November 17, 2014 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >
November 17, 2014 [00:31:57] *** masked has quit IRC[00:37:59] *** masked has joined #project-FiFo[02:31:43] *** trentster has joined #project-FiFo[02:31:43] *** ChanServ sets mode: +o trentster[03:10:27] <masked> hey trentster[03:10:56] <masked> turns out internode some how allocated me a subnet that was already allocated to someone else, thats why it wasn't working![05:11:23] *** marianoguerra has joined #project-FiFo[05:11:24] *** marianoguerra has joined #project-FiFo[05:13:28] *** marianoguerra has quit IRC[06:13:07] <trentster> hey masked[06:13:24] <trentster> aha internode issue revealed[06:18:54] *** TechByTom has quit IRC[06:29:26] *** marianoguerra has joined #project-FiFo[06:53:53] *** marianoguerra has quit IRC[06:56:03] *** marianoguerra has joined #project-FiFo[06:56:03] *** marianoguerra has joined #project-FiFo[07:14:37] *** marianoguerra has quit IRC[07:15:07] *** marianoguerra has joined #project-FiFo[08:59:10] *** marianoguerra has quit IRC[09:05:14] *** alcir has joined #project-FiFo[09:05:28] <Licenser> Morning fifo land![09:05:52] <trentster> Licenser: welcome back!mate - you were missed[09:06:11] <Licenser> :)[10:05:25] <DigiDaz> Morning all![10:05:47] <DigiDaz> Ping trentster[10:06:26] <DigiDaz> My fusionpbx zone ended up failing miserably in production and I have had to switch back to vmware :([10:34:03] <trentster> DigiDaz: howdy - you kidding?[11:37:38] <trentster> DigiDaz: ping[11:57:46] <DigiDaz> trentster: pong pm[13:02:06] *** marianoguerra has joined #project-FiFo[13:41:47] <trentster> DigiDaz: prv[13:48:30] *** marianoguerra has quit IRC[13:49:21] *** marianoguerra has joined #project-FiFo[13:49:22] *** marianoguerra has joined #project-FiFo[14:03:58] <Licenser> tla![14:20:34] <trentster> Licenser: tla?[14:20:47] <Licenser> Three Letter Acronyms[14:59:47] *** olauzon has joined #project-FiFo[16:30:03] *** trentster has quit IRC[16:43:50] *** kwalker has joined #project-FiFo[16:48:53] *** conan_the_destro has joined #project-FiFo[16:55:41] *** kwalker has quit IRC[16:59:16] *** trentster has joined #project-FiFo[16:59:19] *** ChanServ sets mode: +o trentster[17:18:16] *** alcir has quit IRC[17:24:46] *** lfb_ has quit IRC[17:48:29] *** ipalreadytaken has joined #project-FiFo[17:48:29] *** ChanServ sets mode: +o ipalreadytaken[17:49:16] <ipalreadytaken> DigiDaz: you still around?[17:49:38] *** kwalker has joined #project-FiFo[18:10:09] *** kwalker has quit IRC[18:28:26] <DigiDaz> ipalreadytaken: Yep[18:28:37] <ipalreadytaken> can i quiz you about your setup?[18:29:44] <DigiDaz> yes[18:31:01] <ipalreadytaken> are there any network differences between your smartos hypervisor and vmware box?[18:32:42] <DigiDaz> None, in fact the vmware box is currently very much more loaded than the smartos one was[18:33:18] <ipalreadytaken> when you look at the switch ports do you notice any dropped packets or other bad stats?[18:34:15] <DigiDaz> ipalreadytaken: Its something I haven't yet checked but had entered my head[18:35:21] <ipalreadytaken> yea it sounds like it could be a layer one issue. something like the server is having to retransmit packets but you dont notice until it get under load[18:35:27] <DigiDaz> But that said, there is a KVM machine with about 20 clients running on that too, I have had no complaints whatsoever from anyone today[18:36:39] <DigiDaz> ipalreadytaken: I had a chat with trentster today out of channel, I think it load on sqlite personally, trenster thinks it could have been io throttling too[18:37:16] <ipalreadytaken> did that particular box used to run vmware? and if so did you ever notice anything wrong w/ it?[18:37:40] <ipalreadytaken> sqlite is always slow, but if you are running the same thing in vmware it should be better on a zone[18:37:41] <DigiDaz> There wer about maybe 40 users on that box with about 26 queues[18:38:30] <ipalreadytaken> what are the specs of the zone that you where running fusion on?[18:38:55] <DigiDaz> I cannot understand it, still doesn't make any sense[18:39:19] <DigiDaz> Probably a single core with 2GB ram I think, looking back now[18:40:04] <DigiDaz> the single core was an accident it should have been two but there would have been very light cpu load[18:40:22] <DigiDaz> RAM is never a problem[18:41:18] <DigiDaz> At the time these problems where showing I wiuld be surprised if there wher more than 7 concurrent calls.[18:42:23] <DigiDaz> I can easily throw 100 concurrent at it now at 5 calls per second which is a fairly heavy load, heavier than I ever have in production and it will easily handle it[18:44:44] <DigiDaz> ipalreadytaken: Also, AFAIK, the problems were not there on KVM[18:44:49] <DigiDaz> brb[18:49:23] <ipalreadytaken> when you say 100 cucurrent now do you mean on the smart os vmware?[18:55:18] <DigiDaz> ipalreadytaken: On the smartos zone[18:55:39] <ipalreadytaken> so what is different?[18:56:05] <DigiDaz> For me, purely the queuing, I cannot simulate that[18:56:38] <DigiDaz> So you think the sqlite should perform better in a zone than on kvm?[18:56:47] <ipalreadytaken> generally everything should[18:57:48] *** mattronix has quit IRC[18:58:17] <DigiDaz> I'll have to look at some sipp scripting and see if I can replicate it as I definitely want to nail this[18:58:28] <ipalreadytaken> can you openthe sqllite db and see if its valid?[18:58:38] <ipalreadytaken> *not corrupt[18:59:02] *** mattronix has joined #project-FiFo[19:02:14] <ipalreadytaken> the one thing i could think of is that prehaps freeswitch has some sort of bad polling setup where it falls back to a bad method (read db every x seconds) if some linux specific poll method is not advailable. that could starve io/proc[19:17:36] <ipalreadytaken> DigiDaz: if you find a way to test (putting calls in the queue) can you run prstat while the issue occurs and see if there is a particular process that hogs the proc?[19:27:36] <ipalreadytaken> if you have any more insight just leave a note here. I will look at the logs[19:29:17] *** ipalreadytaken has quit IRC[19:30:57] *** fxhp has quit IRC[19:44:34] *** fxhp has joined #project-FiFo[19:46:33] *** marianoguerra has quit IRC[19:51:23] *** mattronix has quit IRC[19:52:32] *** mattronix has joined #project-FiFo[20:57:42] *** benoitc_ is now known as benoitc[20:57:55] *** benoitc has quit IRC[20:57:55] *** benoitc has joined #project-FiFo[23:05:30] *** marianoguerra has joined #project-FiFo[23:05:31] *** marianoguerra has joined #project-FiFo[23:18:38] *** nemith has quit IRC[23:19:08] *** nemith has joined #project-FiFo[23:55:44] *** masked has quit IRC[23:57:49] *** masked has joined #project-FiFo