00:05.28 | *** join/#asterisk akp55 (~akp55@pool-71-115-0-190.rcmdva.fios.verizon.net) |
00:36.57 | *** join/#asterisk CatCow97 (~mine9@c-73-96-109-206.hsd1.or.comcast.net) |
01:16.28 | *** join/#asterisk silverwhitefish (~hidden@47.203.212.155) |
01:52.07 | *** join/#asterisk Enitin (enitin@gateway/vpn/privateinternetaccess/enitin) |
01:52.27 | *** join/#asterisk tsal (~tsal@i59F4D8CC.versanet.de) |
01:54.25 | *** join/#asterisk mr44er1 (~mr44er@dynamic-046-114-003-254.46.114.pool.telefonica.de) |
01:59.08 | *** join/#asterisk akp55 (~akp55@pool-71-115-0-190.rcmdva.fios.verizon.net) |
03:03.11 | *** join/#asterisk infobot (ibot@96-86-209-99-static.hfc.comcastbusiness.net) |
03:03.11 | *** topic/#asterisk is #asterisk The Open Source PBX and Telephony Platform (asterisk.org) -=- LTS: 18.4.0, 16.18.0 (2021/05/06) Final Bugfix: 13.38.2, 17.9.3 (2021/03/04); DAHDI: 3.0.0 (2018/11/15); libpri 1.6.0 (2017/01/27) -=- Wiki: wiki.asterisk.org -=- Code of Conduct: bit.ly/1hH6P22 |
03:44.36 | *** join/#asterisk ghoti (~paul@dsl-rb-64-118-20-57.wtccommunications.ca) |
05:07.08 | *** join/#asterisk Ner0Zer0 (~Ner0Zer0@87.253.63.54) |
05:12.21 | *** join/#asterisk pppingme (~pppingme@unaffiliated/pppingme) |
06:19.10 | *** join/#asterisk akp55 (~akp55@pool-71-115-0-190.rcmdva.fios.verizon.net) |
06:21.50 | *** join/#asterisk Ravenheart (Ravenheart@95-43-79-254.ip.btc-net.bg) |
06:38.01 | *** join/#asterisk lankanmon (~LKNnet@cpeb4fbe4e331bd-cm9050cadd5190.cpe.net.cable.rogers.com) |
06:52.36 | *** join/#asterisk thansen (~thansen@192.74.130.86) |
10:21.04 | *** join/#asterisk akp55 (~akp55@pool-71-115-0-190.rcmdva.fios.verizon.net) |
11:26.56 | *** join/#asterisk tsal (~tsal@i59F4A05F.versanet.de) |
11:35.13 | *** join/#asterisk HannaM (~quassel@p54849510.dip0.t-ipconnect.de) |
11:43.40 | *** join/#asterisk sa02irc (~mbax@155-079-043-212.ip-addr.inexio.net) |
12:00.52 | *** join/#asterisk andrewyager (~andrewyag@114.141.97.1) |
12:51.15 | *** part/#asterisk ldm (~ldm@hacksoc/member) |
13:27.25 | *** join/#asterisk drathir_tor (~drathir@gateway/tor-sasl/drathir) |
13:33.24 | *** join/#asterisk mvanbaak (~mvanbaak@asterisk/contributor-and-bug-marshal/mvanbaak) |
13:46.59 | *** join/#asterisk Bullit (~Bullit01@042-236-158-163.dynamic.caiway.nl) |
14:06.38 | *** join/#asterisk CatCow97 (~mine9@c-73-96-109-206.hsd1.or.comcast.net) |
14:44.04 | *** join/#asterisk rpifan_ (~rpifan@p200300d2670b9500bf2835dac112de5e.dip0.t-ipconnect.de) |
15:23.43 | *** join/#asterisk jayjo (~jayjo@unaffiliated/jayjo) |
15:42.45 | *** join/#asterisk akp55 (~akp55@pool-71-115-0-190.rcmdva.fios.verizon.net) |
15:52.00 | *** join/#asterisk lankanmon (~LKNnet@cpeb4fbe4e331bd-cm9050cadd5190.cpe.net.cable.rogers.com) |
16:03.36 | *** join/#asterisk lankanmon (~LKNnet@cpeb4fbe4e331bd-cm9050cadd5190.cpe.net.cable.rogers.com) |
16:13.32 | *** join/#asterisk chitholian (~chitholia@2400:ca03:f065:c0b0::70b) |
16:14.30 | chitholian | hello |
16:26.33 | *** join/#asterisk lankanmon (~LKNnet@cpeb4fbe4e331bd-cm9050cadd5190.cpe.net.cable.rogers.com) |
16:35.03 | *** join/#asterisk _0x5eb_ (~seb@69.63.121.78.rev.sfr.net) |
16:40.32 | *** join/#asterisk Jesterboxboy (~Thunderbi@84-115-150-8.cable.dynamic.surfer.at) |
16:41.37 | chitholian | How can I handle 600 concurrent calls with 22.8GHz (6 x 3.8GHz) server (12 vCPUs) with g729 pass-thru mode ? Currently I am running 100 calls (200 channels) with PJSIP Asterisk 18.2.0, It consumes 35% CPU (7.9GHz out of 22.8GHz). |
16:42.45 | Bullit | Over the last couple of W we have heard so much about you astrix and I cant make you rotate your inability to have 100-200 contained transistorvibrationsensors seems to be related to previous topics with 220+ reactions |
16:43.22 | sibiria | chitholian: then you are clearly transcoding |
16:43.27 | chitholian | I need to create at least 4 VPS in the 3.8 GHz 6 CPU Xeon machine (VMWare ESXi) each running at least 150 concurrent calls. |
16:43.41 | sibiria | with just passthrough even a Raspberry Pi can manage over 100 calls |
16:44.34 | Bullit | Noble al White |
16:44.40 | chitholian | But how can I know if asterisk is transcoding ? I only allowed g729 and g723 and all INVITE have both codecs ? |
16:44.42 | Bullit | Maersk Resolute |
16:44.48 | sibiria | my test platform is a Pine64 with 1GB RAM - roughly as powerful as the old RPi 3 - and it can manage about 150 calls without transcoding channels |
16:45.31 | Bullit | TechTalk Scatter OR Alice and Bob |
16:45.40 | Bullit | Vos Tracker |
16:45.42 | igcewieling | pjsip show channelstats should show you the codecs of the channels at least |
16:45.43 | Bullit | Vos Trapper |
16:45.47 | Bullit | Vos Trader |
16:46.02 | sibiria | why are you using g.729 and g.723 by the way? |
16:46.07 | igcewieling | Hmm. we appear to have a troll. |
16:46.14 | sibiria | are you terribly bandwidth-constrained? |
16:46.17 | Bullit | is that the standard for the amount of electrical interference over coms |
16:46.24 | Bullit | what length are you expecting |
16:47.52 | Bullit | asteri><? respond asterisk* we don´t have this problem in the relation between the amount of consultants to a job in the harbor but the problem in satelite management teams kicking off before actually starting |
16:48.23 | chitholian | sibiria: Using g729 and g723 are used becaused only these are allowed in my termination (outgoing) trunks. |
16:49.41 | chitholian | "pjsip show channelstats" shows all channel using g729 |
16:51.39 | sibiria | really, only g729 and g723... seems like an odd VoIP provider |
16:51.54 | sibiria | both of those codecs are, well, terrible |
16:52.38 | chitholian | But, does codec matters in pass-thru mode ? |
16:53.47 | sibiria | if you do nothing but pass audio frames between two ends then no |
16:53.52 | sibiria | but that's not what's happening here |
16:54.03 | chitholian | Here is my dialplan. |
16:54.08 | chitholian | [general] |
16:54.09 | chitholian | static=yes |
16:54.11 | chitholian | writeprotect=no |
16:54.12 | chitholian | autofallthrough=yes |
16:54.13 | chitholian | clearglobalvars=no |
16:54.15 | chitholian | [globals] |
16:54.17 | chitholian | [dialer] |
16:54.18 | chitholian | exten => _X.,1,Goto(f1context,${EXTEN},1) |
16:54.19 | sibiria | no... |
16:54.20 | chitholian | [f1context] |
16:54.21 | chitholian | exten => _X.,1,NoOp(${CALLER_USERNAME} from ${CHANNEL(pjsip,remote_addr)}) |
16:54.22 | chitholian | <PROTECTED> |
16:54.24 | chitholian | <PROTECTED> |
16:54.26 | chitholian | <PROTECTED> |
16:54.27 | chitholian | <PROTECTED> |
16:54.28 | chitholian | <PROTECTED> |
16:54.30 | chitholian | <PROTECTED> |
16:54.32 | chitholian | exten => h,1,AGI(agi://127.0.0.1/cdr.php,${CDR(uniqueid)}) |
16:54.33 | chitholian | [answer] |
16:54.34 | chitholian | exten => s,1,Set(theCallID=${ARG1}) |
16:54.36 | chitholian | <PROTECTED> |
16:54.38 | chitholian | <PROTECTED> |
16:57.32 | chitholian | No other process (mysql, fast-agi) consumes too much CPU as asterisk. |
16:58.09 | sibiria | inspect the SDP of both ends when establishing a call, and enable full logging with debug info, to see if asterisk is letting you know something about the transcoding |
16:58.54 | chitholian | Condition goes worse when call per second increases (e.g. 8-10 CPS) although most of the calls get CONGESTION dial status. Only around 100 calls are active at a time. |
