Log for #openttdcoop.devzone on 6th May 2015:
Times are UTC Toggle Colours
05:47:32  *** LSky` has joined #openttdcoop.devzone
06:26:22  *** tycoondemon has quit IRC
06:30:45  *** tycoondemon has joined #openttdcoop.devzone
11:32:22  *** Supercheese is now known as Guest4321
11:32:26  *** Supercheese has joined #openttdcoop.devzone
11:37:59  *** Guest4321 has quit IRC
14:00:07  <DevZone> Project xussrset - Trains from Russia build #947-push: SUCCESS in 4 min 13 sec:
14:47:26  *** Alberth has joined #openttdcoop.devzone
15:08:51  <^Spike^> somehow... i need to figure out when ppl push commits alot....
15:14:29  <Taede> does devzone crumble at that time?
15:27:16  <^Spike^> everything we run :)
15:29:13  <^Spike^> including stable/ps/irc/bnx :)
15:29:15  <^Spike^> bnc*
15:29:27  <^Spike^> but it has to be done that's mostly the point :D
15:29:39  <Taede> ouch
15:30:41  <Taede> restart brotx, and then parse logs?
15:32:48  <Taede> or would that worsen the situation
15:38:37  <Taede> unless hg has its own logging
15:40:07  <Alberth> some are using ssh, others are using http(s)
15:47:54  <Alberth> how many projects are automagically built on push?
15:57:39  <Alberth> revisions do have a creation time stamp, but I'd hope that does not get changed on push
16:01:52  *** oskari89 has joined #openttdcoop.devzone
16:22:47  <^Spike^> well not much use http/https atm Alberth cause of the issues we have(/had) with the old system for that
16:23:02  <^Spike^> i think there is 1 project which has an exception and uses rhodecode atm to push to using web
16:23:24  <^Spike^> we want everything to migrate to either rhodecode or the OS fork of it
16:23:35  <^Spike^> but rhodecode native does not have ssh support
16:23:42  <^Spike^> which is a shame cause that's exactly the feature we want
16:24:43  <Alberth> I wonder whether you can use ssh login, but that doesn't differentiate between push and pull probably
16:24:49  <^Spike^> nop
16:25:06  <^Spike^> and considering eints also uses ssh every 5 mins :)
16:25:13  <^Spike^> (yeah i notice that in my logs :D)
16:25:21  <^Spike^> exactly every 5 mins a peak in ssh logins :D
16:25:49  <Alberth> oh, ok :)   I didn't configure the eints setup
16:26:03  <^Spike^> oh well it has to be done either way
16:26:06  <^Spike^> and no time is a good time
16:26:12  <^Spike^> just like every customer with a single setup so :)
16:26:21  <^Spike^> better act like i do against my own customers: Fuck it has to be done :D
16:26:26  <^Spike^> i don't say that to them though :D
16:27:15  <Taede> hehe
16:28:19  <^Spike^> think i'll just give it a go tonight i guess :)
16:48:12  <DevZone> Project 2cc Bus Set build #37-push: STILL FAILING in 8 sec:
16:50:19  <DevZone> Project 2ccts build #462-push: SUCCESS in 1 min 51 sec:
17:27:15  *** gelignite has joined #openttdcoop.devzone
17:29:32  *** frosch123 has joined #openttdcoop.devzone
17:30:02  <Alberth> quak
17:30:29  <frosch123> moin
17:32:30  <frosch123> dis something happen between 14:00 and 17:00 on devzone? :p
17:32:57  <Alberth> I coded (but not tested at all) a change in eints to save / load a project as a set of files (1 project file, and a bunch of language files). It always  loads and saves all files.
17:33:40  <frosch123> that's not supposed to be an answer to my question, right? :)
17:34:09  <Alberth> that's correct, and I wouldn't know, as I wasn't here at that time
17:34:29  <frosch123> eints sent me 15 love mails :p
17:35:01  <Alberth> :)
17:35:09  <frosch123> though it's a somewhat one-sided relation, they are more like love-request mails
17:35:35  <Alberth> it's stalking :p
17:37:55  <Alberth> I guess I have to also stop eints from saving all files that didn't change?
17:38:38  <frosch123> sounds like a next step :)
17:43:39  <Alberth> having all languages in memory isn't a problem?
17:43:57  <frosch123> nope
17:43:59  <Alberth> or I have to implement some lru caching as well
17:44:40  <frosch123> memory usage was: "in memory" < "loading xml" <<<< "saving xml"
17:46:15  <Alberth> so switching to something json as well?
17:46:45  <Alberth> I wonder if it's useful to use a trivial line-based format, and do a straight dump
17:46:57  <frosch123> well, it also took long :)
17:47:06  <frosch123> so, i think saving per-language is good in any case
17:48:07  <Alberth> I added a flag to disable sharing of text, as that's useless within a single language anyway
17:49:26  <Alberth> although I doubt that has any impact in speed :)
18:04:11  <^Spike^> frosch123 define something :D
18:04:18  <^Spike^> i did some updates on some of the vms
18:04:24  <^Spike^> not devzone for some reasons
18:04:31  <^Spike^> or rather almost all vms
18:04:35  <^Spike^> did i break eints? :D
18:04:52  <frosch123> you restarted the psql vm or so
18:04:58  <^Spike^> ehm pgsql did restart yes
18:05:02  <^Spike^> cause of update :)
18:05:09  <^Spike^> eints hated that? :D
18:05:13  <frosch123> and eints lost its connection (i thought we fixed that)
18:05:25  <^Spike^> well this has been a good test and proof that is wasn't? :)
18:05:33  <frosch123> ^Spike^: it only sent me an email whenever someone tried to login :)
18:05:42  <^Spike^> oh :)
18:05:51  <^Spike^> next time i'll add some logins ;)
18:06:06  <frosch123> don't you get them as well?
18:06:09  <frosch123> or only pm?
18:06:09  <^Spike^> ehm
18:06:23  <^Spike^> tbh nop
18:06:28  <^Spike^> could be pm
18:06:34  <^Spike^> but i think he's not been here :D
18:06:49  <^Spike^> and i think he didn't see his ottdc mail considering sometimes alot of mails get sent to sysadmin :D
18:06:51  <frosch123> oh, he checks mails every day
18:07:02  <^Spike^> even the ottdc ones? :D
18:07:09  <frosch123> he will panic on his vacation, and think that devzone is dead, and he has to return immediately
18:07:26  <^Spike^> ah vacation
18:07:28  <^Spike^> i forgot about that :D
18:07:49  <^Spike^> frosch123: what are the odds of eints not restarting after a vm restart? :)
18:07:56  <^Spike^> basicly all other vms i can fix myself :D
18:08:32  <frosch123> somewhen it opened the db connection only once, so whenever pysql exited, eints needed to restart
18:08:43  <^Spike^> ah
18:08:43  <frosch123> though i thought we changed that to attempt a reconnect :p
18:08:50  <^Spike^> hehe
18:08:58  <^Spike^> well i want to update cimarosa aswell aka reboot it etc etc etc
18:09:13  <frosch123> well, rebooting eints vm takes 2 hours
18:09:19  <frosch123> it clones all repos in that case
18:09:24  <frosch123> so, just restarting eints is better :p
18:09:35  <^Spike^> well considering i need to update cimarosa itself :)
18:09:39  <^Spike^> it will be 2 hours :D
18:09:50  <frosch123> :)
18:10:00  <^Spike^> basicly something that has to happen cause well kernel updates and tooling updates i want/need :)
18:10:26  <frosch123> i am sure it is because of the leap second
18:10:56  <^Spike^> ntp handles date/time aka i don't care about leap seconds :D
18:11:38  <frosch123> during the last leap second a lot of kernels crashed
18:11:43  <frosch123> or something like that
18:11:51  <^Spike^> i think that was a few years back or so
18:11:58  <^Spike^> cause last leap second wasn't too long ago
18:12:07  <^Spike^> and i have routers that are up longer then most leap seconds :)
18:12:17  <^Spike^> linux os based that is :)
18:12:24  <frosch123> well, i have no idea how the kernel knows about leap seconds anyway
18:12:31  <frosch123> they are not predcitable
18:12:32  <^Spike^> they prob don't
18:12:42  <frosch123> os, if they are not hard-coded, then they do not know about them
18:12:43  <^Spike^> and ntp are prob like: Oh you're 1 second in front or behind
18:12:45  <^Spike^> let me fix that!
