| Sunday, July 1st, 2012, 00:00 UTC | ||
| [00:00:17] | gary_buhrmaster (gary_buhrmaster!~gtb@olympus.slac.stanford.edu) has joined #mythtv | |
| [00:02:57] | jams (jams!~jams@cpe-184-58-217-97.wi.res.rr.com) has quit (Read error: Operation timed out) | |
| [00:03:30] | Seeker` (Seeker`!~cjo20@unaffiliated/seeker) has quit (Remote host closed the connection) | |
| [00:08:43] | Seeker` (Seeker`!~cjo20@unaffiliated/seeker) has joined #mythtv | |
| [00:11:53] | knightr_: | beirdo, Hi! Are there currently any multiloc (multi location?) weather pages/maps in MythWeather? |
| [00:12:06] | Beirdo: | no |
| [00:12:20] | Beirdo: | nor will tehre be any time soon :) |
| [00:12:22] | knightr_: | yikes, that's fast service! :-) |
| [00:12:23] | gary_buhrmaster (gary_buhrmaster!~gtb@olympus.slac.stanford.edu) has left #mythtv () | |
| [00:12:36] | Beirdo: | that's a fairly major change |
| [00:13:01] | knightr_: | There is code for that currently in the setup pages, right? |
| [00:13:11] | knightr_: | (or is that for something else?) |
| [00:13:15] | Beirdo: | huh? |
| [00:13:39] | knightr_: | in WeatherSetup.cpp (or something like that) |
| [00:13:48] | Beirdo: | there is support for adding individual weather pages, but for single locations per page |
| [00:14:48] | Beirdo: | oh wonderful |
| [00:14:54] | Beirdo: | fix one bug, cause another |
| [00:16:36] | knightr_: | isn't that part of the game... |
| [00:18:12] | knightr_: | it's that test in weatherSetup.cpp I wondered about: |
| [00:18:14] | knightr_: | if (!si->multiLoc && !si->types.empty()) |
| [00:18:38] | Beirdo: | I dunno what that's supposed to mean |
| [00:18:55] | Beirdo: | unless I'm just totally missing the point :) |
| [00:19:05] | knightr_: | looks like some partial support for not having a location was coded in... |
| [00:20:03] | knightr_: | bah, I'll fake the situation just in case to test my code... |
| [00:20:35] | knightr_: | (I'm modifiying something in how it's translated....) |
| [00:20:48] | knightr_: | Thank you Gavin! |
| [00:21:00] | Beirdo: | no problem :) |
| [00:21:09] | Beirdo: | now if I can get this to unbreak itself |
| [00:21:49] | knightr_: | I'm quite confident you'll find the solution... |
| [00:22:02] | knightr_: | ttyl |
| [00:22:13] | Beirdo: | iao |
| [00:22:16] | Beirdo: | Ciao even |
| [00:22:24] | gary_buhrmaster (gary_buhrmaster!~gtb@olympus.slac.stanford.edu) has joined #mythtv | |
| [00:23:43] | knightr_: | До Ñвидани Ñ! |
| [00:27:39] | voyo (voyo!~voyo@213-134-188-198.home.aster.pl) has joined #mythtv | |
| [00:29:02] | Beirdo: | might have to revert (or fix) that last commit |
| [00:29:18] | Beirdo: | it caused the logserver to go wonky in normal runs... somehow |
| [00:29:33] | Beirdo: | and showed me another bug |
| [00:29:42] | voyo: | hello, Im trying to run mythtv over VAAPI on ATI card, but it wont go... have question – is that supposed to run at all ? |
| [00:30:16] | Beirdo: | oh, interesting |
| [00:30:29] | gary_buhrmaster: | voyo: You want #mythtv-users (the user support channel). This is the developer channel. |
| [00:30:31] | Beirdo: | voyo: wrong channel. See /topic |
| [00:31:29] | Beirdo: | I reverted my change, it STILL is going crazy |
| [00:31:36] | voyo: | Beirdo: yeah I know . I just I wanted answer from developer. If its supposed to work at all . yes/no. |
| [00:31:39] | Beirdo: | OK, what else got changed in the last couple days |
| [00:32:00] | Beirdo: | voyo: then ask in the correct channel, please |
| [00:36:17] | dblain_ (dblain_!~dblain@c-76-127-227-175.hsd1.ma.comcast.net) has joined #mythtv | |
| [00:36:58] | Beirdo: | how in the heck is both the backend and the logsever using over 100% CPU, I wonder? |
| [00:39:33] | pheld (pheld!~heldal@cl-5.osl-01.no.sixxs.net) has quit (Quit: Leaving.) | |
| [00:40:35] | knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Read error: Connection reset by peer) | |
| [00:42:23] | danielk22: | Beirdo: There are 107 mythlogserver instances on my backend.. should I git pull for that commit you just did to fix that? |
| [00:42:38] | Beirdo: | not yet, I think there's an issue in that :) |
| [00:42:46] | Beirdo: | should have it fixed shortly though |
| [00:42:52] | danielk22: | ok |
| [00:42:53] | danielk22: | thx |
| [00:42:57] | Beirdo: | I seemed to have gotten logging into a spinloop |
| [00:43:03] | Beirdo: | and I don't know how :) |
| [00:43:40] | Beirdo: | for now, killall -9 should fix it short-term |
| [00:44:06] | danielk22: | wow, yeah, each one is using 15% of CPU.. |
| [00:44:17] | Beirdo: | 19322 gjhurlbu 20 0 488m 35m 6636 S 229 0.9 5:35.57 mythlogserver |
| [00:44:20] | Beirdo: | not here :) |
| [00:44:25] | Beirdo: | 200%+ |
| [00:44:38] | Beirdo: | it's truly a WTF?! |
| [00:44:51] | danielk22: | killall -9 == 12 respawned quickly. |
| [00:45:01] | jwh: | you linux users enjoying leap second related failure? |
| [00:45:04] | Beirdo: | ahh, you might as well pull |
| [00:45:09] | danielk22: | make that 16 |
| [00:45:27] | danielk22: | err, now 21 |
| [00:45:42] | Beirdo: | yeah, the last last commit keeps that from happening |
| [00:46:04] | Beirdo: | beats me how the heck it's doing that unless somehow now our timers went wonky |
| [00:50:47] | Beirdo: | hmm, I wonder if what I'm seeing isn't in fact a leap-second related thing |
| [00:51:38] | Beirdo: | gonna powercycle the backend box for good measure |
| [00:56:33] | Beirdo: | yeah, that fixed it, it seems |
| [00:56:45] | Beirdo: | jwh: thanks for mentioning that :) |
| [00:56:59] | jwh: | it's pretty hilarious actually |
| [00:57:04] | jwh: | numerous services failing because of it |
| [00:57:10] | Beirdo: | danielk22: seems a power cycle is in order, basically |
| [00:57:11] | jwh: | :D |
| [00:57:22] | Beirdo: | or at least a reboot. mysql was wonky too |
| [00:58:33] | Beirdo: | now to reboot the frontend too |
| [01:04:48] | dekarl (dekarl!~dekarl@p4FE8547E.dip.t-dialin.net) has joined #mythtv | |
| [01:04:50] | Beirdo: | I'll leap-second YOU |
| [01:05:09] | Beirdo: | I am SO glad I'm not in charge of servers at work now |
| [01:05:19] | jwh: | lol |
| [01:05:33] | jwh: | I'm glad I don't run linux anywhere except for dvb duties :P |
| [01:06:48] | dekarl1 (dekarl1!~dekarl@p4FE85352.dip.t-dialin.net) has quit (Ping timeout: 265 seconds) | |
| [01:09:01] | xris (xris!~xris@mythtv/developer/xris) has quit (Quit: Terminated with extreme prejudice - dircproxy 1.2.0) | |
| [01:20:12] | knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv | |
| [01:20:21] | Beirdo: | whatever, one reboot fixed it for me |
| [01:20:37] | J-e-f-f-A (J-e-f-f-A!~J-e-f-f-A@unaffiliated/j-e-f-f-a) has joined #mythtv | |
| [01:22:29] | David_Miller: | leap second! |
| [01:22:33] | David_Miller: | dammnnnnn you! |
| [01:22:49] | David_Miller: | curious: any non ubuntu users have leap second issues? |
| [01:23:12] | jwh: | few reports of centos installs being broken |
| [01:23:36] | Beirdo: | it is likely an eglibc bug |
| [01:23:36] | jwh: | ultimately not everyone runs ntpd though, so it's fairly minimal |
| [01:23:48] | Beirdo: | everyone should be |
| [01:23:53] | jwh: | what |
| [01:23:56] | jwh: | it's a kernel bug |
| [01:23:56] | jwh: | ... |
| [01:24:21] | jwh: | why would glibc induce timing problems, panics and break thread locking? :P |
| [01:24:34] | Beirdo: | because the thread code is all in glibc |
| [01:24:35] | Beirdo: | duh |
| [01:24:47] | jwh: | oh |
| [01:24:57] | Beirdo: | and it didn't cause panics here, just timers constantly firing |
| [01:25:02] | jwh: | I may have to do way with the kernel then if glibc manages cpu scheduling ;) |
| [01:25:24] | Beirdo: | it has nothing to do with cpu scheduling |
| [01:25:29] | jwh: | does it not? |
| [01:25:37] | Beirdo: | quite likely not |
| [01:25:43] | jwh: | is that not what the kernel does when it arbitrates thread locks? |
| [01:25:50] | cocoa117 (cocoa117!~cocoa117@188-222-31-239.zone13.bethere.co.uk) has quit (Quit: Leaving) | |
| [01:25:58] | Beirdo: | thread locks worked fine |
| [01:26:04] | Beirdo: | timers didn't |
| [01:26:08] | jwh: | last I checked the kernel manages the hardware, not the userland interface |
| [01:26:21] | Beirdo: | as in pthread timers |
| [01:26:28] | jwh: | yes, because locking is broken |
| [01:26:33] | jwh: | so threads can't sleep |
| [01:27:03] | Beirdo: | don't think that's accurate |
| [01:27:23] | jwh: | so supply something other than "I'm not sure" or "maybe" |
| [01:27:58] | Beirdo: | more likely, the reference counter used for all the timers took a step in the wrong direction, so every timer would appear to be expired constantly at expiry |
| [01:28:02] | Beirdo: | anyways, whatever |
| [01:28:11] | Beirdo: | I have actual coding to do |
| [01:28:28] | jwh: | almost, but not even close |
| [01:28:29] | Beirdo: | now that I'm sure the bug is NOT in what I committed right when it went sideways |
| [01:28:38] | Beirdo: | I don't care why |
| [01:28:53] | jwh: | the best bit is, it's the logging that causes the panic |
| [01:28:58] | Beirdo: | I have better things to use my brain for right now |
| [01:28:59] | jwh: | not the actual timer problem |
| [01:29:09] | jwh: | ah, limited power? ;) |
| [01:33:45] | Beirdo: | let's see... next victim is db logging |
| [01:33:55] | xris (xris!~xris@xris.forevermore.net) has joined #mythtv | |
| [01:33:56] | xris (xris!~xris@xris.forevermore.net) has quit (Changing host) | |
| [01:33:56] | xris (xris!~xris@mythtv/developer/xris) has joined #mythtv | |
| [01:37:42] | gary_buhrmaster1 (gary_buhrmaster1!~gtb@2001:470:80e4:3::36) has joined #mythtv | |
| [01:49:59] | danielk22: | Beirdo: A restart seems to have made the logserver settle down. I noticed mysqld going bonkers on any other machine. The mysqld one is probably a regression, I'm sure it would have been noticed in 2008 if that bug was in the code then. The zeromq has probably never been through a leap second.. |
| [01:50:22] | jwh: | https://lkml.org/lkml/2012/6/30/122 |
| [01:51:08] | danielk22: | jwh: That page doesn't load, but I'm not running any old kernels... |
| [01:51:23] | jwh: | define old? it's at least in 3.3 |
| [01:51:29] | jwh: | older unaffected |
| [01:51:39] | Beirdo: | danielk22: it's not zeromq, it's Linux |
| [01:52:05] | Beirdo: | I'm running 2.6.34 |
| [01:52:22] | Beirdo: | I'm not convinced it's not libc portion that's primarily borked |
| [01:52:58] | danielk22: | Beirdo: AFAIK versions after 2.6.29 are not affected by the Linux bug. |
| [01:53:30] | Beirdo: | right, but they are by any bugs in glibc or Qt or any other layer |
| [01:54:26] | Beirdo: | and I've seen glibc take out threading stuff before in other interesting ways |
| [01:54:32] | Beirdo: | the kernel was still working fine |
| [01:54:38] | jwh: | heh, lkml is hosed, everyone hitting it because their machines have broken |
| [01:55:48] | danielk22: | speaking of which I should restart the machine I'm IRCing on... mysql and firefox going bonkers.. |
| [01:55:56] | jwh: | yeah |
| [01:55:58] | Beirdo: | seeya soon :) |
| [01:56:06] | jwh: | mysql and seemingly most/all mozilla products are broken |
| [01:56:11] | jwh: | and java, naturally |
| [01:56:21] | danielk22: | Beirdo: You probably are not running glibc.. ubuntu switched after debian switched.. |
| [01:56:27] | Beirdo: | to eglibc |
| [01:56:32] | Beirdo: | yeah, it's crap |
| [01:56:52] | danielk22: | It's a bit buggier... |
| [01:56:56] | Beirdo: | they have undocumented alignment requirements in some places |
| [01:57:12] | danielk22: | I wouldn't call it crap though :) |
| [01:57:13] | Beirdo: | that can really screw threading (like for pthread_mutex) |
| [01:57:14] | J-e-f-f-A (J-e-f-f-A!~J-e-f-f-A@unaffiliated/j-e-f-f-a) has quit (Read error: No route to host) | |
| [01:57:23] | Beirdo: | heh, well, relative crap :) |
| [01:57:37] | Beirdo: | still works mostly |
| [01:57:52] | danielk22 (danielk22!~danielk@96.57.9.142) has quit (Quit: Leaving.) | |
| [01:57:59] | Beirdo: | wouldn't surprise me if this bug were in their implementation of timers in pthreads |
| [01:58:28] | jwh: | if you're not running ntpd and not on an affected kernel and it didn't break at exactly midnight UTC+0, then yes probably |
| [01:58:42] | J-e-f-f-A (J-e-f-f-A!~J-e-f-f-A@unaffiliated/j-e-f-f-a) has joined #mythtv | |
| [01:59:07] | danielk22 (danielk22!~danielk@96.57.9.142) has joined #mythtv | |
