00:03 | sweetpi has quit IRC | |
00:11 | M4gic5t0rM has joined #ltsp | |
00:45 | chris143 has joined #ltsp | |
00:47 | <chris143> having problems with client audio on ubuntu 10.04 ltsp-standalone-server.. SOUND=True, but no sound... verified the client audio capabilities already using 8.04 ltsp server that works with audio.. fresh install of 10.04 not an upgrade.. any ideas?????
| |
00:50 | installed 10.04 alternate CD but did not choose LTSP at install.. later I installed ltsp-standalone-server... all other client features work (localdev, video etc.) but no sound.. where should I look?
| |
01:01 | having problems with client audio on ubuntu 10.04 ltsp-standalone-server.. SOUND=True, but no sound... verified the client audio capabilities already using 8.04 ltsp server that works with audio.. fresh install of 10.04 not an upgrade.. any ideas?????
| |
01:22 | _UsUrPeR_ has quit IRC | |
01:23 | intelliant has joined #ltsp | |
01:29 | _UsUrPeR_ has joined #ltsp | |
01:31 | alkisg has joined #ltsp | |
01:34 | <alkisg> nutron: I think yes, on recent ltsp versions you can use wildcards on mac addresses in lts.conf. But why do you need a different setting for them? (also, distro/version?)
| |
01:38 | nutron: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=550989
| |
01:39 | <intelliant> hi!
| |
01:39 | anyone really experienced with ubuntu ltsp-cluster setup using 10.04 ?
| |
01:40 | with auth against non-ubuntu hosted openLDAP and /home over NFS ?
| |
01:48 | <johnny> chris143, that's enough, you don't have to say it that many times
| |
01:48 | did you try any other clients? perhaps something with different hardware?
| |
02:02 | dobber has joined #ltsp | |
02:14 | <chris143> i tried another client that is workign well on ltsp with 8.04 server
| |
02:14 | and that client had no audio
| |
02:15 | i also tried the (new) hw client on the old 8.04 ltsp server and the new client's audio works great there...
| |
02:16 | so there is something wrong with the new 10.04 ltsp server...
| |
02:20 | btw, sound does work when logged in locally on the ltsp server
| |
02:25 | <alkisg> chris143: it might be either a module problem for that specific client (even if it was working in 8.04), that's why johnny told you to try with another client.
| |
02:27 | <chris143> i also tried a 3rd completely different hw client (different chip maker) abd it too works on the old one and not on the new 10.04 server
| |
02:27 | it seems all machines (client) work on the old 8.04 ltsp server and none work on the new 10.04 ltsp server
| |
02:28 | interesting local apps are not working either... but local devices are (flash drive) with no problems...
| |
02:30 | i have a shell accessible on the client machine
| |
02:30 | i tried an lsmod and found many snd_* modules loaded (when running the image/connected to the new 10.04 ltsp server
| |
02:32 | poff has joined #ltsp | |
02:33 | <alkisg> chris143: local apps are not working? with ltsp-localapps xterm, you don't get an xterm?
| |
02:34 | F-GT has quit IRC | |
02:34 | <chris143> nope
| |
02:34 | LOCAL_APPS=True in lts.conf
| |
02:34 | is set
| |
02:35 | <alkisg> You don't need that, it's by default
| |
02:35 | <chris143> i tried looking in some logs on the client machine.. scrolled to the bottom and heard a beeping noise! (in vi)
| |
02:35 | now this was on the client in a shell (console)
| |
02:35 | so no sound through the ltsp server though... anyway it show that the local client kernel has configured sound...
| |
02:38 | <alkisg> There's sound card "beep" and there's speaker "beep", if it was the latter then that doesn't mean that sound is configured properly
| |
02:38 | But, pulseaudio, localapps etc, that's a lot of problems, maybe ltsp-build-client didn't finish correctly?
| |
02:38 | <chris143> sound card beep through external speakers plugged in to the green back panel mini jack
| |
02:39 | F-GT has joined #ltsp | |
02:39 | <chris143> it didn't report any errors...
| |
02:39 | i could re-try
| |
02:39 | i suppose
| |
02:40 | <alkisg> Is pulseaudio running on the client?
| |
02:40 | <chris143> i installed ltsp via ltsp-standalone-server package post ubu install
| |
02:41 | there is a process id 3750 called pulse after a ps aux | grep pulse.. so i believe so...
| |
02:41 | the process has a bunch of arguments
| |
02:45 | bobby_C has joined #ltsp | |
02:46 | <alkisg> Can you also post (or completely remove) your lts.conf, it case it complicates things?
| |
02:46 | <chris143> i have some progress!
| |
02:46 | sound!
| |
02:47 | <alkisg> How?
| |
02:47 | <chris143> i saved (like a good buoy) and went back to the original i386.img file
| |
02:47 | boy
| |
02:47 | anyway... i loaded that image via tftp
| |
02:47 | and i have sound
| |
02:47 | the only thing i has changed in the chroot env
| |
02:47 | <alkisg> i386.img is loaded with nbd, but anyway, that's good progress
| |
02:48 | <chris143> was adding x11vnc and a custom ldm theme
| |
02:48 | the theme works... in the newer image
| |
02:48 | could x11vnc cause a problem?
| |
02:48 | <alkisg> It shouldn't
| |
02:48 | <chris143> i has even booted the client wihtout running the vnc on it
| |
02:49 | hmmm.
| |
02:49 | local apps work too :-)
| |
02:49 | xterm
| |
02:49 | oh.. i also set a root password in the chroot
| |
02:49 | on the newer image
| |
02:51 | although that may not be necessary.. as i can get a shell on the original image
| |
02:52 | irule has quit IRC | |
03:00 | FGXR6 has joined #ltsp | |
03:07 | biomorph has joined #ltsp | |
03:11 | otavio has quit IRC | |
03:38 | alkisg has quit IRC | |
03:39 | alkisg has joined #ltsp | |
03:46 | alkisg has quit IRC | |
03:55 | <intelliant> for ubuntu ltsp-cluster I need not authenticate the ltsp-root against LDAP, right?
| |
03:58 | <Appiah> no you dont have to
| |
03:58 | you can setup local users if you want
| |
03:58 | bobby_C has quit IRC | |
04:12 | bobby_C has joined #ltsp | |
04:18 | alkisg has joined #ltsp | |
04:30 | bobby_C has quit IRC | |
04:40 | biomorph has quit IRC | |
05:15 | bobby_C has joined #ltsp | |
05:20 | bobby_C has quit IRC | |
05:33 | M4gic5t0rM has quit IRC | |
05:41 | vagrantc has quit IRC | |
06:02 | bobby_C has joined #ltsp | |
06:16 | ogra_cmpc has quit IRC | |
06:17 | chris143 has quit IRC | |
06:20 | mae_tae has joined #ltsp | |
06:23 | <mae_tae> alkisg, regarding the thin client, i got TFTP Error - File Not Found, how do i resolve that? This is the one being displayed on client side PXE-TO1 : File Not Found
| |
06:23 | PXE-E3D : TFTP Error - File Not Found
| |
06:24 | tftp-hda was already installed, when i type this tftp localhost, actually i can connect but when i type get pxelinux.0 it says transfer time out
| |
06:28 | otavio has joined #ltsp | |
06:28 | otavio has joined #ltsp | |
06:36 | wwx has quit IRC | |
06:40 | wwx has joined #ltsp | |
06:41 | <alkisg> mae_tae: on the ltsp server, try: sudo status tftpd-hpa
| |
06:41 | wwx has quit IRC | |
06:41 | wwx has joined #ltsp | |
06:41 | <alkisg> And, sudo netstat -nap|grep :69
| |
06:44 | <mae_tae> alkisg, actually when i type this tftp localhost, i can get inside as in tftp> something like that, but when i type there get pxelinux.0, seems that file is not located there
| |
06:47 | alkisg, where should that file be exactly be placed (pxelinux.0)? mine is i place it in /var/lib/tftpboot
| |
06:50 | <intelliant> Appiah: doing the entire installation from scratch
| |
06:50 | currently the ltsp-build-client is under execution
| |
06:50 | hoping that the ldap auth goes through fine
| |
06:50 | sometime to it as yet
| |
06:54 | <mae_tae> alkisg, when i type this netstat -a |grep tftp <--- its really there
| |
07:02 | pmatulis has joined #ltsp | |
07:11 | M4gic5t0rM has joined #ltsp | |
07:16 | M4gic5t0rM has quit IRC | |
07:18 | <mae_tae> alkisg, when i type ls in tftp> ls, it seems that ls is a bad command, nothing displays in there
| |
07:29 | [GuS] has joined #ltsp | |
07:57 | alkisg has quit IRC | |
08:14 | biomorph has joined #ltsp | |
08:19 | biomorph has quit IRC | |
08:25 | biomorph has joined #ltsp | |
08:33 | mae_tae has quit IRC | |
08:56 | cyberorg has quit IRC | |
08:58 | ogra_cmpc has joined #ltsp | |
08:58 | cyberorg has joined #ltsp | |
09:01 | alkisg has joined #ltsp | |
09:02 | <intelliant> The control-center.py file is missing required for ltsp-cluster
| |
09:02 | wget http://bazaar.launchpad.net/%7Eltsp-cluster-team/ltsp-cluster/ltsp-cluster-control\/download/head%3A/controlcenter.py-20090118065910-j5inpmeqapsuuepd-3/control-center.py
| |
09:02 | any new location known?
| |
09:07 | <cyberorg> intelliant, on suse it is installed with ltsp-cluster-control package /srv/www/ltspc/scripts/control-center.py
| |
09:08 | <intelliant> cyberorg: got an alternate location
| |
09:08 | http://bazaar.launchpad.net/%7Eltsp-cluster-team/ltsp-cluster/ltsp-cluster-control/download/head%3A/controlcenter.py-20090118065910-j5inpmeqapsuuepd-3/control-center.py
| |
09:09 | no difference though :)
| |
09:09 | it justworked
| |
09:09 | making the last attempt with ubuntu :)
| |
09:11 | <[GuS]> Hi guys, did someone made work the SOUND with pulseaudio in thin clients using Gentoo+LTSP5?
| |
09:13 | intelliant has quit IRC | |
09:19 | biomorph has quit IRC | |
09:23 | alkisg has quit IRC | |
09:30 | alkisg has joined #ltsp | |
09:31 | mikkel has joined #ltsp | |
09:34 | garymc has joined #ltsp | |
09:34 | <garymc> Hi guys, im looking at a new upto date server. (Well a newer one ) that can run a 64bit version of LTSP
| |
09:35 | Is the 64bit software any better than the 32bit and what are the differnces?
| |
09:35 | and what is the package that is a self install, with ubuntu on it?
| |
09:38 | Gadi has joined #ltsp | |
09:40 | Gadi has quit IRC | |
09:42 | [GuS] has quit IRC | |
09:42 | |GuS| has joined #ltsp | |
09:43 | Gadi has joined #ltsp | |
09:43 | poff has quit IRC | |
09:56 | |GuS| is now known as [GuS] | |
09:57 | <[GuS]> Guys, did someone made work the SOUND with pulseaudio in thin clients using Gentoo+LTSP5?
| |
10:04 | <alexqwesa> [GuS]: check PULSE_SERVER variable and check what sound not muted
| |
10:08 | <[GuS]> alexqwesa: let me check (sound is not muted btw, and PULSE returns the thin client ip and the pulseaudio port, i must check PULSE_SERVER)
| |
10:10 | * [GuS] booting thin client.. | |
10:10 | <garymc> Also at the minute my sound only works for a short while on my thin clients.
| |
10:11 | <[GuS]> i saw in other distros, like UBuntu, that there is a pulseaudio daemon running on server, but not in gentoo. This could be related alexqwesa?
| |
10:11 | <garymc> Im running vers 10 ubuntu ltsp
| |
10:12 | <Gadi> you dont need pulseaudio daemon on the server - only the client
| |
10:13 | you do need pulseaudio client libs on the server
| |
10:13 | if you have pulseaudio-utils on the server side, you can use command line tools like "paplay <file.wav>" to test
| |
10:13 | <[GuS]> alexqwesa: PULSE_SERVER says: tcp:192.168.1.76:4713 which is the thin client ip
| |
10:14 | Gadi: ok
| |
10:15 | <alexqwesa> [GuS]: look good, also apps should be configured to use pulseaudio (server side)
| |
10:15 | <[GuS]> alexqwesa: indeed, they are also compiled to use pulseaudio
| |
10:16 | and in preferences (of audacious for example) has pulseaudio
| |
10:17 | but when i play a song, audacious freeze for example
| |
10:18 | <alexqwesa> hmm, see pulseaudio logs
| |
10:19 | staffencasa has joined #ltsp | |
10:21 | <alexqwesa> [GuS]: nmap -p 4713 192.168.1.76 (on server side - should be open)
| |
10:21 | <[GuS]> ok
| |
10:21 | alexqwesa: is closed :S
| |
10:21 | 4713/tcp closed unknown
| |
10:21 | <alexqwesa> port on client should be open
| |
10:21 | check firewall
| |
10:22 | <[GuS]> it does not have a firewall...
| |
10:22 | <alexqwesa> on client: check is deamon run?
| |
10:23 | *demon
| |
10:25 | <[GuS]> i dont see any pulse daeon running
| |
10:26 | also in the console of the thin client i see that alsa is using the thin clint driver, but under gnome i see the server sound driver
| |
10:27 | <alexqwesa> [GuS]: on client console run pulseaudio -vvv
| |
10:28 | and check port again
| |
10:28 | <[GuS]> ok
| |
10:28 | alexqwesa: under the desktop console right? or the chroot?
| |
10:29 | damn... many errors
| |
10:29 | <alexqwesa> login to client via ssh and run pulseaudio -vvv or switch on client CTRL+ALT+F1
| |
10:30 | <[GuS]> ok, that is what i am asking
| |
10:35 | alexqwesa: pasting last stuff http://pastebin.com/DBxzxvGD
| |
10:37 | for short: http://pastebin.com/JfQt1KUD
| |
10:42 | <alexqwesa> so, you have pulseaudio bug(or maybe alsa) on client side, i'm not expert in PA - google'ing by this errors messages or ask someone else (also better post full log)
| |
10:44 | <[GuS]> i am trying to save the output by >> but it does not save :S
| |
10:45 | <alexqwesa> [GuS]: pulseaudio -vvv 2>&1 >file.log
| |
10:45 | [GuS]: also, you use prebuilt image or custom?
| |
10:46 | <[GuS]> alexqwesa: the stage?
| |
10:46 | is a prebuilt
| |
10:46 | stage3-i686-20100511.tar.bz2
| |
10:48 | dobber has quit IRC | |
10:49 | <alexqwesa> [GuS]: also paste /var/log/messages
| |
10:52 | <[GuS]> ok
| |
10:53 | alexqwesa: still does not save the output of pulseaudio -vvv :S
| |
10:53 | also, now i'e changed the soundcard, and does noty say anymore the bug in the soundriver, but just in consolekiyt
| |
10:53 | consolekit*
| |
10:54 | bah.. nevermind... the soundcard is not detected... back into the old one.
| |
10:55 | mistik1 has joined #ltsp | |
10:55 | <[GuS]> also i am running a "thin client" into virtualbox, but i have the same problem on a existing thin client...