18:12:58  <^Spike^> and i don't know any system getting killed
18:12:59  <^Spike^> oh..
18:13:00  <^Spike^> not true
18:13:06  <^Spike^> i can think of several
18:13:14  <^Spike^> but when ntp fixes date/time meh :)
18:14:12  <Alberth> hmm, maybe that's in the tzdata updates? :)
18:14:19  <^Spike^> big chance :D
18:14:27  <Alberth> I always wonder how that needs so many updates
18:14:40  <^Spike^> cause tzdata changes more then you would realize/think
18:15:01  <^Spike^> i'm amazed when i read an article about something with tzdata and some company being a pita about something license wise
18:15:13  <^Spike^> think that was 1-2 years ago i read it
18:15:40  <frosch123> ah, ntp notifies the kernel that it should insert a second at midnight
18:15:47  <frosch123> so, the trigger is communicated via ntp
18:15:54  <^Spike^> see ntp! :)
18:15:59  <frosch123> but executed by the kernel
18:16:06  <frosch123> i.e. not adjustments afterwards
18:16:11  <Alberth> ^Spike^: when money or power gets into play, stuff gets messy very quickly :(
18:16:17  <^Spike^> yep
18:16:43  <^Spike^> trying to remember when the nightlies build...
18:16:53  <Alberth> at night?  :p
18:16:56  <Rubidium> daily?
18:16:57  <^Spike^> somehow i rather break the server now... compared to in 2-3 hours :D
18:17:13  <^Spike^> you guys are of no help with those answers :D
18:17:38  <Rubidium> well, the openttd nightly runs at 20:00 CE(S)T
18:17:59  <frosch123> ^Spike^: nightlies are no longer "built", it's only a symlnik getting updated
18:18:22  <^Spike^> then pm changed that
18:18:25  <^Spike^> somehow makes sense...
18:18:26  <^Spike^> and me happy :D
18:18:35  <^Spike^> i hate expanding the fs vm the whole time :)
18:18:38  <^Spike^> thnx to V
18:18:39  <^Spike^> ;)
18:18:52  <Alberth> :)
18:19:16  <Alberth> but yeah, newgrfs are exploding lots :)
18:20:35  <frosch123> Alberth: funnily it's the source bundle that causes trouble :p
18:20:52  <^Spike^> yep :)
18:21:31  <^Spike^> oh well time to see if this thing survives a reboot
18:21:41  <^Spike^> almost...
18:21:54  <^Spike^> or else kick support again and again ask them what the *(&$(*#@$ screen says
18:22:01  <^Spike^> and not getting the answer: we kicked it
18:22:23  <^Spike^> case of always make sure you can explain why you did what you had to do
18:22:26  <^Spike^> and not just do it :)
18:23:45  *** ^ekipS^ has joined #openttdcoop.devzone
18:24:03  <frosch123> bye webster
18:24:10  <^Spike^> too soon
18:24:13  <^Spike^> yum update running :)
18:24:39  <frosch123> [20:24] [Notice] -Webster- Error: "bye" is not a valid command. <- it just reacts to its name?
18:24:46  <frosch123> i didn't even use ! or @ :p
18:24:47  <^Spike^> calc Webster 1+1
18:24:51  <^Spike^> nop :)
18:24:55  <^Spike^> Webster calc 1+1
18:24:55  <Webster> ^Spike^: 2
18:24:58  <^Spike^> that it does :)
18:25:03  <^Spike^> else we just do this:
18:25:04  <^Spike^> KenjiE20
18:25:04  <^Spike^> KenjiE20
18:25:05  <^Spike^> KenjiE20
18:25:06  <^Spike^> ;)
18:25:33  <^Spike^> now a case of rebooting..