| [01:59:26] | Beirdo: | if you're not running ntpd, you should be ashamed of yourself |
| [02:00:06] | danielk22: | Beirdo: I think it must be some common service/library that is busted. For mythlogserver/firefox/mysqld to all be affected. There aren't many things they all use.. |
| [02:00:40] | jwh: | probably more implementation than library specific |
| [02:01:06] | Beirdo: | danielk22: pthreads in glibc :) |
| [02:01:40] | Beirdo: | or usleep() in glibc |
| [02:01:48] | jwh: | I can't even do any diagnosis on my ubuntu box (no ntpd so it hasn't broken), it zeroed the current log and the previous one |
| [02:01:58] | Beirdo: | as that's what went sideways in mythsignalingtimer |
| [02:02:01] | jwh: | probably fired twice |
| [02:02:50] | Beirdo: | the repeated firings of mythlogserver launching was usleep returning immediately and indicating timeout over and over |
| [02:04:00] | jwh: | yeah, in futex()? |
| [02:04:42] | danielk22: | Beirdo: Hmm, usleep() can always return immediately if there is a signal. |
| [02:05:04] | Beirdo: | possible, yeah |
| [02:05:13] | Beirdo: | repeatedly and constanly? |
| [02:05:37] | Beirdo: | sounds like a bug in the system somewhere, but not in our code |
| [02:05:41] | danielk22: | Beirdo: No, it shouldn't do that. |
| [02:07:08] | Beirdo: | agreed :) |
| [02:07:36] | Beirdo: | anyways, I put extra code into the logserver launch so it will only launch ONCE in a one second period |
| [02:07:51] | Beirdo: | although I doubt it will happen again in the near future |
| [02:07:59] | danielk22: | I'll be glad when we can use Qt timers again.. |
| [02:08:10] | danielk22: | Beirdo: It won't happen for at least 6 months :) |
| [02:08:22] | Beirdo: | yeah, hopefully they don't wedge themselves too on the next leap second |
| [02:08:32] | Beirdo: | heh, so true |
| [02:08:43] | jwh: | I don't know why ISC didn't adopt the google method |
| [02:08:48] | jwh: | this would have been entirely avoided |
| [02:08:49] | Beirdo: | my commit timing was so nasty |
| [02:09:38] | danielk22: | jwh ? |
| [02:10:00] | danielk22: | Beirdo: FYI Everything righted itself here after a reboot. |
| [02:10:03] | jwh: | danielk22: google modified their ntpd behaviour to change by a a handful of milliseconds in the run up to midnight |
| [02:10:06] | jwh: | instead of jumping backwards |
| [02:10:29] | Beirdo: | danielk22: did for me too, thankfully |
| [02:11:26] | danielk22: | jwh: Why would ntpd jump backawards? The adjustment was forward one second... glibc supports the 60th second.. |
| [02:11:47] | jwh: | danielk22: forward rather |
| [02:11:50] | jwh: | but the result is the same |
| [02:12:24] | jwh: | 23:59:60 isn't a valid time spec, so what happens to data written during that second |
| [02:12:29] | jwh: | like database entries |
| [02:12:39] | Beirdo: | it most certainly IS valid |
| [02:12:47] | jwh: | (it's not) |
| [02:12:48] | Beirdo: | as is :61 for double leap seconds |
| [02:12:56] | jwh: | glibc might say it is, but it's not a valid time spec in normal operation |
| [02:12:59] | danielk22: | jwh: Well the google method would mean things would be off for time.. |
| [02:13:14] | jwh: | danielk22: only by a fraction of a second, like milliseconds |
| [02:13:16] | Beirdo: | any code that doesn't handle it is not good code |
| [02:13:33] | jwh: | danielk22: but it also means at midnight it will already be :60 |
| [02:13:45] | Beirdo: | if you want to argue that it's not a valid second, talk to the astronomers, etc. |
| [02:13:51] | danielk22: | Anyway this isn't our problem to worry about this time :) |
| [02:14:16] | Beirdo: | yeah, it caused me a mild panic at first, but it's all seemingly good |
| [02:14:32] | jwh: | arguably your code isn't good either because it couldn't handle the leap second ;) |
| [02:14:35] | Beirdo: | and now the logserver should very quickly exit if started when it's already running |
| [02:15:18] | Beirdo: | jwh: you don't know what you are talking about, and now you are getting annoying. Please stop while you are behind. |
| [02:15:26] | jwh: | explain? |
| [02:15:41] | jwh: | only a few pieces of software actually had issues, mostly terrible ones |
| [02:16:35] | Beirdo: | terrible... like...mysql, firefox.... I see |
| [02:16:47] | jwh: | mozilla products, java |
| [02:16:49] | Beirdo: | just let me code in peace, please |
| [02:16:52] | jwh: | mysql isn't all that either |
| [02:18:02] | Beirdo: | this is a channel for development, not for dissing our code, nor that of other projects by non-developers of our code. |
| [02:18:21] | Beirdo: | constructive criticism is good |
| [02:19:03] | jwh: | constructive criticism is usually ignored |
| [02:19:22] | aloril (aloril!~aloril@dsl-tkubrasgw3-fe7ef900-153.dhcp.inet.fi) has quit (Remote host closed the connection) | |
| [02:19:32] | Beirdo: | you are welcome to go look at the code and determine how it should be fixed if you think it's so bad. Otherwise, I stand by my "you don't know what you are talking about" |
| [02:19:45] | danielk22: | Heh, java is one of the best engineered pieces of software on the planet. If you consider that terrible then you don't know what you are talking about. |
| [02:19:58] | jwh: | yeah like I'm going to waste *that* much time |
| [02:20:10] | Beirdo: | danielk22: yeah, but Java apps can be extremely poorly written :) |
| [02:20:31] | jwh: | danielk22: some questionable design decisions, though |
| [02:20:34] | Beirdo: | jwh: then why are you even in the development channel? |
| [02:20:54] | jwh: | Beirdo: so I can see how things are progressing, of course |
| [02:23:38] | Beirdo: | oh yah, before db logging, let's redo the install paths for external library includes |
| [02:23:50] | Beirdo: | I think I even have a ticket for that |
| [02:24:36] | aloril (aloril!~aloril@dsl-tkubrasgw3-fe7ef900-153.dhcp.inet.fi) has joined #mythtv | |
| [02:28:04] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Remote host closed the connection) | |
| [02:32:23] | XDS2010 (XDS2010!users.1218@gateway/web/irccloud.com/x-gmhoqegwnoexavzj) has joined #mythtv | |
| [02:47:44] | Beirdo: | ahhh, 3 tickets closed |
| [02:48:03] | Beirdo: | superm1: I fixed the include/lib stuff you had tickets for. Should look better now |
| [02:56:49] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Remote host closed the connection) | |
| [02:59:24] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv | |
| [03:05:42] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Remote host closed the connection) | |
| [03:11:17] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has joined #mythtv | |
| [03:11:48] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has quit (Changing host) | |
| [03:11:48] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv | |
| [03:14:50] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Remote host closed the connection) | |
| [03:15:51] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has joined #mythtv | |
| [03:22:02] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has quit (Ping timeout: 246 seconds) | |
| [03:29:48] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has joined #mythtv | |
| [03:31:47] | voyo (voyo!~voyo@213-134-188-198.home.aster.pl) has quit (Read error: Connection reset by peer) | |
| [03:35:20] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has quit (Ping timeout: 246 seconds) | |
| [04:02:44] | kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has quit (Quit: kormoc) | |
| [04:40:04] | xris (xris!~xris@mythtv/developer/xris) has quit (Quit: Terminated with extreme prejudice - dircproxy 1.2.0) | |
| [04:42:22] | xris (xris!~xris@xris.forevermore.net) has joined #mythtv | |
| [04:42:22] | xris (xris!~xris@xris.forevermore.net) has quit (Changing host) | |
| [04:42:22] | xris (xris!~xris@mythtv/developer/xris) has joined #mythtv | |
| [04:44:33] | xris (xris!~xris@mythtv/developer/xris) has quit (Client Quit) | |
| [04:53:47] | zombor (zombor!~zombor_@65.29.231.135) has joined #mythtv | |
| [04:53:47] | zombor (zombor!~zombor_@65.29.231.135) has quit (Changing host) | |
| [04:53:47] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [04:55:24] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [04:58:19] | xris (xris!~xris@xris.forevermore.net) has joined #mythtv | |
| [04:58:19] | xris (xris!~xris@xris.forevermore.net) has quit (Changing host) | |
| [04:58:19] | xris (xris!~xris@mythtv/developer/xris) has joined #mythtv | |
| [05:41:44] | Unhelpful (Unhelpful!~quassel@rockbox/developer/Unhelpful) has quit (Ping timeout: 260 seconds) | |
| [05:42:36] | Unhelpful (Unhelpful!~quassel@rockbox/developer/Unhelpful) has joined #mythtv | |
| [06:11:36] | zombor (zombor!~zombor_@65.29.231.135) has joined #mythtv | |
| [06:11:37] | zombor (zombor!~zombor_@65.29.231.135) has quit (Changing host) | |
| [06:11:37] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [06:25:02] | pheld (pheld!~heldal@cl-5.osl-01.no.sixxs.net) has joined #mythtv | |
| [06:28:24] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [06:47:39] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has joined #mythtv | |
| [06:47:39] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has quit (Changing host) | |
| [06:47:39] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv | |
| [06:48:24] | stuartm: | Beirdo: overnight my frontend went mad and started spawning dozens, if not hundreds of mythlogserver processes |
| [06:50:12] | stuartm: | http://pastebin.com/52iMtmGC |
| [06:50:51] | stuartm: | heh my load averages were over 400 |
| [06:56:02] | stuartm: | unable to kill most of them because their handlers are trying to write the last long lines to the database, but mysql is refusing connections because it's overloaded |
| [07:03:02] | Beirdo: | that was due to the leap second |
| [07:03:09] | Beirdo: | reboot will fix it fast |
| [07:03:19] | Beirdo: | many many things went apeshit |
| [07:03:33] | Beirdo: | including mythtv code, mysql, firefox, and on and on |
| [07:04:05] | Beirdo: | and I put in code to make sure that the mass spawning can't happen anymore as well |
| [07:04:16] | Beirdo: | sucked, didn't it? :) |
| [07:05:36] | Beirdo: | caught us all off our guard... |
| [07:08:36] | stuartm: | very sucky :) I'm going to have to reboot, this machine is very sick, too much was pushed into swap :/ |
| [07:11:39] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Remote host closed the connection) | |
| [07:12:59] | Goga777 (Goga777!~Goga777@95-30-19-81.broadband.corbina.ru) has joined #mythtv | |
| [07:13:28] | Goga777 (Goga777!~Goga777@95-30-19-81.broadband.corbina.ru) has quit (Read error: Connection reset by peer) | |
| [07:19:20] | gary_buhrmaster1: | Beirdo: Rather than reboot, just setting the date will (usually) fix the issue (ntpd may be confused, best to stop/start it tpp): # date --set="`date`" |
| [07:25:11] | stoffel (stoffel!~quassel@pD9E42108.dip.t-dialin.net) has joined #mythtv | |
| [07:29:55] | Beirdo: | not if the machine's already hosed |
| [07:30:20] | Beirdo: | and yeah, it might work, but reboot is more guaranteed ;) |
| [07:42:04] | SteveGoodey (SteveGoodey!~steve@host86-160-43-123.range86-160.btcentralplus.com) has joined #mythtv | |
| [07:48:13] | J-e-f-f-A (J-e-f-f-A!~J-e-f-f-A@unaffiliated/j-e-f-f-a) has quit (Read error: No route to host) | |
| [07:48:46] | J-e-f-f-A (J-e-f-f-A!~J-e-f-f-A@unaffiliated/j-e-f-f-a) has joined #mythtv | |
| [07:58:13] | stoffel (stoffel!~quassel@pD9E42108.dip.t-dialin.net) has quit (Read error: Connection reset by peer) | |
| [08:09:33] | J-e-f-f-A (J-e-f-f-A!~J-e-f-f-A@unaffiliated/j-e-f-f-a) has quit (Quit: Oh No!!!! ;-)) | |
| [08:14:22] | J-e-f-f-A (J-e-f-f-A!~J-e-f-f-A@pool-71-184-254-79.bstnma.fios.verizon.net) has joined #mythtv | |
| [08:54:33] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has joined #mythtv | |
| [09:17:09] | amessina (amessina!~amessina@2001:470:1f11:a4:d6be:d9ff:fe8d:7c1e) has joined #mythtv | |
| [10:02:37] | MavT (MavT!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (Ping timeout: 240 seconds) | |
| [10:39:28] | MavT (MavT!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv | |
| [10:41:10] | SteveGoodey (SteveGoodey!~steve@host86-160-43-123.range86-160.btcentralplus.com) has quit (Remote host closed the connection) | |
| [11:35:38] | MaverickTech (MaverickTech!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv | |
| [11:36:31] | MavT (MavT!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (Ping timeout: 244 seconds) | |
| [11:41:08] | dblain_ is now known as dblain | |
| [11:41:10] | dblain (dblain!~dblain@mythtv/developer/dblain) has joined #mythtv | |
| [11:41:10] | dblain (dblain!~dblain@c-76-127-227-175.hsd1.ma.comcast.net) has quit (Changing host) | |
| [11:45:00] | SteveGoodey (SteveGoodey!~steve@host86-160-43-123.range86-160.btcentralplus.com) has joined #mythtv | |
| [12:02:41] | tomimo (tomimo!~kurre@xdsl-83-150-88-111.nebulazone.fi) has quit (Remote host closed the connection) | |
| [12:02:56] | kurre2 (kurre2!~tomimo@83.150.88.111) has quit (Remote host closed the connection) | |
| [12:13:36] | kurre2 (kurre2!~tomimo@xdsl-83-150-88-111.nebulazone.fi) has joined #mythtv | |
| [12:14:05] | tomimo (tomimo!~kurre@xdsl-83-150-88-111.nebulazone.fi) has joined #mythtv | |
| [12:25:12] | SteveGoodey (SteveGoodey!~steve@host86-160-43-123.range86-160.btcentralplus.com) has quit (Remote host closed the connection) | |
| [13:19:52] | amessina (amessina!~amessina@2001:470:1f11:a4:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!) | |
| [13:21:32] | amessina (amessina!~amessina@2001:470:1f11:a4:d6be:d9ff:fe8d:7c1e) has joined #mythtv | |
| [13:38:38] | gnexus (gnexus!~gnexus@92.61.201.66) has joined #mythtv | |
| [13:39:10] | gnexus: | anybody here? |
| [13:40:48] | andreax (andreax!~andreaz@p54BF2114.dip.t-dialin.net) has joined #mythtv | |
| [13:41:00] | rsiebert_ (rsiebert_!~quassel@g225048241.adsl.alicedsl.de) has joined #mythtv | |
| [13:41:31] | rsiebert (rsiebert!~quassel@g225052223.adsl.alicedsl.de) has quit (Ping timeout: 246 seconds) | |
| [13:42:16] | gnexus (gnexus!~gnexus@92.61.201.66) has left #mythtv () | |
| [13:47:36] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [13:49:20] | sraue (sraue!~stephan@xbmc/staff/sraue) has quit (Quit: Ex-Chat) | |
| [14:08:45] | sraue (sraue!~stephan@xbmc/staff/sraue) has joined #mythtv | |
| [14:24:23] | gary_buhrmaster1 (gary_buhrmaster1!~gtb@2001:470:80e4:3::36) has left #mythtv () | |
| [14:41:32] | gigem (gigem!~david@mythtv/developer/gigem) has quit (Quit: WeeChat 0.3.8) | |
| [14:43:01] | gigem (gigem!~david@mythtv/developer/gigem) has joined #mythtv | |
| [14:56:30] | sraue (sraue!~stephan@xbmc/staff/sraue) has quit (Ping timeout: 252 seconds) | |
| [15:09:30] | amessina (amessina!~amessina@2001:470:1f11:a4:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!) | |
| [15:19:24] | sraue (sraue!~stephan@xbmc/staff/sraue) has joined #mythtv | |
| [15:57:04] | amejia (amejia!~andres@xbmc/staff/amejia) has joined #mythtv | |
| [17:04:15] | taylorr (taylorr!~taylorr@cpe-173-095-145-027.nc.res.rr.com) has joined #mythtv | |
| [17:04:15] | taylorr (taylorr!~taylorr@unaffiliated/elmojo) has joined #mythtv | |
| [17:04:15] | taylorr (taylorr!~taylorr@cpe-173-095-145-027.nc.res.rr.com) has quit (Changing host) | |
| [17:11:55] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has joined #mythtv | |
| [17:11:55] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has quit (Changing host) | |
| [17:11:55] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv | |
| [17:13:27] | stuartm: | well that wonky leap second handling really screwed things up, the events last night corrupted some sectors on the disk meaning it never came back up after a reboot and for some reason the keyboard wouldn't work after post so I couldn't get in to do repairs :( |
| [17:13:34] | _Techie_-_AFK_ (_Techie_-_AFK_!~techie@165.33.69.111.dynamic.snap.net.nz) has quit (Ping timeout: 260 seconds) | |
| [17:14:28] | stuartm: | I thoughtlessly threw away every live/repair disc months ago and left myself no way to fix the problem ... still everything seems ok now |
| [17:14:59] | andreax (andreax!~andreaz@p54BF2114.dip.t-dialin.net) has quit (Read error: Connection reset by peer) | |
| [17:15:28] | stuartm: | first thing I do is create a bootable usb thumb drive just for situations such as this |
| [17:16:33] | stuartm: | or I will, just as soon as I can find one ... they are all hiding atm |
| [17:16:45] | _Techie_-_AFK_ (_Techie_-_AFK_!~techie@165.33.69.111.dynamic.snap.net.nz) has joined #mythtv | |
| [17:40:24] | skd5aner (skd5aner!~skd5aner@cpe-071-071-242-134.carolina.res.rr.com) has quit (Ping timeout: 252 seconds) | |
| [17:40:52] | skd5aner (skd5aner!~skd5aner@cpe-071-071-242-134.carolina.res.rr.com) has joined #mythtv | |
| [17:47:55] | andreax (andreax!~andreaz@p54BF2114.dip.t-dialin.net) has joined #mythtv | |
| [18:00:54] | Guyverix (Guyverix!~chubbard@c-50-135-0-96.hsd1.wa.comcast.net) has joined #mythtv | |
| [18:01:27] | Guyverix: | Gavin, you on here today? |
| [18:15:11] | stuartm: | heh, and in all of that it somehow managed to change my timezone to Belfast, 1 hour ahead |
| [18:21:28] | sraue (sraue!~stephan@xbmc/staff/sraue) has quit (Ping timeout: 252 seconds) | |
| [18:34:20] | sraue (sraue!~stephan@xbmc/staff/sraue) has joined #mythtv | |
| [18:41:37] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has quit (Ping timeout: 250 seconds) | |
| [18:44:13] | Guyverix (Guyverix!~chubbard@c-50-135-0-96.hsd1.wa.comcast.net) has left #mythtv () | |
| [18:44:53] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Remote host closed the connection) | |
| [18:51:40] | SteveGoodey (SteveGoodey!~steve@host86-160-43-123.range86-160.btcentralplus.com) has joined #mythtv | |
| [18:54:34] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has joined #mythtv | |
| [18:54:34] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has quit (Changing host) | |
| [18:54:34] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv | |
| [18:57:17] | Guyverix (Guyverix!~chubbard@c-50-135-0-96.hsd1.wa.comcast.net) has joined #mythtv | |
| [19:04:14] | fully_human (fully_human!~fully_hum@50-39-144-123.bvtn.or.frontiernet.net) has joined #mythtv | |
| [19:05:44] | fully_human: | Hello. I am trying to configure mythbackend. However, I'm getting an error that the directories /var/mythtv/whatever//.test are not writable. However, they are, they exist, and just to make sure I did a chmod and chown again (recursively). Any idea why I'm getting this? Thanks. |
| [19:06:10] | Guyverix: | what user did you chmod them to? |
| [19:06:16] | fully_human: | mythtv:mythtv |
| [19:07:22] | Guyverix: | thats chown.. What are they chmodded to? 775 ?? |
| [19:07:52] | fully_human: | Oh, g+wrx. |
| [19:08:43] | stuartm: | o+wrx? |
| [19:09:13] | fully_human: | I set the directory /var/mythtv/ to 775 now and still getting the error. |
| [19:09:38] | fully_human: | I find it odd that there's that double "//" before the directory. |
| [19:09:39] | Guyverix: | try a 776, just for grins and test again |
| [19:09:47] | Guyverix: | 5 is read / execute |
| [19:10:01] | Guyverix: | that is odd too, but will not hurt anything |
| [19:10:05] | fully_human: | Now it's saying none of the paths exist. |
| [19:11:13] | Guyverix: | miss a colon in the path definition? |
| [19:11:35] | fully_human: | No. |
| [19:12:01] | fully_human: | I mean the paths of the mythtv directory (e.g. /var/mythtv/defaults) |
| [19:12:04] | Guyverix: | hrm... |
| [19:12:15] | Guyverix: | well, cheat.. |
| [19:12:21] | fully_human: | But when I changed it back to 775, I got the .test error again. |
| [19:12:32] | Guyverix: | chmod recursive 777, and then chown.. |
| [19:12:38] | Guyverix: | then start taking away |
| [19:13:47] | Guyverix: | also do an id mythtv and make sure that mythtv user really is in the mythtv group (dont know why it would not be though) |
| [19:14:09] | fully_human: | With 777, I'm still getting the same error. |
| [19:14:30] | Guyverix: | okay, now thats weird |
| [19:14:37] | fully_human: | Yup. |
| [19:14:50] | Guyverix: | could your backend be running as your normal userid, and not mythtv? |
| [19:15:10] | fully_human: | Possibly. I was going to try to restart that next. |
| [19:15:19] | Guyverix: | give that a shot |
| [19:15:48] | Guyverix: | I am getting some food wiith my wife, but when I get back I will try to help more if I can in PM if yuo need it. |
| [19:16:04] | fully_human: | Thanks. I may grab a bite to eat, too. |
| [19:23:34] | fully_human (fully_human!~fully_hum@50-39-144-123.bvtn.or.frontiernet.net) has quit (Ping timeout: 244 seconds) | |
| [19:33:18] | stuartm: | ps -Af | grep mythbackend |
| [19:36:47] | Sharky112065: | I know im not supposed to bother developers in this channel, but I think there is a bug in mythfilldatabase in the pull I did yesterday. Can someone help me in the mythtv-users channel? |
| [19:44:35] | Beirdo: | hey, Guyverix... anything break at work due to the leap second? :) |
| [19:44:45] | Beirdo: | well, ex-work for me |
| [19:48:43] | mzanetti (mzanetti!~mzanetti@2a01:4f8:130:8481::2) has left #mythtv ("Konversation terminated!") | |
| [20:01:52] | gary_buhrmaster: | stuartm: While the leap second is all powerful, I doubt it was able to corrupt disk sectors (broken software, on the other hand :-) |
| [20:03:07] | stuartm: | gary_buhrmaster: I never said the leap second was responsible, but mysql, mythlogserver or one of the other processes that went crazy as a result of the switch |
| [20:05:39] | stuartm: | it's funny that the leap second stuff managed to achieve what the y2k 'bug' didn't |
| [20:11:51] | gnexus (gnexus!~gnexus@92.61.201.66) has joined #mythtv | |
| [20:13:32] | gary_buhrmaster: | stuartm: Well, y2k was overhyped, but I do not think it is in the same order of magnitude even for the impacted systems (since this is mostly a "reboot and its over" fix). On the other hand, the year 2038 will be a bigger thing. |
| [20:15:03] | gnexus (gnexus!~gnexus@92.61.201.66) has left #mythtv () | |
| [20:15:07] | stuartm: | it's still a serious enough problem if all mission-critical system and their backups fall over simultaneously :) |
| [20:21:08] | stuartm: | reports are sketchy because it's Sunday and the traditional press are nursing hangovers, but I'd seen mention that outfits such as Netflix and Facebook suffered downtime |
| [20:23:51] | stuartm: | probably any site using mysql, although simply restarting mysql wasn't enough, which suggests the problem was at a lower level ... |
| [20:24:20] | gary_buhrmaster: | There was the Netflix outage on Friday due to the Amazon No. Va. EC2 power outage (apparently the Netflix chaos monkeys are not erratic enough) |
| [20:26:41] | kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has joined #mythtv | |
| [20:29:25] | gary_buhrmaster: | Not all sites. It required specific kernel versions, or specific glibc versions, or specific java versions, or specific app versions..... and it also depended on the load. If the change(s) occurred when no one was looking .... Unfortunately, those versions are widely deployed. |
| [20:32:59] | gary_buhrmaster: | I do agree that since the average communications (propaganda) department in the large corps are still asleep we will not hear until Monday about how they have everything under control. |
| [20:39:46] | gregL (gregL!~greg@cpe-74-76-125-87.nycap.res.rr.com) has quit (Ping timeout: 246 seconds) | |
| [20:40:48] | andreax1 (andreax1!~andreaz@p54BF1C06.dip.t-dialin.net) has joined #mythtv | |
| [20:42:09] | andreax (andreax!~andreaz@p54BF2114.dip.t-dialin.net) has quit (Ping timeout: 246 seconds) | |
| [20:48:28] | amejia (amejia!~andres@xbmc/staff/amejia) has quit (Read error: Connection reset by peer) | |
| [20:48:35] | amejia_ (amejia_!~andres@xbmc/staff/amejia) has joined #mythtv | |
| [20:50:10] | Guyverix: | bak! |
| [20:50:29] | Guyverix: | leap-second errors required a reboot across the board! |
| [20:50:32] | Guyverix: | hehehe |
| [20:54:23] | superm1: | Beirdo: cool, i'll take a look |
| [20:55:14] | Guyverix: | Beirdo: I think I found an issue with the default package, and wanted to confirm before attempting to submit a fix |
| [20:57:15] | amejia (amejia!~andres@xbmc/staff/amejia) has joined #mythtv | |
| [20:57:15] | amejia_ (amejia_!~andres@xbmc/staff/amejia) has quit (Read error: Connection reset by peer) | |
| [20:57:27] | Beirdo: | Guyverix: yeah, I called Kevin yesterday to give him a heads up |
| [20:57:46] | Guyverix: | I was wondering why I was seeing so much mail from him |
| [20:57:57] | Guyverix: | heh, he must have just been sitting waiting for it |
| [20:58:12] | Beirdo: | nah, he got caught by surprise too |
| [20:58:19] | Guyverix: | heh, that guy has got to get away from his desk once in a while |
| [20:58:33] | Beirdo: | but I figured better to know on the weekend than first thing Monday |
| [20:58:45] | Guyverix: | yea |
| [20:58:47] | amejia_ (amejia_!~andres@xbmc/staff/amejia) has joined #mythtv | |
| [20:58:51] | amejia (amejia!~andres@xbmc/staff/amejia) has quit (Client Quit) | |
| [20:58:59] | Guyverix: | it could do weird stuff to VoIP CDR |
| [20:59:03] | Beirdo: | I was thinking "SO glad I don't have to deal with those servers anymore" :) |
| [20:59:07] | Guyverix: | hahaah |
| [20:59:24] | Guyverix: | it looks liike it actually damaged replication too |
| [20:59:32] | Beirdo: | not surprising |
| [20:59:38] | Guyverix: | one of my primaries dropped with insane ouitput |
| [20:59:47] | Beirdo: | mysql went haywire on my mythbackend box |
| [20:59:52] | Guyverix: | but then again bizops was doing normal crap trying to break them |
| [20:59:57] | Beirdo: | so not surprising it would there too |
| [21:00:27] | Beirdo: | anyways, hoping to get more coding done here today... after a nap, I think. |
| [21:00:29] | Guyverix: | ya |
| [21:00:57] | Guyverix: | BTW, had you noticed in the forums that people were having serious performance issues with the default MySQL instance? |
| [21:02:24] | Beirdo: | yeah |
| [21:02:27] | Beirdo: | people are stupid |
| [21:02:30] | Guyverix: | I have a fix for you |
| [21:02:35] | Beirdo: | they won't follow directions |
| [21:02:41] | Beirdo: | cool |
| [21:02:46] | Guyverix: | true, but this time, I dont think it is the end users fault |
| [21:03:03] | Guyverix: | you have no inodb config set for your mysql conf file. |
| [21:04:22] | Beirdo: | that's the end user's fault |
| [21:04:34] | Beirdo: | and the distro's fault |
| [21:04:51] | Beirdo: | mind you, only mythweather uses innodb right now |
| [21:05:43] | Guyverix: | then looks like mythbuntu goofed somethign up |
| [21:06:09] | Guyverix: | it is not defining myisam as a default, and mysql 5.5 uses innodb as the default if unspecified |
| [21:06:44] | Beirdo: | oooh |
| [21:06:48] | Beirdo: | now that WILL suck |
| [21:07:03] | Guyverix: | The file I am sending dropped my indexing of 3 TB to less than 10 seconds |
| [21:07:03] | Beirdo: | and ubuntu STILL doesn't setup a sane default innodb config |
| [21:07:11] | Guyverix: | and sped up the entire system |
| [21:07:15] | Guyverix: | nope |
| [21:07:21] | Guyverix: | it is using the internal defaults only |
| [21:07:36] | Guyverix: | the file has all the settings needed if you want it |
| [21:07:42] | Beirdo: | I've pointed it out before, and people don't listen :) |
| [21:07:43] | Guyverix: | it will work on any default installation |
| [21:08:04] | Guyverix: | grin, people rarely do.. |
| [21:08:06] | Beirdo: | sure, punt it on over. You could put it on the wiki for all if you want |
| [21:08:14] | Guyverix: | ah, ok.. |
| [21:08:37] | Guyverix: | was not sure what the proper way to ddo this was since I ususally am one of the guys whining, not fixing. |
| [21:08:38] | Guyverix: | hahaha |
| [21:09:04] | Guyverix: | you have an xdcc pending, but I will just put it on the wiki |
| [21:09:51] | Beirdo: | heh, sounds good :) |
| [21:10:33] | Defense|Twin (Defense|Twin!~jepz@e177239246.adsl.alicedsl.de) has joined #mythtv | |
| [21:17:37] | Defense|Twin (Defense|Twin!~jepz@e177239246.adsl.alicedsl.de) has quit (Remote host closed the connection) | |
| [21:25:56] | Beirdo: | be back in a bit. nap time |
| [21:26:02] | SteveGoodey (SteveGoodey!~steve@host86-160-43-123.range86-160.btcentralplus.com) has quit (Remote host closed the connection) | |
| [21:39:21] | Guyverix: | Heh, here ya go for the MySQL fix: http://www.mythtv.org/wiki/Tune_MySQL |
| [21:40:36] | sphery: | Guyverix: btw, I have a patch ready to go that sets the session storage_engine to myisam (for now) so that temp tables will use myisam until distros get us good innodb configs |
| [21:41:01] | Guyverix: | grin, there is your good innodb config |
| [21:41:19] | sphery: | Guyverix: also, it sounds like you may have created your DB using MySQL 5.5 and 0.24-fixes or below... With 0.25, we now specify MyISAM engine for all tables (just a couple of temp tables that don't yet have it). |
| [21:41:46] | Guyverix: | huh, interesting... |
| [21:41:58] | sphery: | that said, I'm sure you're fine with an innodb-based DB--even if it needs more care than myisam :) |
| [21:42:00] | Guyverix: | wondering if this is becuase I did an upgrade to Ubuntu 12.04 |
| [21:42:40] | Guyverix: | Heh, yea.. It does need a bit of thinking before implementation... |
| [21:43:44] | Guyverix: | I have been doing a lot of tuning at work for the last few months on innodb (dammit) so at least I knoew this time what I was looking for with the performance bottleneck |
| [21:43:47] | sphery: | if you used MySQL 5.5 to create your initial DB with a version of MythTV before 0.25, you'd have gotten InnoDB tables... It's also possible that distros ship a "blank" database, which may have been made with 5.5/0.24 or whatever, so might use InnoDB |
| [21:44:34] | Guyverix: | I upgraded to 12.04, and then upgraded from .24 to .25.. (although the whole thing has been upgrade after upgrade from .18) |
| [21:44:35] | sphery: | yeah, Bei rdo mentioned who you were--so I'm thinking we just need to rope you into helping us get things set up for making our switch to InnoDB :) |
| [21:44:46] | Guyverix: | no problem at all.. |
| [21:45:18] | Guyverix: | what I put on the wiki will override the defaults that came with the 5.5.24 MySQL instance only in a few areas. |
| [21:45:25] | sphery: | that config is actually a big help--now I have something to point users at when they have DB issues... out of curiosity, do you have DB data on ext4 + barriers? |
| [21:46:15] | Guyverix: | no, I try to keep things simple at home. Just a normal ext3 partition |
| [21:46:47] | sphery: | without barriers? (cat /proc/mounts) |
| [21:47:34] | Guyverix: | no barriers (for now) |
| [21:47:44] | Guyverix: | it is actually on the root partition |
