EG: debug logs |
| 13:38.31 | Dovid | isnt #asterisk for gen questions and #asterisk-dev for bugs ? |
| 13:38.57 | pabelanger | Dovid: No, #asterisk-bugs is. -dev is for development |
| 13:41.44 | *** join/#asterisk-dev Corydon76-home (four@c-69-137-80-31.hsd1.tn.comcast.net) |
| 13:41.45 | *** mode/#asterisk-dev [+o Corydon76-home] by ChanServ |
| 13:42.09 | schmidts | any idea what could cause this big memory usage increase? |
| 13:46.15 | kpfleming | schmidts: what memory increase? |
| 13:47.00 | kpfleming | you are aware of the patch that tilghman has on reviewboard to drastically reduce memory consumption, right? |
| 13:48.27 | Marquis42 | schmidts: You don't have to be using iax2, just loading it. chan_iax2 allocates 32768 a02 containers (with an extra 3k of memory each) upon loading. |
| 13:51.05 | kpfleming | objects, not containers |
| 13:51.12 | Marquis42 | Err, yeah |
| 13:52.20 | Marquis42 | And also, while iax2 is the big offender due to its initial allocations, it also means that *every* ao2 object is using an extra 3k of memory. |
| 13:53.50 | kpfleming | and every other object with an embedded ast_mutex |
| 13:54.31 | Marquis42 | Oh that's right, because that's actually the source of the issue. |
| 13:55.09 | schmidts | ok now i knew which review you meant, i will try this out. sound like a good reason |
| 13:58.22 | schmidts | thx! |
| 14:05.55 | *** join/#asterisk-dev _schmidts (~schmidts@213.235.212.193) |
| 14:08.20 | *** join/#asterisk-dev hehol (~hehol@2001:1438:1009:200:9c80:629c:c23e:2b0e) |
| 14:44.21 | *** join/#asterisk-dev elguero (~miguel323@12.187.84.162) |
| 14:52.02 | *** join/#asterisk-dev putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 14:52.02 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:54.55 | *** join/#asterisk-dev jeffg (~jeffg@pdpc/supporter/professional/jeffg) |
| 15:00.40 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 15:09.44 | *** join/#asterisk-dev kay2 (~key2@olm03.cvf.fr) |
| 15:19.13 | *** join/#asterisk-dev The_Boy_Wonder (~vossel@asterisk/batman-developer/dvossel) |
| 15:19.13 | *** mode/#asterisk-dev [+o The_Boy_Wonder] by ChanServ |
| 15:37.20 | *** join/#asterisk-dev datachomper (~Adium@nat/digium/x-kxebimsffdyuvedt) |
| 15:45.17 | *** join/#asterisk-dev sruffell (~sruffell@asterisk/the-kernel-guy/sruffell) |
| 15:45.17 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 16:41.27 | *** join/#asterisk-dev marv (~timr@router.asteriasgi.com) |
| 16:42.23 | *** join/#asterisk-dev rrb3942 (~rbullock@208.34.105.161) |
| 17:14.49 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 17:14.49 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 18:16.12 | *** join/#asterisk-dev russellb (~russell@asterisk/digium-open-source-team-lead/russellb) |
| 18:16.12 | *** mode/#asterisk-dev [+o russellb] by ChanServ |
| 18:39.13 | *** join/#asterisk-dev Tim_Toady (~moi@178.128.134.192.dsl.dyn.forthnet.gr) |
| 18:47.20 | *** join/#asterisk-dev The_Boy_Wonder (~vossel@asterisk/batman-developer/dvossel) |
| 18:47.20 | *** mode/#asterisk-dev [+o The_Boy_Wonder] by ChanServ |
| 18:51.14 | *** join/#asterisk-dev The_Boy_Wonder (~vossel@asterisk/batman-developer/dvossel) |
| 18:51.14 | *** mode/#asterisk-dev [+o The_Boy_Wonder] by ChanServ |
| 20:29.54 | putnopvut | I was looking in main/udptl.c and found that the ast_udptl has an undocumented 'verbose' field. The only place that it is used is in udptl_build_packet. If 'verbose' is non-zero then a newline is printed to stderr. Since there's no code path that ever sets 'verbose' to anything non-zero and it seems pretty dumb to output a newline to stderr whenever a UDPTL packet is created, would anyone take issue with my removing ast_ud |
| 20:29.54 | putnopvut | ptl::verbose from all branches? |
| 20:30.51 | putnopvut | The reason I ask instead of just doing it is that it's not really a bug fix so much as a cleanup, and so since it's not really hurting anything, I can understand if I should just put the change in trunk. |
| 20:35.39 | seanbright | like fprintf(stderr, ...)? |
| 20:37.55 | putnopvut | yeah |
| 20:38.02 | seanbright | toss it |
| 20:38.20 | putnopvut | all right! |
| 20:38.23 | seanbright | my stamp of approval carries some serious weight around here |
| 20:38.25 | seanbright | DO IT |
| 20:57.03 | kpfleming | yeah, that sounds like it should go, putnopvut |
| 21:05.23 | *** join/#asterisk-dev datachomper (~Adium@nat/digium/x-ijdrkvyclduvlbgr) |
| 21:12.24 | *** join/#asterisk-dev philipp64|laptop (~chatzilla@192.55.64.175) |
| 21:23.51 | *** join/#asterisk-dev sruffell (~Adium@asterisk/the-kernel-guy/sruffell) |
| 21:23.51 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 22:22.26 | *** join/#asterisk-dev datachomper (~Adium@nat/digium/x-kktsrknmdxqrqmbr) |
| 22:40.03 | *** join/#asterisk-dev guilhermebr (~Guilherme@189.63.57.56) |
| 22:51.19 | *** join/#asterisk-dev sruffell (~Adium@asterisk/the-kernel-guy/sruffell) |
| 22:51.19 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 22:51.49 | *** join/#asterisk-dev sruffell (~sruffell@asterisk/the-kernel-guy/sruffell) |
| 22:51.49 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 23:52.20 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 23:52.20 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |