| 00:24.10 | *** join/#asterisk-dev kamanashisroy (n=kamanash@202.56.7.193) |
| 00:34.31 | *** join/#asterisk-dev outtolunc (n=me@c-67-170-211-86.hsd1.ca.comcast.net) |
| 00:47.01 | *** join/#asterisk-dev d3wayne (n=deeewayn@76.29.245.9) |
| 00:47.01 | *** mode/#asterisk-dev [+o d3wayne] by ChanServ |
| 00:52.46 | *** part/#asterisk-dev RoyK (n=roy@91.149.31.29) |
| 01:27.21 | *** join/#asterisk-dev moy (n=moyhu@189.169.83.74) |
| 01:29.33 | *** join/#asterisk-dev IgorG (n=IgorG@195.162.49.23) |
| 01:40.29 | *** join/#asterisk-dev mog (n=mog@c-68-62-219-86.hsd1.al.comcast.net) |
| 01:40.29 | *** mode/#asterisk-dev [+o mog] by ChanServ |
| 02:03.33 | *** join/#asterisk-dev RoyK (n=roy@91.149.31.29) |
| 02:03.57 | *** part/#asterisk-dev RoyK (n=roy@91.149.31.29) |
| 02:28.43 | *** join/#asterisk-dev AdamWest (n=Leif@CPE001d7e2f9574-CM0012c9db3d2e.cpe.net.cable.rogers.com) |
| 03:50.44 | snuff-home | thinks enum is now on the right track.. |
| 04:20.59 | *** join/#asterisk-dev outtolunc (n=me@c-67-170-211-86.hsd1.ca.comcast.net) |
| 05:51.59 | *** join/#asterisk-dev IgorG (n=IgorG@195.162.49.23) |
| 06:07.38 | *** join/#asterisk-dev steliosk (n=Stelios@85.72.68.184) |
| 06:09.27 | *** join/#asterisk-dev sergee (n=serg@voip1.west-call.com) |
| 07:10.46 | *** part/#asterisk-dev mog (n=mog@c-68-62-219-86.hsd1.al.comcast.net) |
| 07:57.11 | *** join/#asterisk-dev Daviey (n=dave@ubuntu/member/daviey) |
| 08:33.30 | *** join/#asterisk-dev qdk_ (n=qdk@85.235.253.139) |
| 08:40.03 | *** part/#asterisk-dev jmls_home (n=jmls_hom@mail.tessera.co.uk) |
| 08:52.18 | *** join/#asterisk-dev RoyK (n=roy@fw.fortel.no) |
| 09:04.01 | *** join/#asterisk-dev seanbright (n=sean@c-69-251-175-43.hsd1.md.comcast.net) |
| 09:16.55 | *** join/#asterisk-dev fcois (n=fcois@bagnolet.acropolistelecom.net) |
| 09:17.01 | fcois | hello |
| 09:17.13 | fcois | I need help for an 'asterisk appliance aa50' |
| 09:18.01 | *** join/#asterisk-dev magumbade (n=magumbad@p5497C4A4.dip.t-dialin.net) |
| 09:18.57 | fcois | I need to add the tftp command in the busybox 1.0 system, know you how to do? |
| 09:21.58 | *** join/#asterisk-dev IgorG (n=IgorG@195.162.49.23) |
| 10:02.27 | fcois | need help for an 'asterisk appliance aa50' |
| 10:03.38 | snuff-home | ask in #asterisk |
| 10:04.12 | fcois | I tryed but nobody speak |
| 10:09.44 | *** part/#asterisk-dev fcois (n=fcois@bagnolet.acropolistelecom.net) |
| 10:28.36 | *** join/#asterisk-dev jmls_work (n=jmls_wor@mail.tessera.co.uk) |
| 10:53.10 | *** join/#asterisk-dev DagMoller (n=aguirre@unaffiliated/dagmoller) |
| 10:55.44 | *** join/#asterisk-dev DagMoller (n=aguirre@unaffiliated/dagmoller) |
| 11:37.38 | *** join/#asterisk-dev kamanashisroy (n=kamanash@202.56.7.139) |
| 11:46.32 | *** join/#asterisk-dev lmadsen (n=Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 11:46.33 | *** mode/#asterisk-dev [+o lmadsen] by ChanServ |
| 11:49.07 | *** join/#asterisk-dev ctooley (n=ctooley@mo-76-0-41-172.dhcp.embarqhsd.net) |
| 11:55.32 | *** join/#asterisk-dev caio1982 (i=caio1982@CAcert-br/caio1982) |
| 12:07.42 | *** join/#asterisk-dev _ys (i=yuri@91.151.196.254) |
| 12:14.33 | *** join/#asterisk-dev agx (n=AGX@host63-216-static.34-88-b.business.telecomitalia.it) |
| 12:23.20 | *** join/#asterisk-dev jeffg (i=w0nrbHQl@pdpc/supporter/active/jeffg) |
| 12:32.47 | *** join/#asterisk-dev russellb (n=russell@asterisk/developer-and-stable-maintainer/drumkilla) |
| 12:32.47 | *** mode/#asterisk-dev [+o russellb] by ChanServ |
| 12:35.44 | *** join/#asterisk-dev kamanashisroy (n=kamanash@202.56.7.139) |
| 12:36.55 | *** join/#asterisk-dev shido6 (n=shido6@74-130-51-159.dhcp.insightbb.com) |
| 12:37.45 | shido6 | in gdb, I can find out what context, fullnamme,, timezone attachfmt but how do i find out what msgnum asteirsk was working on when it crashed |
| 12:38.29 | shido6 | im using bt full , then f 5 then p vmu->mailbox, etc but what do i use to find out the msgnum it was working on |
| 12:38.43 | shido6 | and cidname and cidnum |
| 12:39.16 | shido6 | ok i have the current director p vms->curdir |
| 12:44.00 | *** join/#asterisk-dev Iamnacho (i=Iamnacho@ip68-103-153-140.ks.ok.cox.net) |
| 13:18.35 | *** join/#asterisk-dev zerohalo (n=zeroHalo@pool-72-70-66-31.bstnma.east.verizon.net) |
| 13:26.48 | *** join/#asterisk-dev moy (n=moyhu@nat/ibm/x-540ff69dae155230) |
| 13:30.26 | *** join/#asterisk-dev kamanashisroy (n=kamanash@202.56.7.135) |
| 13:33.52 | *** join/#asterisk-dev anthm (n=anthm@CPE-72-131-113-50.wi.res.rr.com) |
| 14:05.35 | *** join/#asterisk-dev joshaidan (n=joshaida@S0106001c1023e838.tb.shawcable.net) |
| 14:11.42 | *** join/#asterisk-dev anonymouz666 (n=anonymou@201.19.205.41) |
| 14:27.25 | *** join/#asterisk-dev mog (n=mog@c-68-62-219-86.hsd1.al.comcast.net) |
| 14:27.25 | *** mode/#asterisk-dev [+o mog] by ChanServ |
| 14:30.06 | *** join/#asterisk-dev sruffell (n=sruffell@adsl-074-185-078-025.sip.bhm.bellsouth.net) |
| 14:30.06 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 14:30.07 | *** join/#asterisk-dev putnopvut (n=putnopvu@216.207.245.1) |
| 14:30.07 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:46.01 | *** join/#asterisk-dev russellb (n=russell@asterisk/developer-and-stable-maintainer/drumkilla) |
| 14:46.01 | *** mode/#asterisk-dev [+o russellb] by ChanServ |
| 14:54.41 | *** join/#asterisk-dev joetester2 (n=joeteste@216.191.34.13) |
| 14:57.42 | *** join/#asterisk-dev snuff-home (n=snuff-ho@60-240-117-7-nsw-pppoe.tpgi.com.au) |
| 15:30.24 | putnopvut | seanbright: ping |
| 15:35.09 | *** join/#asterisk-dev DarkRift (n=dark@bas10-montreal02-1177584428.dsl.bell.ca) |
| 15:38.18 | *** join/#asterisk-dev CunningPike (n=arodgers@204.239.12.183) |
| 15:44.04 | shido6 | can someone look at line 3899 of app_voicemail.c , how would I check msg_cat value before calling ast_Variable_update() |
| 15:45.07 | putnopvut | shido6: what version of Asterisk? |
| 15:45.37 | shido6 | 1.4.18 |
| 15:45.40 | shido6 | http://pastebin.ca/1010643 |
| 15:46.56 | putnopvut | shido6: so what do you mean by "value" before? Do you want to know the address that the pointer is pointing to, or something else? |
| 15:47.05 | putnopvut | s/do/did/ |
| 15:47.13 | putnopvut | crap |
| 15:48.02 | shido6 | I want to check the msg category before calling the vatiable update ..... otherwise the context info is lost and asterisk crashes |
| 15:48.06 | shido6 | for example..... |
| 15:48.11 | shido6 | a user logs in |
| 15:48.14 | shido6 | listens to msg 1 |
| 15:48.18 | shido6 | and wants to forward it |
| 15:48.24 | shido6 | upon forwarding |
| 15:48.26 | shido6 | asterisk crashes |
| 15:49.36 | putnopvut | Ah, so you want to check if the category still exists before the ast_variable_update? |
| 15:50.13 | putnopvut | It looks like you're doing that in your pastebin by insuring that msg_cat is non-NULL when calling ast_category_get. |
| 15:51.57 | *** join/#asterisk-dev qdk_ (n=qdk@0x573fe57e.bynqu2.broadband.tele.dk) |
| 15:52.04 | shido6 | http://pastebin.ca/1010648 |
| 15:52.24 | shido6 | when i insure its not null i break other stuff |