| [21:48:18] | Guyverix: | rootfs / rootfs rw 0 0 |
| [21:48:38] | sphery: | ok, thanks. was just wondering. the ext4 (which enabled barriers by default--and some distros are now doing so with ext3) + innodb was really killing performance for us, so I'm hoping someone will come up with a good plan for that, since I'm not a fan of telling users to disable barriers. I did push a change to use myisam for the mythfilldatabase temp tables because of that |
| [21:48:58] | sphery: | I was hoping you knew some trick/magic :) |
| [21:49:26] | Guyverix: | well another thing that would make life easier for the MyISAM guys would be an override in the myth.cfg file for the temp tables |
| [21:49:48] | sphery: | what override? |
| [21:49:49] | Guyverix: | they do not get so big that an increase in both the cache keys and tmp table space would degrade the host |
| [21:50:23] | Guyverix: | table_cache = 265 for one |
| [21:51:18] | Guyverix: | also the key_buffer_size should be bigger |
| [21:51:19] | sphery: | that in a session variable for connections, I assume (or can it be set on the mythconverg DB alone?) |
| [21:51:34] | sphery: | similar to http://code.mythtv.org/cgit/mythtv/tree/mytht . . . con.cpp#n182 |
| [21:51:36] | Guyverix: | it is a global |
| [21:52:00] | sphery: | so has to be set on the server? if so, we'd probably have to get distros to change it |
| [21:52:20] | Guyverix: | Dont the distros get the myth.cnf from you guys? |
| [21:52:27] | Guyverix: | or did they make that themselves? |
| [21:53:03] | sphery: | we only provide mc.sql – http://code.mythtv.org/cgit/mythtv/tree/mythtv/database/mc.sql |
| [21:53:26] | Guyverix: | I see.. |
| [21:53:45] | Guyverix: | Okay, that explains so many of the defaults for MySQL being used. |
| [21:54:37] | sphery: | still useful info--and if we can document it on the wiki, we can point users there in the meantime, and distro's mythtv packagers will have a reference to use for tickets when considering the changes/trying to get approval from mysql packagers |
| [21:55:10] | Guyverix: | yea. I hope I made it generic enough.. I put some logging paths in my config |
| [21:55:41] | sphery: | Yeah, I think the main problem is that there's never been a documented configuration suggested, so it's all been kind of a "go with defaults unless a specific bug requires a change" type approach |
| [21:56:22] | Guyverix: | That would do it.. Also now that you have connection pools pumping data in faster there will have to be a bit of config work done |
| [21:57:02] | Guyverix: | with the innodb config I dropped in there my 3TB index takes ~3–10 seconds to scan for updates vs 40–50 seconds |
| [21:57:04] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [21:57:59] | Guyverix: | Hopefully MySQL 5.6 will make it out soon. The thread pooling will also speed things up internally for Myth |
| [21:58:48] | sphery: | nice... users on *buntu were finding mythfilldatabase run times of ~40min or so due to innodb temp tables + ext4/barriers, so we really need to find something to help before we can make a switch |
| [21:59:20] | Guyverix: | Well the barriers should be less of an issue than you would expect. |
| [21:59:42] | Guyverix: | the biggest problem is that the defaults for innodb barely allow the engine to run and thats it |
| [22:00:21] | Guyverix: | There is one thing however.. My Memory footprint increased from 300MB resident to 500MB resident.. |
| [22:00:45] | sphery: | that would be nice--if it turns out that a proper config makes it usable, even with barriers, then I wouldn't be afraid of making the switch, since we could just point users to a proper config (even if their distros don't have it) |
| [22:00:46] | Guyverix: | dont know if you guys would be worried about that though.. I would expect most Myth installations to be at a minimum of 1GB RAM |
| [22:01:34] | sphery: | hehe, yeah, some users will be worried about that (those who run master backend + mysql on a pogoplug or similar), but not sure if any devs will lose sleep over it |
| [22:01:44] | Guyverix: | heh |
| [22:02:16] | Guyverix: | actually, I am going to run a mythfill and see what my real perf stats are like since innodb actually has a real performance engine built in |
| [22:02:56] | knightr_: | Beirdo, you might want to look at that shared library problem on the -users ML, I am not sure if that's done on purpose or if there's anything we could do about it.. |
| [22:03:14] | gigem: | stuartm: My proposed fix for the rest of the scheduler/recorder deadlock can be found at http://pastebin.com/vUJpmwdv . Please double check it for me. Note: it also fixes a minor problem with an AddHistory() call. |
| [22:03:17] | sphery: | Guyverix: 0.25-fixes and master now specify myisam for mfdb temp tables... feel free to back that out and try some testing/timing |
| [22:04:10] | sphery: | Guyverix: http://code.mythtv.org/cgit/mythtv/commit/?id . . . 8da31fe83a07 |
| [22:04:35] | Guyverix: | okay, I will see what I can do for some base tuning |
| [22:04:43] | sphery: | thx |
| [22:09:15] | gary_buhrmaster (gary_buhrmaster!~gtb@olympus.slac.stanford.edu) has left #mythtv () | |
| [22:11:10] | Guyverix: | Confirmed with a full mythfilldatabase that the initial fill is about 50% faster... going to work on getting hard numbers for it |
| [22:40:05] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has joined #mythtv | |
| [22:45:42] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [22:57:14] | amejia_ (amejia_!~andres@xbmc/staff/amejia) has quit (Read error: Connection reset by peer) | |
| [22:57:23] | amejia_ (amejia_!~andres@xbmc/staff/amejia) has joined #mythtv | |
| [23:04:38] | sraue (sraue!~stephan@xbmc/staff/sraue) has quit (Remote host closed the connection) | |
| [23:13:48] | sraue (sraue!~stephan@xbmc/staff/sraue) has joined #mythtv | |
| [23:30:57] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has quit (Ping timeout: 265 seconds) | |
| [23:57:52] | sphery: | Guyverix: if you get a chance, testing with --dd-grab-all would be interesting, too (grabs and populates all 14 days in one pull) |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.