17:00.28 | sibiria | for fastagi, are you using some sort of "slow agi" wrapper? |
17:00.35 | sibiria | because PHP is incredibly slow at starting up |
17:01.04 | sibiria | how is the fastagi listener invoking those PHP scripts? |
17:05.14 | chitholian | With xientd |
17:05.17 | chitholian | service fastagi |
17:05.18 | chitholian | { |
17:05.20 | chitholian | <PROTECTED> |
17:05.21 | chitholian | <PROTECTED> |
17:05.23 | chitholian | <PROTECTED> |
17:05.25 | chitholian | <PROTECTED> |
17:05.26 | chitholian | <PROTECTED> |
17:05.28 | chitholian | <PROTECTED> |
17:05.29 | chitholian | <PROTECTED> |
17:05.30 | chitholian | <PROTECTED> |
17:05.32 | chitholian | <PROTECTED> |
17:05.33 | chitholian | <PROTECTED> |
17:05.35 | chitholian | <PROTECTED> |
17:05.36 | chitholian | } |
17:07.39 | chitholian | If AGI (php program) is slow will the slowness go to asterisk ? Shouldn't it go to xinetd or PHP ? |
17:10.53 | chitholian | If asterisk is consuming more CPU for blocking FastAGI will AMI (async AGI) or ARI client solve the problem ? |
17:18.36 | sibiria | proper fastagi would solve the problem - or slowagi with a faster interpreter like lua or perl |
17:18.56 | sibiria | php, just like python, takes a very long time to start up. they're bad choices for slowagi applications |
17:19.28 | sibiria | but i'd start with looking at the SDP of all the call legs first, and debug info from the log |
17:28.22 | *** join/#asterisk chithol97 (~Android@2400:ca03:f065:c0b0:dd0a:c803:ceb2:ad6f) |
17:30.28 | chithol97 | Is there any difference (as CPU usage) Between FastAGI and Async AGI? |
18:06.08 | *** join/#asterisk sinaowolabi (~Sina@102.134.114.1) |
18:18.17 | *** join/#asterisk rpifan (~rpifan@p200300d2670b950003e715f232381e72.dip0.t-ipconnect.de) |
18:20.21 | *** join/#asterisk silverwhitefish (~hidden@47.203.212.155) |
18:24.41 | *** join/#asterisk InterLinked (~ambassado@cpe-24-209-155-151.wi.res.rr.com) |
18:29.36 | InterLinked | Is there a debugging method that allows tracing the code up to/through a crash in an easily readable way? I dumped the core but when I open up core-full.txt, I don't even see any mention of the function file that caused the crash. I'm getting constantly changing stuff, like smallbin double linked list corrupted, then a seg fault next time, and so on. Thanks! |
18:30.16 | InterLinked | I have some ast_logs in my code to try to follow along with print statements, but that isn't getting me terribly far. |
18:48.18 | *** join/#asterisk sinaowolabi (~Sina@160.152.55.235) |
19:29.36 | *** join/#asterisk akp55 (~akp55@pool-71-115-0-190.rcmdva.fios.verizon.net) |
20:20.25 | *** join/#asterisk sa02irc (~mbax@155-079-043-212.ip-addr.inexio.net) |
20:47.14 | *** join/#asterisk DodgeThis (~DodgeThis@246.102.90.149.rev.vodafone.pt) |
21:22.45 | *** join/#asterisk infobot (ibot@96-86-209-99-static.hfc.comcastbusiness.net) |
21:22.45 | *** topic/#asterisk is #asterisk The Open Source PBX and Telephony Platform (asterisk.org) -=- LTS: 18.4.0, 16.18.0 (2021/05/06) Final Bugfix: 13.38.2, 17.9.3 (2021/03/04); DAHDI: 3.0.0 (2018/11/15); libpri 1.6.0 (2017/01/27) -=- Wiki: wiki.asterisk.org -=- Code of Conduct: bit.ly/1hH6P22 |
22:00.37 | *** join/#asterisk sinaowolabi (~Sina@160.152.55.235) |
22:01.43 | *** join/#asterisk andrewya_ (~andrewyag@114.141.97.1) |
22:09.10 | *** join/#asterisk Typhon (~Typhon@dslb-088-067-143-148.088.067.pools.vodafone-ip.de) |
22:40.46 | *** join/#asterisk markong (~markong@37.120.141.114) |
23:31.27 | *** join/#asterisk akp55 (~akp55@pool-71-115-0-190.rcmdva.fios.verizon.net) |
23:38.47 | *** join/#asterisk pchero (~pchero@211.178.226.108) |