| 15:52.54 | shido6 | brb |
| 15:52.56 | putnopvut | What other stuff do you break? |
| 15:57.40 | russellb | while : ; do sudo asterisk -rx "sip reload" ; done |
| 15:57.44 | russellb | come on, crash, i dare you! |
| 15:58.06 | file | crashes russellb |
| 15:59.27 | Corydon76-dig | russellb: start that in two different windows, so that they collide with each other |
| 15:59.44 | russellb | good idea. |
| 15:59.52 | Qwell | fork bomb it |
| 16:00.01 | file | the actual reload operation can't collide, can it? |
| 16:00.12 | Qwell | in sip it can...sort of |
| 16:00.23 | Qwell | it spawns a thread and then unlocks the reload lock |
| 16:00.27 | Qwell | (iirc) |
| 16:00.46 | russellb | depends what you mean by collide |
| 16:00.56 | russellb | if you reload while a reload is in progress, it knows it and says NO |
| 16:01.03 | Qwell | maybe no thread |
| 16:01.17 | *** join/#asterisk-dev outtolunc (n=me@adsl-66-218-53-172.dslextreme.com) |
| 16:01.19 | russellb | the reload is done within the monitor thread i think |
| 16:02.04 | file | pretty sure as well |
| 16:02.50 | file | dances |
| 16:17.37 | snuff-home | lol @ falves latest comment |
| 16:18.03 | russellb | what is it? |
| 16:18.06 | russellb | i'm too lazy to look |
| 16:18.17 | putnopvut | snuff-home: which issue? |
| 16:18.31 | russellb | can't muffinman show comments now? |
| 16:18.35 | file | M12566#86539 |
| 16:18.36 | MuffinMan | [12566#0086539] Commented by falves11 on 05-07-08 at 10:44: Is there a volunteer that wishes to help capture the debugging information, since Digium's Leif has no time and I don't feel like paying support for bugs? I can dedicate a machine only for analizing this bug. |
| 16:18.37 | Qwell | he can! |
| 16:18.51 | russellb | O.O |
| 16:19.44 | russellb | lmadsen: how dare you not have time to help everyone who wants help for free!!!!!11 |
| 16:20.15 | Juggie | eugh, so sleepy. |
| 16:22.49 | snuff-home | crosses fingers in hope that enum works as required now for new patch |
| 16:23.31 | jtodd | snuff-home: I'll take a poke at it in a few minutes. |
| 16:24.36 | snuff-home | cool.. :) i know i fixed isn.. least the 2 tests i had 13/14 which were ur 5/6.. they worked for me anyway |
| 16:25.27 | snuff-home | well snooze time.. leave a msg etc jtodd with how u went |
| 16:34.46 | lmadsen | russellb: I know... I'm a real bastard |
| 16:35.04 | lmadsen | russellb: I got an email to my work address asking... I had to respectfully decline |
| 16:36.42 | Juggie | lmadsen loves money too much |
| 16:36.48 | Juggie | how is the condo hunt going lmadsen |
| 16:36.49 | lmadsen | heck ya |
| 16:36.58 | lmadsen | when you live in Toronto you don't have much of a choice |
| 16:37.06 | Juggie | no i imagine |
| 16:37.10 | Juggie | 300,000+ prob is it? |
| 16:37.14 | lmadsen | Juggie: found a great place that overlooks the skydome, but the guy wants about $20-$25k too much |
| 16:37.16 | Juggie | or maybe 200,000 for 600sq feet |
| 16:37.28 | lmadsen | $235-$250k for 550sqft |
| 16:37.34 | Juggie | figures |
| 16:37.34 | lmadsen | or ya, up to 600sqft |
| 16:37.45 | Juggie | expensive area |
| 16:37.54 | lmadsen | indeed |
| 16:38.04 | Juggie | maybe maurice has a condo he will sell for cheep :) |
| 16:38.15 | lmadsen | :D |
| 16:38.18 | Juggie | you know what sucks |
| 16:38.23 | Juggie | when you incorrectally budget your bills |
| 16:38.30 | lmadsen | Juggie: lets move this to -doc |
| 16:55.43 | Juggie | anyone know the ldap driver asterisk wants for centos? |
| 16:56.23 | *** join/#asterisk-dev atis_work (n=atis_wor@193.238.212.171) |
| 16:57.06 | seanbright | openldap? |
| 16:57.20 | Juggie | its installed but configure still fails |
| 16:57.34 | Juggie | maybe there is a openldap-devel |
| 16:57.36 | Juggie | looks |
| 16:58.54 | *** join/#asterisk-dev linuxer_igor (n=igor@mlsrj200152096p202.mls.com.br) |
| 16:59.07 | Juggie | indeed there is |
| 17:03.42 | lmadsen | good work |
| 17:07.06 | *** join/#asterisk-dev outtolunc (n=me@adsl-66-218-53-172.dslextreme.com) |
| 17:12.01 | seanbright | holy hell, i can't parse what is going on in #asterisk _at all_ |
| 17:12.03 | Juggie | seanbright, does my patch on 12599 look fine |
| 17:14.44 | seanbright | Juggie: the documentation says that, or just the log message? |
| 17:16.13 | seanbright | Juggie: looks like just the log message says asterisk is used as a default |
| 17:16.33 | seanbright | Juggie: the sample config makes no mention of it |
| 17:18.43 | seanbright | Juggie: i'd say commit it to trunk, but just fix the log message in 1.6.0 (but i'm probably not the person to make that call) |
| 17:20.46 | *** join/#asterisk-dev bbryant (n=brett@216.207.245.1) |
| 17:20.46 | *** mode/#asterisk-dev [+o bbryant] by ChanServ |
| 17:22.38 | Juggie | anynoe else seeing a coredump in trnk? |
| 17:22.45 | Juggie | i'm getting one in res_odbc |
| 17:23.16 | seanbright | does run trunk |
| 17:23.25 | seanbright | man... i CANNOT type today |
| 17:23.28 | seanbright | doesn't run trunk |
| 17:23.32 | Juggie | *** glibc detected *** asterisk: munmap_chunk(): invalid pointer: 0x0964bc10 *** |
| 17:23.34 | lmadsen | seanbright: I was the same way this morning too |
| 17:23.35 | Juggie | fancy :) |
| 17:24.09 | seanbright | Juggie: compile with DONT_OPTIMIZE and all that jazz? |
| 17:24.22 | Juggie | i'm about to |
| 17:24.25 | Juggie | so i can see whats up |
| 17:25.27 | Juggie | hmmm |
| 17:25.30 | Juggie | the core is actually ok |
| 17:25.49 | Juggie | (without dont_optimize) |
| 17:26.12 | Juggie | relatively anyway |
| 17:26.19 | lmadsen | heh |
| 17:28.23 | Juggie | Corydon76-dig: ping |
| 17:28.56 | Juggie | looks to me like res_odbc is coreing if it loads and no config file exists |
| 17:29.05 | Juggie | it then tries to free memory it never read, and kaboom |
| 17:29.11 | Juggie | er, never filled. |
| 17:29.40 | Juggie | hmmm |
| 17:29.43 | Juggie | nm thats not it |
| 17:31.36 | Juggie | it is broken though |
| 17:31.39 | Juggie | wonder if 1.6 is too |
| 17:33.29 | *** kick/#asterisk-dev [file!n=russell@asterisk/developer-and-stable-maintainer/drumkilla] by russellb (russellb) |
| 17:33.33 | Juggie | haha... |
| 17:33.49 | Juggie | russellb, did you get realtime setup? |
| 17:34.16 | *** join/#asterisk-dev file (n=file@asterisk/developer-and-muffin-lover/file) |
| 17:34.16 | *** mode/#asterisk-dev [+o file] by ChanServ |
| 17:38.18 | *** join/#asterisk-dev kamanashisroy (n=kamanash@202.56.7.134) |
| 17:52.45 | *** join/#asterisk-dev dlynes (n=chatzill@dsl-vlan468-66-18-244-66.nucleus.com) |
| 17:57.30 | Juggie | ahha |
| 17:58.37 | russellb | Juggie: yup |
| 17:58.47 | Juggie | found the bug in res_odbc |
| 17:58.55 | Juggie | when the connection fails, it sets obj to null |
| 17:59.24 | Juggie | then odbc_obj_destructor does an ast_free(obj); |
| 17:59.31 | Corydon76-dig | That's not a bug |
| 17:59.38 | Juggie | well its core dumping |
| 17:59.47 | Corydon76-dig | free(NULL) is legitimate |
| 18:01.13 | russellb | free(NULL) bothers me :-p |
| 18:01.21 | file | free(tibet) |
| 18:01.25 | putnopvut | free(beer) |