18:25:35  <^Spike^> almost...
18:25:37  <^Spike^> start ping first
18:26:14  <^Spike^> there going down
18:26:20  <^ekipS^> i think
18:32:04  *** Webster has joined #openttdcoop.devzone
18:32:32  <frosch123> @services op
18:32:46  <frosch123> isn't it op normalls?
18:33:01  <^ekipS^> ehm can check that when i have my own account again :)
18:33:23  <frosch123> isn't this your true self? :p
18:33:35  <^ekipS^> :)
18:34:38  *** SmatZ has joined #openttdcoop.devzone
18:34:40  <^ekipS^> there connecting
18:34:43  *** ^Spike^ has joined #openttdcoop.devzone
18:34:57  <^ekipS^> and other notes to self: Services don't start by themselves......
18:35:11  *** DevZone has joined #openttdcoop.devzone
18:35:38  *** Terkhen has joined #openttdcoop.devzone
18:35:40  <^Spike^> @services op
18:35:44  <^Spike^> Webster hates me
18:35:47  <^Spike^> :)
18:36:08  *** avdg has joined #openttdcoop.devzone
18:36:08  *** Yexo has joined #openttdcoop.devzone
18:36:19  *** planetmaker has joined #openttdcoop.devzone
18:36:38  *** tneo has joined #openttdcoop.devzone
18:37:08  *** Hirundo has joined #openttdcoop.devzone
18:37:28  <^Spike^> @services op
18:37:38  *** Ammler has joined #openttdcoop.devzone
18:37:39  <^Spike^> @whoami
18:37:43  <Webster> ^Spike^: I don't recognize you.
18:37:47  <^Spike^> that explains :)
18:38:08  *** XeryusTC has joined #openttdcoop.devzone
18:38:08  *** V453000 has joined #openttdcoop.devzone
18:39:08  *** ^ekipS^ has quit IRC
18:43:20  *** ChanServ sets mode: +o Webster
18:43:39  <^Spike^> how to know i've restarted the gameservers before
18:43:45  <^Spike^> history
18:43:49  <^Spike^> why i that script called
18:43:51  <^Spike^> less <scriptname>
18:43:53  <^Spike^> ah..
18:43:53  <KenjiE20> @services op
18:43:57  <^Spike^> that restarts the service
18:44:00  <^Spike^> it's op already KenjiE20
18:44:04  <^Spike^> you did it from bots aswell
18:44:05  <^Spike^> that worked
18:44:21  <KenjiE20> Ah
18:44:32  <KenjiE20> Went across all did it?
18:44:47  <^Spike^> not that i saw
18:45:00  <^Spike^> it did around the same time you did the services op <chan> command
18:45:08  <^Spike^> don't know if it has an all option :D
18:45:13  <^Spike^> with chan only does current chan
18:45:33  <KenjiE20> Ahh, it didn't reply so wasn't sure
18:45:42  <^Spike^> :)
18:45:47  <KenjiE20> Weechat smart filter :p
18:46:00  <^Spike^> that or atleast it doesn't say in chan :D
18:46:07  <KenjiE20> Probably hide the op
18:46:07  <^Spike^> cause for bots there was no chan response either :D
18:46:41  * KenjiE20 shrugs
18:47:24  <KenjiE20> Its back hpw it should be therefore eff it
18:47:24  <^Spike^> oh frosch123 see you in 2 hours sort of...? :)
18:47:32  <^Spike^> something with eints? :D
18:47:36  <^Spike^> hehe
18:47:41  <^Spike^> atleast ts3 autostarted :D
18:47:45  <frosch123> :)
18:49:07  * KenjiE20 disappears into the sun once more  :p
18:49:22  <^Spike^> your sun is called gtaV ;)
18:49:45  <KenjiE20> That or Kerbin
18:50:01  <^Spike^> steam says first one :D
18:50:10  <KenjiE20> Yes
18:51:37  <^Spike^> ok why is a syslog daemon hogging all the cpu...