| 18:01.37 | anonymouz666 | free(kevin) |
| 18:01.45 | russellb | i actually submitted a patch to a certain C++ based project that removed a bunch of checks for non-NULLness before delete |
| 18:01.48 | russellb | since it isn't necessary |
| 18:01.52 | russellb | and it got rejected |
| 18:02.04 | putnopvut | Did they say why? |
| 18:02.11 | russellb | said it was a stupid idea |
| 18:02.18 | putnopvut | lol |
| 18:02.31 | putnopvut | I wouldn't call it stupid, but perhaps not worth implementing. |
| 18:02.33 | russellb | compatability with non-standard C++ crap crap .. |
| 18:02.39 | russellb | but i had already done it, heh |
| 18:03.07 | Corydon76-dig | Because in C++, delete != free |
| 18:03.23 | Corydon76-dig | Well, it is on Linux, but it isn't on other platforms |
| 18:03.25 | Juggie | Corydon76-dig, i've got a core dump in res_odb in trunk atm |
| 18:03.32 | russellb | it is valid to do delete NULL |
| 18:03.33 | Juggie | which is new. as i haev not touched my config files |
| 18:03.35 | putnopvut | Juggie: what's the backtrace look like? |
| 18:03.43 | Juggie | #4 0x001930eb in free () from /lib/libc.so.6 |
| 18:03.43 | Juggie | #5 0x06509205 in odbc_obj_destructor (data=0x9fb6bc8) at res_odbc.c:103 |
| 18:03.43 | Juggie | #6 0x08076fc3 in __ao2_ref (user_data=0x9fb6bc8, delta=-1) at astobj2.c:254 |
| 18:03.56 | Corydon76-dig | Juggie: yay |
| 18:04.24 | Corydon76-dig | Juggie: bug number? |
| 18:04.25 | putnopvut | Juggie: What is obj? |
| 18:04.39 | putnopvut | As in what's the value of obj? |
| 18:04.40 | Juggie | its an odbc_obj |
| 18:04.42 | russellb | an object, obviously |
| 18:04.45 | Juggie | and its null because the connection didnt work |
| 18:05.28 | russellb | gives putnopvut a large tractor |
| 18:05.38 | putnopvut | butt |
| 18:05.53 | Corydon76-dig | Is 'tractor' the safe word today? |
| 18:06.28 | putnopvut | Juggie: I don't see how obj is NULL in that backtrace. |
| 18:06.38 | putnopvut | It should be 0x9fb6bc8 |
| 18:06.58 | putnopvut | Which maybe already free'd. |
| 18:07.30 | russellb | yay for breakage |
| 18:07.34 | russellb | high fives Corydon76-dig |
| 18:07.38 | russellb | at least you didn't break chan_sip :( |
| 18:07.45 | Juggie | its definitally null, or it should be |
| 18:07.50 | file | wait... |
| 18:08.01 | russellb | waits |
| 18:08.11 | Juggie | its set to null @ line 549 when the connection fails and then the function returns it. |
| 18:08.28 | file | you don't free the object in the destructor callback |
| 18:08.36 | *** join/#asterisk-dev joetester2 (n=joeteste@216.191.34.13) |
| 18:08.38 | putnopvut | Juggie: that's not the same pointer. |
| 18:08.56 | putnopvut | look at odbc_obj_destructor. |
| 18:09.10 | putnopvut | obj is local to that function and is set equal to data (which is 0x9fb6bc8) |
| 18:09.19 | putnopvut | Nothing that is called from there sets obj NULL. |
| 18:11.05 | file | bad bad... |
| 18:11.14 | Corydon76-dig | file: circular references |
| 18:11.15 | seanbright | ok, i got allison to record "dick wilson" |
| 18:11.24 | putnopvut | lol? |
| 18:11.31 | seanbright | heh |
| 18:11.39 | seanbright | decides to make a remix |
| 18:11.40 | Corydon76-dig | file: that's why it doesn't destroy it in THAT destructor |
| 18:12.07 | Corydon76-dig | Basically, I deactivated one-half of the cascade |
| 18:12.12 | file | Corydon76-dig: I'm looking at odbc_obj_destructor, you are freeing the object passed in... |
| 18:12.39 | russellb | that is bad, mmmkay? |
| 18:12.39 | file | ponders |
| 18:12.42 | Juggie | ya thats the line thats crashing |
| 18:12.51 | file | very bad. |
| 18:13.06 | Juggie | just reverted all my debuging and did a new build to make sure the line numbers matched up |
| 18:13.13 | Juggie | and the ast_Free is what is crashing |
| 18:13.26 | Corydon76-dig | Sorry, my mind is in another repo at the moment |
| 18:13.44 | Corydon76-dig | Please leave a message, and I'll get back to you. <beep> |
| 18:14.04 | russellb | Hey Corydon76-dig, fix that shit, kthxbye |
| 18:14.30 | Juggie | is it as simple as just removing the ast_free? |
| 18:14.32 | russellb | I demand that it may or may not bee too urgent |
| 18:14.38 | file | Juggie: should be. |
| 18:17.30 | Juggie | file that fixed the crash as you would expect |
| 18:17.55 | Corydon76-dig | Juggie: question is, does it leak memory on a reload? |
| 18:18.41 | Juggie | whats the command to see module memory, i cant remember. |
| 18:18.52 | putnopvut | memory show allocations |
| 18:18.53 | russellb | memory show summary |
| 18:18.57 | putnopvut | or that one :) |
| 18:19.05 | russellb | memory show summary res_odbc.c |
| 18:19.06 | putnopvut | yeah, the summary is better. |
| 18:19.14 | Juggie | hrmf i have no memory |
| 18:19.18 | Juggie | i presume i'm missing some flag |
| 18:19.22 | russellb | you have to compile with MALLOC_DEBUG |
| 18:19.23 | putnopvut | Juggie: build with MALLOC_DEBUG |
| 18:19.35 | Juggie | standby. |
| 18:20.09 | russellb | stands |
| 18:20.20 | file | pushes russellb |
| 18:20.44 | Corydon76-dig | No, I guess that is okay, still |
| 18:20.44 | Juggie | :) i also have a patch for another reported bug on the tracker russellb, but i'm not sure what the intended functionality is, and its ldap so i dont know much about it either |
| 18:21.02 | Corydon76-dig | Juggie is the new LDAP maintainer! |
| 18:21.07 | Juggie | No way :) |
| 18:21.10 | putnopvut | lol |
| 18:21.15 | Corydon76-dig | You know more than I do |
| 18:21.26 | russellb | Juggie: congratulations! |
| 18:21.32 | file | Juggie: congrats! |
| 18:21.39 | russellb | my name is in there, but only because i reworked most of the code because it was insanely ugly |
| 18:21.43 | russellb | not because i know anything about ldap |
| 18:21.52 | Juggie | i know less.... so i deny! :) |
| 18:21.54 | file | you have been voluntold. |
| 18:21.55 | putnopvut | I know what LDAP stands for, but that's about it. |
| 18:22.02 | russellb | putnopvut: same |
| 18:22.02 | Juggie | i do not even remember that |
| 18:22.13 | putnopvut | Lightweight Directory Access Protocol |
| 18:22.13 | Juggie | the D stands for directory |
| 18:22.19 | russellb | Corydon76-dig: you told me you'd maintain it when you committed it, sooo ... ;) |
| 18:22.21 | Corydon76-dig | Juggie: if you use LDAP at all, you have more experience than the rest of us combined |
| 18:22.22 | ctooley | What's wrong with the LDAP module? |
| 18:22.46 | russellb | it exists. |
| 18:22.52 | Juggie | just some silly bug with the debug message in the code saying it was setting a value to 'asterisk' |
| 18:22.55 | Juggie | if it wasnt configured |
| 18:23.04 | Corydon76-dig | ctooley: there's a few bugs open, and most of us have never used it |
| 18:23.06 | ctooley | :) There's nothing exactly "wrong" with LDAP other than being overly complicated and flakey |
| 18:23.08 | Juggie | but then it did var[0]=\0; |
| 18:23.15 | Juggie | so it set it to nothing :) |
| 18:23.24 | Juggie | so the notice and code dont match |
| 18:23.25 | ctooley | goes off to look at the bugs. |
| 18:23.43 | putnopvut | starts setting up OS X tests... |
| 18:28.00 | Juggie | Corydon76-dig, are you going to commit this res_odbc fix? |
| 18:28.53 | russellb | Juggie: you have commit don't you? |
| 18:29.01 | Juggie | i know but i dont want us both to commit it |
| 18:29.25 | seanbright | Juggie: that can't happen |
| 18:29.41 | putnopvut | Yeah, whoever tries to commit second will be told their file is out of date. |
| 18:37.01 | mvanbaak | hello all :) |
| 18:37.12 | file | waves to mvanbaak |
| 18:37.32 | mvanbaak | howz life here ? |
| 18:37.43 | mvanbaak | <--- took a bunch of days off to enjoy his new boat |
| 18:38.35 | putnopvut | Life's good, as always. |
| 18:39.49 | mvanbaak | nice |
| 18:43.58 | ctooley | Corydon76-dig, I can't tell if 12523 is still a problem or not from his last comment. |
| 18:44.18 | ctooley | I can set up an environment to try to replicate a problem but I can't tell if there actually is one. |
| 18:44.42 | *** join/#asterisk-dev jmls_home (n=jmls_hom@mail.tessera.co.uk) |
| 18:45.02 | jmls_home | mutters because he's lost his backlog _again_ |
| 18:45.30 | jsmith-teaching | jmls_home: That's OK... I'm trying to kill a bug or two while teaching this bootcamp class |
| 18:45.39 | jmls_home | putnopvut: hopeful ping |
| 18:45.57 | putnopvut | jmls_home: hi. I looked at the code and your backtrace some more this morning. |
| 18:46.02 | jmls_home | jsmith-teaching: as interesting as the one I found yesterday ? |
| 18:46.14 | jmls_home | waits with baited breath .... |
| 18:46.20 | putnopvut | jmls_home: it's not good. |
| 18:46.26 | jsmith-teaching | jmls_home: No... what one did you find yesterday? |
| 18:46.32 | jmls_home | define "not good" |
| 18:46.38 | putnopvut | There are two odd things about you're situation. |
| 18:46.56 | jmls_home | jsmith-teaching a channel hang in an exclusive macro ;) |
| 18:47.01 | jmls_home | go on ... |
| 18:47.20 | putnopvut | First, it should be impossible for autoservice to get in the state it was in, and second, even if it did get in that state, it doesn't make sense that it would hang. |
| 18:47.26 | jmls_home | heh |
| 18:47.36 | jmls_home | does the impossible |
| 18:47.45 | jmls_home | twice in two days |
| 18:47.48 | putnopvut | seriously, I have no idea how that could happen. |
| 18:48.03 | jmls_home | bugger. that is not good |
| 18:48.05 | Corydon76-dig | Memory corruption |
| 18:48.32 | Corydon76-dig | And you know what we do with memory corruption.....valgrind! |
| 18:48.37 | putnopvut | Corydon76-dig: did you see where the hang was happening? It's in ast_autoservice_stop. |
| 18:49.11 | putnopvut | I learned last night that the 'i' in valgrind is pronounced with a short vowel sound, like "grinned" |
| 18:49.29 | jmls_home | Corydon76-dig: it's something that happened twice in two days on one channel out of probably 50000+ per day |
| 18:50.20 | jmls_home | *and* both at the same point in the dialplan |
| 18:55.38 | *** join/#asterisk-dev atis_work (n=atis_wor@193.238.212.171) |
| 18:58.11 | *** join/#asterisk-dev CrashHD (n=CrashHD@65.74.161.225) |
| 19:04.32 | Juggie | on a svn update, whats G mean on a file. |
| 19:05.17 | Juggie | ah |
| 19:05.21 | Juggie | should read docs, merged. |
| 19:07.00 | Juggie | thanks for commiting that patch russ. |
| 19:08.51 | *** join/#asterisk-dev outtolunc (n=me@adsl-66-218-53-172.dslextreme.com) |
| 19:11.23 | *** join/#asterisk-dev jameswf-home (n=james@dsl093-157-131.phx1.dsl.speakeasy.net) |
| 19:20.26 | *** join/#asterisk-dev DagMoller (n=aguirre@unaffiliated/dagmoller) |
| 19:36.10 | Qwell | so, if we have the manager users list..why do we parse the file in authenticate()? |
| 19:40.17 | lmadsen | putnopvut: hrmmm... I like the long 'i' sound better |
| 19:40.20 | jsmith-teaching | In case they changed their password? |
| 19:42.17 | Qwell | jsmith-teaching: without a reload? |
| 19:42.26 | ctooley | that's what I was thinking, doesn't that require a reload? |
| 19:42.28 | Qwell | that goes against every other config file |
| 19:42.34 | Qwell | ctooley: actually, it doesn't currently |
| 19:42.45 | Qwell | but, one would sort of expect it to, I'd think |
| 19:43.13 | ctooley | so manager.conf is a "realtime" config file? |
| 19:43.21 | Qwell | only in some cases |
| 19:43.27 | ctooley | nice |
| 19:43.32 | Qwell | if you do manager show users, no, it'll use what's in memory |
| 19:44.46 | Juggie | does it read the config every time a user logs in? |
| 19:44.53 | Qwell | yes |
| 19:45.01 | Juggie | h0t |
| 19:45.29 | jsmith-teaching | Well, I'm just saying how it *is*, now how it *should be* |
| 19:45.38 | Qwell | jsmith-teaching: right |
| 19:45.57 | Corydon76-dig | I don't believe it's that way in trunk |
| 19:46.11 | ctooley | heh, I actually assumed since there was no reload for the manager it required a restart. ;) |
| 19:46.15 | jsmith-teaching | wonders if the change got documented |
| 19:46.39 | Qwell | Corydon76-dig: yeah, you're right |
| 19:47.24 | Corydon76-dig | So it's already fixed; it just hasn't propagated to a release yet. |
| 19:47.47 | Qwell | so then... |
| 19:47.47 | Corydon76-dig | It was part of the major overhaul to manager, starting in 1.6.0 |
| 19:48.04 | Qwell | M12594 |
| 19:48.05 | MuffinMan | [new] [Asterisk] Core/ManagerInterface 0012594: [patch] Add manager users from users.conf to the users LIST reported by bkruse (Karma: +4.50) http://bugs.digium.com/view.php?id=12594 |
| 19:48.19 | mvanbaak | 98.7KB/s |
| 19:48.22 | mvanbaak | ugh, my upload sux |
| 19:50.02 | Corydon76-dig | Qwell: right, he's talking about 1.4 |
| 19:50.30 | Qwell | I guess rather than try to rework it, I can just use his method.. |
| 19:52.14 | *** join/#asterisk-dev Cresl1n (n=matt@216.207.245.1) |
| 19:52.14 | *** mode/#asterisk-dev [+o Cresl1n] by ChanServ |
| 19:53.17 | jsmith-teaching | Well if it isn't Cresl1n! |
| 19:53.24 | Cresl1n | hey all! |
| 19:53.27 | Cresl1n | jsmith-teaching: !!!! |
| 19:55.18 | *** join/#asterisk-dev atis_work (n=atis_wor@193.238.212.171) |
| 20:03.20 | mvanbaak | am I cool or what? http://images.vanbaak.info/displayimage.php?album=44&pos=36 |
| 20:11.37 | *** join/#asterisk-dev jmls (n=asterisk@host217-36-208-155.in-addr.btopenworld.com) |
| 20:11.58 | jmls | now I'm getting really annoyed with IRC |
| 20:12.15 | mvanbaak | lol |
| 20:12.15 | jmls | what did I miss ? (20:10 onwards) |
| 20:12.27 | mvanbaak | erm, a lot |
| 20:12.33 | jmls | ... |
| 20:12.50 | mvanbaak | 3 times page_up on 1600x1200 font set to 8pt |
| 20:13.06 | jmls_home | bugger. |
| 20:13.16 | mvanbaak | mostli manager stuff |
| 20:13.20 | bkruse | Qwell: In Authenticate() (I believe) It says /* XXX Why are we parsing this twice? We already have the user data */ and it goes against how all the other config files work, however, it has been like that forever :/ |
| 20:13.22 | mvanbaak | s/mostli/mostly/ |
| 20:13.42 | Qwell | bkruse: yeah, attaching new patch right now |
| 20:13.44 | Qwell | give it a try? |
| 20:13.49 | Qwell | adds hasmanager support |
| 20:14.17 | bkruse | Qwell: Yes |
| 20:14.31 | bkruse | We need that ability for the GUI anyways (It is a jira bug somewhere....) |