18:51:39  <^Spike^> unles....
18:51:41  <^Spike^> unless*
18:52:01  <^Spike^> nop not what i expected...
18:52:48  <^Spike^> ok now it is what i expected
18:52:50  <^Spike^> frosch123! :)
18:52:52  <^Spike^> btw...
18:53:00  <^Spike^> how much is it you save in the tmp folder?
18:53:08  <^Spike^> all repose?
18:53:10  <frosch123> all repository clones
18:53:10  <^Spike^> repos*
18:53:12  <^Spike^> hmmm
18:53:14  <^Spike^> size?
18:53:17  <^Spike^> roughly
18:53:21  <frosch123> that's why it has to redo them when reboot
18:53:29  <frosch123> 10 gib?
18:53:42  <^Spike^> damn
18:53:53  <^Spike^> was thinking something with /tmp and tmpfs :)
18:53:55  <frosch123> shall i move them to /var? :p
18:54:09  <^Spike^> depends on what you think is better :)
18:54:18  <frosch123> well, it won't have to reclone them then
18:54:19  <^Spike^> having them in a persistent folder does increase startup times? :)
18:54:39  <frosch123> fine, i move them
18:54:47  <^Spike^> you got the space for it
18:54:55  <^Spike^> and it does save you 2 hours of recloning :)
18:55:10  <^Spike^> cause is there a reason to have them in /tmp and reclone them each restart? :)
18:55:32  <^Spike^> now you know how my collegues feel cause i do this all the time: Why are we doing this? ;)
18:55:40  <^Spike^> on design decisions that is :D
18:56:15  <frosch123> it's recoverable data :)
18:57:43  <^Spike^> seems the reboot took roughly 90 seconds considering i missed +- 90 pings
19:00:19  <^Spike^> frosch123 tbh it's that you an pm always rename the screens... cause i would else just close them :D
19:00:57  <frosch123> where should i put the repos?
19:01:05  <frosch123> ~eints or somewhere in /var ?
19:01:09  <^Spike^> pick a folder...
19:01:14  <^Spike^> /opt would also be nice?
19:01:26  <frosch123>  /srv ? :p
19:01:32  <^Spike^> that's the usual: "we don't know where it belongs"-folder
19:01:45  <^Spike^> pick anyone :D
19:11:28  <frosch123> it's now at 2.8G/44G disk usage
19:11:33  <frosch123> let's see what it is after cloning
19:11:47  <frosch123> s/44/50/
19:12:02  <^Spike^> i still got enough in the pool
19:12:10  <^Spike^> and if you didn't fill it up with tmp
19:12:14  <^Spike^> you shouldn't now either
19:12:21  <frosch123> just so you know how big your tmpfs would have needed to be
19:12:33  <^Spike^> prob too big considering the mem we have for it :D
19:17:21  <^Spike^> oh fuck upstart....
19:17:59  <^Spike^> because the pid file still existed it didn't start a service....
19:18:01  <^Spike^> *sigh*
19:18:23  <^Spike^> case of: Screw check to check if something actually runs on that pid that should match!
19:22:24  <^Spike^> atleast the server update did what i hoped it would do...
19:22:34  <^Spike^> fix java/openvz/elasticsearch issues
19:34:39  *** Alberth has left #openttdcoop.devzone
19:48:07  *** LSky` has quit IRC
19:57:15  <^Spike^> i think for now that is enough playing around with ottdc services... will update mailserver at a later time :D
20:04:07  <frosch123> hmm, looks like my 10GiB guess was quire precise
20:04:16  <frosch123> 10.1G
20:05:10  <frosch123> well, 10.3G, of which 7.8G are yeti :p
20:05:53  <^Spike^> hmm not too bad then :)
20:39:53  <DevZone> Project FIRS Industry Replacement Set build #600-push: STILL FAILING in 1 min 6 sec:
21:17:11  *** gelignite has quit IRC
21:21:36  *** frosch123 has quit IRC
21:25:41  *** oskari89 has quit IRC

Powered by YARRSTE version: svn-trunk