| 20:14.43 | Qwell | added |
| 20:14.59 | Qwell | err, it's wrong. hang on |
| 20:15.01 | bkruse | reloads |
| 20:15.02 | bkruse | k |
| 20:15.09 | jmls_home | putnopvut: do you think my problem could be linked with MacroExclusive itself ? |
| 20:15.19 | bkruse | Show does manager show users work for 1.6.0 and trunk? (people in users.conf) |
| 20:15.38 | Qwell | yes, it's much better in trunk/1.6 |
| 20:15.48 | putnopvut | jmls_home: it could be. The problem looked to be a problem with autoservice, actually, which is used by MacroExclusive. |
| 20:15.54 | bkruse | Qwell: Gotcha, I guess I just did not test since this started as an appliance issue |
| 20:16.25 | jmls_home | I guess not many people used MacroExclusive |
| 20:16.35 | Qwell | better - use -2 |
| 20:17.03 | jmls_home | just wondering if there was some silly / broken / damaged code in the MacroExclusive part to do with locks and autoservice ? |
| 20:17.15 | bkruse | applies |
| 20:17.58 | jmls_home | mutters "Not that I would know damaged C code if it came up like a Great White and bit a huge chunk out of his ass" |
| 20:18.12 | putnopvut | jmls_home: nope, the hang was in autoservice itself. Nothing to do with locking. |
| 20:18.39 | jmls_home | All C looks damaged to me. Especially these goddamned pointer operations .. |
| 20:18.49 | jmls_home | putnopvut: damn |
| 20:19.19 | jmls_home | shakes a stick at his work VPN |
| 20:19.27 | jmls_home | work, damn you |
| 20:20.01 | putnopvut | jmls_home: yeah, it's really strange. Based on the trace that you gave me, it looks like one thread is just looping forever, but it should be impossible for that to happen. |
| 20:20.02 | mvanbaak | pats his OpenBSD ipsec setup |
| 20:20.09 | mvanbaak | works 100% of the time :) |
| 20:20.21 | bkruse | Qwell: works great. |
| 20:22.32 | mvanbaak | any idea when kevin will be here ? |
| 20:22.49 | *** join/#asterisk-dev drummond_ (n=rsd@h-67-103-23-130.phlapafg.covad.net) |
| 20:23.10 | russellb | he's out for the day |
| 20:23.11 | jmls_home | so no idea on what may have caused that ? What is "impossible" ? |
| 20:23.44 | mvanbaak | russellb: thank you |
| 20:23.57 | mvanbaak | russellb: we prolly will have dutch soundfiles soon |
| 20:25.09 | putnopvut | jmls_home: well, it's a bit complicated to explain, but here's a simple explanation. |
| 20:25.27 | mvanbaak | 42 |
| 20:25.32 | mvanbaak | ;) |
| 20:25.33 | putnopvut | mvanbaak: lol |
| 20:25.38 | jmls_home | heh |
| 20:25.52 | jmls_home | sits back in his pupil's chair |
| 20:26.15 | Corydon76-dig | mvanbaak: woot |
| 20:26.27 | putnopvut | jmls_home: so, the autoservice thread's purpose is to just continuously read frames from channels that are currently in the autoservice list. |
| 20:26.37 | mvanbaak | Corydon76-dig: yeah. my wife agreed with the terms kevin stated in his mail |
| 20:26.47 | jmls_home | this thread comes from the channel itself ? |
| 20:26.51 | *** part/#asterisk-dev zerohalo (n=zeroHalo@pool-72-70-66-31.bstnma.east.verizon.net) |
| 20:27.02 | drummond_ | i am having issues trying to build the asterisk addons, can someone help/ |
| 20:27.16 | jmls_home | shhh! |
| 20:27.21 | jmls_home | putnopvut is teaching ... |
| 20:27.22 | jmls_home | ;) |
| 20:27.27 | putnopvut | jmls_home: no, this is a thread that used to just be continuously running at all times, but now is started whenever the first channel goes into autoservice. |
| 20:27.37 | jmls_home | oh, ok |
| 20:27.45 | putnopvut | This thread handles all channels that are in autoservice. |
| 20:28.17 | putnopvut | So, basically, when a channel enters autoservice, the channel is added to the list of channels that are in autoservice, and when it stops autoservice, it is removed from the list. |
| 20:28.24 | jmls_home | ok |
| 20:28.35 | putnopvut | Now, there are some protections added in to handle edge cases. |
| 20:29.07 | jmls_home | nods as if he knows what putnopvut is on about |
| 20:29.15 | jmls_home | *nod* |
| 20:30.13 | putnopvut | Now, I believe the part that's getting stuck is due to a change russellb added recently to be sure that the autoservice thread doesn't have a reference to a channel that was removed from the autoservice list. |
| 20:30.36 | mvanbaak | and when does a channel go into autoservice ? |
| 20:31.07 | jmls_home | blames russellb |
| 20:31.16 | mvanbaak | jmls_home: we all do ;) |
| 20:31.20 | putnopvut | mvanbaak: various times. I think of it going into autoservice when Asterisk has no need to process frames from that channel immediately. Basically if there will be some sort of processing delay. |
| 20:31.24 | mvanbaak | the guy got married ! |
| 20:31.40 | russellb | runs away |
| 20:31.44 | putnopvut | mvanbaak: for instance, in app_queue, autoservice is started on the caller's channel if you have set a memberdelay. |
| 20:32.01 | putnopvut | While that delay is happening, Asterisk pretty much just discards all voice frames that come from the caller's channel. |
| 20:32.18 | mvanbaak | moh, parked calls, etc |
| 20:32.22 | putnopvut | But it will save important frames (like DTMF or hangup) until the channel is removed from autoservice. |
| 20:32.22 | mvanbaak | that kindda channels ? |
| 20:32.51 | putnopvut | mvanbaak: I'd have to double check, but yeah, that sounds right. |
| 20:32.55 | jmls_home | so autoservice is a frame garbage collector ? |
| 20:33.12 | mvanbaak | jmls_home: that's what I'm starting to think indeed |
| 20:33.13 | putnopvut | jmls_home: that's one way to think of it. |
| 20:33.58 | jmls_home | would a CURL call cause an autoservice ? |
| 20:34.09 | putnopvut | jmls_home: yes, I'm almost certain it would. |
| 20:34.15 | Corydon76-dig | Yes, it does |
| 20:34.22 | Corydon76-dig | unless it's not run on a channel |
| 20:34.26 | Corydon76-dig | It's conditional |
| 20:35.00 | jmls_home | so when CURL finishes, the channel is removed from autoservice ? |
| 20:35.07 | mvanbaak | yup |
| 20:35.11 | putnopvut | That's right. |
| 20:35.12 | jmls_home | what about SET ? |
| 20:35.27 | jmls_home | or ARRAY ? |
| 20:35.34 | Corydon76-dig | No, SET or ARRAY shouldn't |
| 20:35.52 | jmls_home | ok. what if the CURL call took such a little time to complete |
| 20:36.27 | jmls_home | so, CURL starts, channel is being added to autoservice, CURL ends before channel has been added |
| 20:36.46 | putnopvut | jmls_home: that can't happen, because it's synchronous. |
| 20:36.56 | jmls_home | bugger |
| 20:37.37 | putnopvut | Channel goes into autoservice. Then the curl call happens. Then it is removed from autoservice. |
| 20:38.08 | jmls_home | try #2. channel is added to autoservice, returns control to CURL. CURL goes away, comes back very quickly, channel is removed from autoservice. All before some process in autoservice is performed on the channel |
| 20:38.31 | putnopvut | jmls_home: That could very well happen, but it shouldn't make a difference. |
| 20:38.39 | putnopvut | But see, there's even more to it in your case. |
| 20:38.50 | jmls_home | sighs. there had to be |
| 20:39.06 | mvanbaak | it has to do with your login name ;) |
| 20:39.08 | jmls_home | I'm just that sort of guy. I attract bugs |
| 20:39.19 | jmls_home | should have seen my first wife .... |
| 20:39.26 | putnopvut | In your backtrace, the autoservice thread is operating on a zero-channel list, even though that SHOULD be impossible. If there are zero channels in the list, it's supposed to sleep until it gets poked saying that there's now a channel in the list. |
| 20:39.59 | jmls_home | what if it got poked, but the channel itself then got removed before it woke up properly |
| 20:40.09 | jmls_home | (it == autoservice) |
| 20:40.17 | mvanbaak | gotta love multithreading |
| 20:41.21 | putnopvut | jmls_home: can't happen due to locks protecting the autoservice list. |
| 20:41.33 | putnopvut | s/locks/a lock/ |
| 20:41.43 | jmls_home | is there a potential hole in the lock protection ? |
| 20:41.56 | jmls_home | like in MacroExclusive ? |
| 20:42.04 | Corydon76-dig | No, it's a mutex |
| 20:42.15 | jmls_home | whoooossh |
| 20:42.21 | putnopvut | jmls_home: If there is, I can't see it. I have probably looked at the backtrace and the code involved about fifty times now and can't see how the problem is occurring. |
| 20:42.29 | Corydon76-dig | You'd have to violate a fundamental law of physics to violate a mutex |
| 20:42.39 | jmls_home | that's the sound I made in university when they started talking about "i" |
| 20:42.43 | Qwell | Corydon76-dig: which law of physics is that? |
| 20:42.52 | Corydon76-dig | Qwell: continuity of time |
| 20:43.22 | Corydon76-dig | You could only violate it if time were disjoint |
| 20:43.38 | putnopvut | Time is only disjoint on weekend nights. |
| 20:45.15 | Corydon76-dig | Mutexes rely on spinlocks, which are the fundamental locking mechanism on most computers |
| 20:46.37 | jmls_home | we have loads of MOH, CURL, parking. I have never seen this execept the two times it's hung on a channel whilst in the MacroExclusive |
| 20:46.46 | mvanbaak | putnopvut: after enough canabis everything is disjoint, no matter what time of day/day of week |
| 20:46.55 | jmls_home | mvanbaak: Class B again .. |
| 20:47.06 | jmls_home | don't come to England :p |
| 20:47.40 | mvanbaak | whehehehehe |
| 20:48.00 | mvanbaak | rolls a bunch of joints and hands them to #asterisk-dev for free |
| 20:48.10 | russellb | O.O |
| 20:48.16 | jmls_home | hasterisk |
| 20:48.23 | jmls_home | damn |
| 20:48.27 | putnopvut | colors russellb's O.O red |
| 20:48.30 | jmls_home | hashterisk |
| 20:48.34 | russellb | putnopvut: nice. |
| 20:48.39 | mvanbaak | woot ! |
| 20:49.31 | mvanbaak | if you guys can attend the next WhatTheHack, please let me know |
| 20:49.40 | mvanbaak | it's this event once every 4 year |
| 20:49.43 | mvanbaak | in .nl |
| 20:49.51 | putnopvut | When is the next one? |
| 20:50.00 | mvanbaak | plenty of devs, plenty of drinks, plenty of canabis, plenty of fun |
| 20:50.12 | putnopvut | s/fun/prostitutes/ |
| 20:50.36 | mvanbaak | 2009 |
| 20:50.38 | mvanbaak | next year |
| 20:51.09 | russellb | bans commits from people visiting .nl |
| 20:51.42 | russellb | s/bans commits/limits commits to team branches/ |
| 20:51.49 | russellb | never know what you might get :-p |
| 20:51.51 | mvanbaak | whehehehehe |
| 20:52.05 | mvanbaak | what about freshly married ppl ? |
| 20:52.15 | jmls_home | run russellb, run |
| 20:53.10 | putnopvut | Interesting...ARRAY() does call autoservice functions. |
| 20:54.16 | mvanbaak | why ? |
| 20:54.43 | putnopvut | I don't know, actually. I'll svn blame it. |
| 20:55.22 | mvanbaak | let me guess: file |
| 20:55.29 | putnopvut | nope, tilghman. |
| 20:55.36 | Qwell | it could take a fairly long amount of time for ARRAY to complete |
| 20:55.37 | file | it is always my fault. |
| 20:55.43 | Qwell | it kinda makes sense |
| 20:55.52 | *** join/#asterisk-dev atis_work (n=atis_wor@193.238.212.171) |
| 20:55.56 | mvanbaak | hhmm, array is that heavy ? |
| 20:56.05 | Qwell | if you pass it 1k args, sure |
| 20:56.15 | mvanbaak | erm, yeah |
| 20:56.25 | Qwell | (even then, probably not *that* much time) |
| 20:56.30 | putnopvut | Yup, that's what the log message indicates. |
| 20:57.17 | mvanbaak | Qwell: if I want to implement realtime in chan_skinny, what do you like to see in different tables ? |
| 20:57.32 | mvanbaak | I'm thinking about devices and lines |
| 20:57.52 | mvanbaak | should we also put the speeddials in a seperate table ? |
| 20:57.56 | Qwell | no |
| 20:58.03 | mvanbaak | or should they be put in the device table ? |
| 20:58.07 | Qwell | device table |
| 20:58.17 | Qwell | not sure about lines |
| 20:58.22 | Qwell | probably |
| 20:58.39 | mvanbaak | hhmm, problem is |
| 20:58.49 | mvanbaak | a device can have N speeddials and N lines |
| 20:59.02 | Qwell | speeddials shouldn't ever be shared |
| 20:59.03 | mvanbaak | some support 1, some support 6 etc |
| 20:59.39 | mvanbaak | so the speeddial field should be a TEXT field with some delimiter |
| 20:59.54 | Qwell | realtime already does delimiting on it's own |
| 21:00.14 | Qwell | used to be ; - I don't think it's changed |
| 21:00.38 | Qwell | so you could have like g729;ulaw;alaw in an allow column in a realtime table |
| 21:00.56 | mvanbaak | because a single speeddial here looks like this: speeddial=> 602314363@skinny_hint_extensions,Sales manager John |
| 21:01.17 | Qwell | so to have 2, it would just be |
| 21:01.34 | Qwell | 1234@skinny_hint_extensions,Alice;1235@skinny_hint_extensions,Bob |
| 21:01.46 | mvanbaak | k |
| 21:01.48 | Qwell | and res_config_* already handles parsing that |
| 21:02.06 | mvanbaak | but lines can have their own settings, that's why I'm thinking about putting them in a seperate table |
| 21:02.16 | Qwell | yeah |
| 21:02.26 | jmls_home | watches putnopvut sniffing around ARRAY with interest |
| 21:03.15 | putnopvut | jmls_home: I'm having another look at this now, trying to come up with a scenario for why this could happen. |
| 21:03.50 | mvanbaak | hhmm, too bad we dont have a build_device function |
| 21:07.31 | jmls_home | gotta get some food, be back in an hour or so |
| 21:07.40 | jmls_home | putnopvut: thanks very much for the time and effort |
| 21:08.39 | Corydon76-dig | jmls_eating: if you had a reproducable case... |
| 21:09.52 | ctooley | Corydon76-dig, reproducable bugs are too easy to fix. Where's the challenge if you know what the problem is? :) |
| 21:13.53 | Corydon76-dig | ctooley: I don't take bugs out; I put them in... ;-) |
| 21:14.09 | ctooley | Corydon76-dig, good, that helps a lot. |
| 21:14.54 | Corydon76-dig | "rebugging" |
| 21:15.06 | Corydon76-dig | aka "programming" |
| 21:15.59 | *** join/#asterisk-dev jmls_home (n=jmls_hom@mail.tessera.co.uk) |
| 21:16.00 | *** join/#asterisk-dev jmls_work (n=jmls_wor@mail.tessera.co.uk) |
| 21:17.41 | mvanbaak | wow, 3 jmls dudes |
| 21:17.45 | mvanbaak | that's ... too much |
| 21:18.18 | Corydon76-dig | clones mvanbaak |
| 21:23.43 | mvanbaak | I'm off |
| 21:23.45 | mvanbaak | latero all |
| 21:33.30 | *** join/#asterisk-dev atis_home (n=chatzill@193.238.213.215) |
| 21:39.53 | seanbright | subject on the rc release e-mail says "rc1" |
| 21:39.58 | seanbright | is probably late to the party |
| 21:40.27 | russellb | i already noticed |
| 21:40.34 | putnopvut | seanbright: heh, the only one who had said anything about it was russellb right after he sent it out. |
| 21:40.35 | russellb | unfortunately, i can't edit the internet |
| 21:40.46 | seanbright | russellb: you can't? you're fired. |
| 21:40.49 | russellb | :( |
| 21:40.58 | file | seanbright: from life? |
| 21:40.58 | seanbright | russellb: ... and re-hired at twice the salary! |
| 21:41.00 | russellb | holds his head in shame |
| 21:41.03 | russellb | w00t! |
| 21:41.13 | file | russellb works for free! |
| 21:41.15 | putnopvut | notes that he needs to mess up more e-mail subjects |
| 21:41.15 | seanbright | russellb: but then, alas, fired yet again. |
| 21:41.19 | seanbright | :) |
| 21:41.26 | russellb | no, _you're_ fired |
| 21:41.33 | seanbright | thank god |
| 21:41.41 | putnopvut | hmm, I wonder if there's ever been a loop in the chain of command like that. |
| 21:41.44 | russellb | does your boss know you work on asterisk all day? |
| 21:41.49 | seanbright | i'm not sure i've ever been so happy! |
| 21:42.18 | seanbright | russellb: i've really just been lurking lately. not doing too much in the way of "workinig on asterisk." |
| 21:42.57 | seanbright | the only commits i've made have been merging in the changes from trunk :) |
| 21:43.23 | russellb | hehe |
| 21:43.27 | *** join/#asterisk-dev RoyK (n=roy@ip-149-4-149-91.dialup.ice.no) |
| 21:43.27 | russellb | I'm just trolling |
| 21:43.30 | seanbright | i know |
| 21:43.34 | russellb | and by the way, I would like to encourage that behavior |
| 21:43.37 | russellb | working on asterisk is quite welcome |
| 21:43.40 | Qwell | trolling behavior? |
| 21:43.55 | *** part/#asterisk-dev RoyK (n=roy@ip-149-4-149-91.dialup.ice.no) |
| 21:44.01 | seanbright | russellb: would love to spend more time on it. oss doesn't pay the bills... at least for me. |
| 21:44.04 | file | Qwell: I thought that that as well, lol |
| 21:44.15 | file | ooh double that |
| 21:44.19 | russellb | yes, trolling is encouraged. |
| 21:44.39 | seanbright | sweet... |
| 21:44.49 | seanbright | digium hardware has irq problems, that's why i always buy sangoma |
| 21:44.51 | seanbright | (like that?) |
| 21:44.57 | russellb | ugh |
| 21:45.00 | seanbright | heh |
| 21:45.09 | russellb | welcome to 3 years ago |
| 21:45.23 | seanbright | some people have trouble letting go |
| 21:45.27 | russellb | yup |
| 21:45.33 | russellb | internet archives, especially |
| 21:45.50 | seanbright | well that b.s. is still making its way into *new* messages |
| 21:45.52 | seanbright | its exhausting. |
| 21:46.03 | seanbright | i've basically stopped reading the users list |
| 21:46.19 | tzanger | heh |
| 21:46.32 | ctooley | unsubscribed a long time ago in an effort to be able to read his email |
| 21:46.42 | seanbright | filters |
| 21:46.45 | tzanger | yeah I should unsubscribe, I post so rarely to it now |
| 21:47.20 | russellb | i occasionally clear out the folder |
| 21:47.22 | russellb | that's about it |
| 21:47.29 | russellb | kevin reads it ... he's a machine |
| 21:47.34 | file | I glance at stuff that may be relevant and interesting |
| 21:47.50 | Qwell | I'm glad we don't have mail quotas :p |
| 21:48.09 | tzanger | mind you I used to be subscribed to lkml too |
| 21:48.13 | tzanger | -users has nothing on that |
| 21:48.13 | seanbright | i read interesting stuff until i read a troll e-mail and just lose faith in humanity |
| 21:48.15 | file | that reminds me... |
| 21:48.22 | seanbright | then i start drinking again. |
| 21:48.35 | file | I'm sad that I can't empty my inbox in Zimbra |
| 21:48.40 | bkruse | seanbright: It is the american cycle |
| 21:48.49 | bkruse | file: I am sad you are using zimbra |
| 21:48.51 | seanbright | bkruse: i'll drink to that |
| 21:49.26 | tzanger | haha |
| 21:51.13 | russellb | i never empty my inbox, heh |
| 21:51.26 | russellb | 7097 messages |
| 21:51.27 | russellb | :-p |
| 21:51.58 | tzanger | pshaw |
| 21:52.04 | file | I am at 619MB usage total, and I emptied all my other folders |
| 21:52.05 | russellb | most of my stuff is filtered though |
| 21:52.08 | tzanger | 33591 messages, 10 unread |
| 21:52.30 | Qwell | Total: 47972 |
| 21:52.31 | russellb | my inbox is bigger than yours! |
| 21:52.33 | tzanger | it'll take a few moments to run du -hs on my Mail folder |
| 21:52.34 | russellb | chuckles to himself |
| 21:52.44 | file | 1.9GB total on my personal mailserver |
| 21:52.50 | russellb | 1.1 GB in use |
| 21:52.56 | seanbright | ok, 4 minutes... and i can't read any more |
| 21:53.01 | seanbright | its a new record |
| 21:53.03 | file | contemplates |
| 21:53.11 | file | if I was Zimbra... how would I allow a user to empty the inbox... |
| 21:53.23 | Qwell | file: empty some other dir |
| 21:53.24 | seanbright | i have 61k in my asterisk-users folder |
| 21:53.27 | Qwell | coughs |
| 21:53.41 | seanbright | 30243 unread messages total |
| 21:53.42 | seanbright | bam |
| 21:53.47 | tzanger | 2.0G |
| 21:53.48 | russellb | Mark as Read. |
| 21:54.00 | file | maybe I can do a mass move... |
| 21:54.10 | russellb | 2165 messages to -users since i last clicked on that folder ... |
| 21:54.14 | tzanger | I occassionally find something interesting in users |
| 21:54.18 | tzanger | but it's not often |
| 21:54.39 | seanbright | heh -> The subject should read "Asterisk 1.4.20-rc2 Now Available" |
| 21:54.45 | seanbright | (from a message to users) |
| 21:55.01 | russellb | gee, thanks |
| 21:55.10 | seanbright | i didn't write it |
| 21:55.37 | seanbright | just found it funny. i'm sure an entire thread will be started based on that. |
| 21:55.51 | seanbright | "the devs don't even RC their own e-mail messages!" |
| 21:56.30 | seanbright | will shut up now |
| 21:57.00 | ctooley | seanbright, editorial mistakes on announcements are not the core competency of the development staff... thankfully. :) |
| 21:59.02 | russellb | what is the core competency? |
| 21:59.12 | Qwell | development, presumably |
| 21:59.15 | russellb | :) |
| 21:59.29 | seanbright | i know. i wasn't trying to come off as a douche. was trying to parody your average -users level email. |
| 21:59.51 | russellb | nods |
| 22:00.11 | ctooley | seanbright, as was I. |
| 22:00.18 | seanbright | *phew* |
| 22:00.21 | ctooley | I apparently forgot the <sarcasm> tags. |
| 22:00.29 | ctooley | :) |
| 22:00.33 | seanbright | heh |
| 22:00.44 | Qwell | IRC is one big <sarcasm> tag |
| 22:01.01 | ctooley | And I'd say, based on my experience being around them at Astricon's, that the core competency is consumption. |
| 22:01.02 | seanbright | noooooooooooooooo |
| 22:01.04 | seanbright | heh |
| 22:01.13 | Qwell | ctooley: consumption? |
| 22:01.18 | Qwell | us? never. |
| 22:02.05 | file | DRINK! |
| 22:02.17 | seanbright | goes home now. |
| 22:02.28 | ctooley | It's 5! |
| 22:02.34 | ctooley | cracks one open. |
| 22:08.56 | Corydon76-dig | file: will you be 21 finally at this year's Astricon? |
| 22:09.04 | file | I'm already 21 |
| 22:09.12 | Corydon76-dig | Well, then yes. |
| 22:09.37 | Corydon76-dig | gets file drunk |
| 22:09.57 | russellb | gets file smashed |
| 22:10.24 | Qwell | smashes file |
| 22:10.36 | file | eep |
| 22:11.19 | Corydon76-dig | If I remember correctly, at Spring VON, on the last day, the waiter carded Qwell and then didn't bother carding file |
| 22:11.39 | Qwell | I remember that |
| 22:11.45 | Qwell | don't remember when/where |
| 22:11.58 | Corydon76-dig | Bottom of the hotel, in that center area |
| 22:12.13 | file | bows |
| 22:12.26 | Corydon76-dig | We were debating eating sushi there or heading out to a burger joint |
| 22:12.28 | Qwell | don't really remember VON. that was forever ago |
| 22:12.44 | Qwell | san jose? |
| 22:12.46 | Corydon76-dig | I believe we left for a burger joint |
| 22:12.47 | Corydon76-dig | Right |
| 22:13.08 | Corydon76-dig | Didn't you drive down with tuxd00d? |
| 22:13.12 | Qwell | up, yeah |
| 22:13.28 | Corydon76-dig | whichever. :-) |
| 22:13.42 | Qwell | that drive was kinda weird :p |
| 22:14.03 | Qwell | I hadn't actually met him yet, at that point. O.o |
| 22:14.11 | Corydon76-dig | Heh |
| 22:14.29 | Qwell | was nice though - tux is cool |
| 22:14.55 | Corydon76-dig | So you got into a car with a strange man? |
| 22:15.13 | file | lol |
| 22:15.18 | Qwell | I wouldn't use those exact words |
| 22:15.23 | Qwell | but yeah, pretty much :p |
| 22:15.37 | ctooley | Off topic: Anyone happen to know what the kernel setting is to release tcp and udp sockets faster? We're killing chan_sip with very short "calls" for routing and running out of file handles. |
| 22:16.17 | Corydon76-dig | ctooley: something like SO_REUSE |
| 22:16.54 | Corydon76-dig | SO_REUSEADDR |
| 22:17.14 | Qwell | Corydon76-dig: any idea what man page those are documented in? |
| 22:17.23 | Corydon76-dig | man 7 socket |
| 22:17.33 | Qwell | ahh |
| 22:17.41 | *** join/#asterisk-dev eurodrew555 (n=andrew@91.84.189.61) |
| 22:17.58 | Corydon76-dig | anything udp-specific would be in 'man 7 udp' |
| 22:18.46 | ctooley | hmmm can I set via a /proc setting or is it something that's set when the socket is actually created? |
| 22:19.00 | Corydon76-dig | I don't know about proc |
| 22:19.06 | Corydon76-dig | Oh, the other one is SO_LINGER |
| 22:19.57 | mvanbaak | http://images.vanbaak.info/thumbnails.php?album=44 |
| 22:20.04 | mvanbaak | totally OT, I know |
| 22:20.09 | mvanbaak | but I dont care ;) |
| 22:20.27 | mvanbaak | I'm the one in the blue-yellow sailing suit |
| 22:20.37 | mvanbaak | with the short hair and blue sunglasses |
| 22:20.40 | ctooley | we're getting more than 100 cps out of chan_sip for routing though. woohoo |
| 22:22.12 | ctooley | SO_LINGER sounds like a good possibility |
| 22:22.12 | codefreeze-lap | file: can you get into swdev? |
| 22:22.43 | russellb | i thought SO_LINGER was only TCP socket related |
| 22:22.52 | russellb | i'm likely confused. |
| 22:23.06 | file | codefreeze-lap: yup... |
| 22:23.07 | *** join/#asterisk-dev jmls_work (n=jmls_wor@mail.tessera.co.uk) |
| 22:23.08 | *** join/#asterisk-dev jmls_home (n=jmls_hom@mail.tessera.co.uk) |
| 22:24.24 | ctooley | russellb, might be. |
| 22:24.32 | Corydon76-dig | ctooley: SO_REUSEADDR is what you use if you want to be able to reuse a port number before the timeout has been reached for possibly delayed packets |
| 22:26.10 | Corydon76-dig | Oh, LINGER is only used to wait synchronously with the socket on a close(2) until it's actually shutdown |
| 22:26.17 | Corydon76-dig | So you don't want SO_LINGER |
| 22:26.25 | mvanbaak | SO_REUSEADDR is something we use in almost all programs |
| 22:26.34 | Corydon76-dig | Generally, yes |
| 22:26.46 | mvanbaak | it gives you the possibility to kill your program, and start it the same second |
| 22:26.46 | Corydon76-dig | because it tends to be safe |
| 22:26.58 | mvanbaak | otherwise you will get 'cannot bind to bla bla' errors |
| 22:26.59 | Corydon76-dig | Right |
| 22:27.30 | mvanbaak | during development, those errors are annoying |
| 22:27.57 | mvanbaak | kill, fix a one line bug, restart |
| 22:28.09 | mvanbaak | most of the time that takes less then the default timeout |
| 22:30.04 | ctooley | The problem we were seeing is that the OS is keeping hold of the udp socket for the SIP communications for too long after the call was gone and we were running out fds. Also, ExternalIVR is opening up 2 sockets for communication and those are hanging around. |
| 22:30.51 | mvanbaak | you can raise the number of allowed fds |
| 22:31.10 | Corydon76-dig | ulimit -n 32768 |
| 22:31.19 | mvanbaak | what Corydon76-dig said |
| 22:31.41 | Corydon76-dig | In 1.6, you can even specify that on the Asterisk CLI and raise the limit while Asterisk is running |
| 22:31.43 | mvanbaak | on a 'real' os you can set it in /etc/login.conf |
| 22:31.59 | ctooley | Yeah, and when I figured out that I had a typo in /etc/security/limits.conf and the ulimit wasn't 65000 it was obvious 1024 wasn't going to work. |
| 22:32.05 | mvanbaak | linux has some pam.d file as well I think |
| 22:32.55 | ctooley | mvanbaak, at least in RHEL /etc/security/limits.conf is where it's set. |
| 22:33.01 | mvanbaak | ah |
| 22:33.31 | mvanbaak | I'm using OpenBSD for services that are exposed to the internet |
| 22:33.57 | mvanbaak | and there you define a login class in /etc/login.conf and put a user in one of those classes |
| 22:34.30 | mvanbaak | both asterisk and postgresql need their own class with raised limits |
| 22:35.53 | mvanbaak | maxproc and openfiles entries |
| 22:55.22 | *** join/#asterisk-dev CoaxD (i=coax@shell1.cornernet.com) |
| 22:56.56 | CoaxD | Okay, what gives? Have a PRI configured 5ESS.. I need to send a facility message (an EECT request) to an Avaya Definity G3... Keep on getting MANDATORY_IE_MISSING cause back whenever i submit ANYTHING... Does anyone happen to have a workign facility request with avaya? |
| 23:04.35 | codefreeze-lap | file: you out there? |
| 23:04.53 | file | I just lost it... |
| 23:05.12 | file | my VPN link is the down |
| 23:05.24 | codefreeze-lap | my vpn flickered, then went out |
| 23:05.38 | codefreeze-lap | about 6 mins ago |
| 23:06.16 | codefreeze-lap | everything else seems ok |
| 23:12.26 | Corydon76-dig | Yeah, mine seems down, too |
| 23:16.03 | *** join/#asterisk-dev jpeeler (n=jpeeler@216.207.245.1) |
| 23:16.33 | *** join/#asterisk-dev angler_ (n=angler@216.207.245.1) |
| 23:16.37 | *** part/#asterisk-dev mog (n=mog@c-68-62-219-86.hsd1.al.comcast.net) |
| 23:17.31 | Corydon76-dig | Oh, that might be why |
| 23:18.07 | file | yup... |
| 23:18.25 | Corydon76-dig | APID was doing maintenance |
| 23:18.52 | Corydon76-dig | Why they were doing maintenance at 6pm on a business day, I'll never understand |
| 23:30.34 | lmadsen | it is not meant to be understood |
| 23:38.53 | *** join/#asterisk-dev bkruse (n=bkruse@216.207.245.1) |
| 23:38.53 | *** mode/#asterisk-dev [+o bkruse] by ChanServ |