| |
10:57 | <alexqwesa> pulseaudio -vvv 2>&1 1>file.log
| |
10:59 | <[GuS]> alexqwesa: same... does not save
| |
11:33 | gbit has joined #ltsp | |
11:33 | gbit has left #ltsp | |
11:36 | mistik1 has quit IRC | |
11:55 | ogra_cmpc has quit IRC | |
11:56 | ogra_cmpc has joined #ltsp | |
11:56 | <[GuS]> alexqwesa: the asound.conf configuration for pulse must go on the server side, right?
| |
11:57 | <johnny> i never made any modifications to asound.con
| |
11:57 | <[GuS]> johnny: well, i did it cause it says in the docs... indeed, i dont know what else more to do... reallly....
| |
11:57 | i tried many things...
| |
11:57 | <johnny> well it works for me
| |
11:57 | talk to the authors
| |
11:57 | <[GuS]> and in #gentoo i dont have much help
| |
11:57 | <johnny> file upstream tickets
| |
11:58 | #gentoo is pointless
| |
11:58 | <[GuS]> johnny: well, believe me that i am doing that stuff
| |
11:58 | <johnny> ltsp for gentoo really needs a new maintainer :(
| |
11:58 | [GuS], weren't you going to switch to ubuntu?
| |
11:58 | if your bos doesn't like it.. tell him to get his butt in here
| |
11:59 | tell him the maintainer told you to switch
| |
11:59 | <[GuS]> johnny: my boss still wants gentoo... he have no plans to change... also in my particular case i've installed ubuntu 10.04 as LTSP inside vbox and works EVERYHING... :(
| |
11:59 | <johnny> and that he's nuking ltsp 4 from orbit ('ts the only way to be sure)
| |
11:59 | tell him the maitnainer told you to do it
| |
11:59 | tell himi he doesn't recommend anyone to use it
| |
11:59 | i'll lie to him for you
| |
12:00 | <[GuS]> :P well... we use an application that he only preparted it for gentoo... so that is difficult... besides he is kind closed minded :S
| |
12:00 | <johnny> it's not fair to force somebody to use something they don't want
| |
12:00 | dude.. packages can be ported
| |
12:00 | it's just a distro.. it's not a whole new OS
| |
12:00 | <[GuS]> johnny: is what i told him.. besides i know how to pack for ubuntu for years...
| |
12:00 | <johnny> tell him to come in here
| |
12:00 | tell him to fix the shit
| |
12:00 | <[GuS]> we are suffering so many problems with our clients cause the use of gentoo... stilll he dont want to change...
| |
12:00 | <johnny> oops.. wrong channel for such language
| |
12:00 | find a new job :)
| |
12:00 | <[GuS]> jhaha yeah...
| |
12:01 | johnny: yeah, like that is easy these days.... at least not in Argentina :P
| |
12:01 | garymc has quit IRC | |
12:01 | dgroos has joined #ltsp | |
12:01 | <johnny> well i hear argentina is doign better nowadays than in the past .. if it weren't for the global troubles
| |
12:01 | <[GuS]> yeah, is better today
| |
12:02 | <johnny> tell your boss to fly me down there :)
| |
12:02 | or pay me to consult :)
| |
12:02 | <[GuS]> haha that will be nice
| |
12:02 | <johnny> then i'll consult him not to use it
| |
12:02 | hah
| |
12:02 | it's usuable. but gentoo itself changes all the time
| |
12:03 | so things work and then stop
| |
12:03 | which is ok for people who keep on top of this stuff
| |
12:03 | but it's not ok for you :(
| |
12:03 | <[GuS]> exactly... is what i told him... Just with one client he let me to install a mail server with *buntu... but for ltsp he still wants me to crash my head inside the monitor to solve it :S
| |
12:04 | everything that i solve or try to solve in ltsp takes hours...
| |
12:04 | <johnny> if you want, i could mail him
| |
12:04 | <[GuS]> is amazing how great worked Ubuntu ltsp and even inside a vbox!! event local devices like pendrives!
| |
12:04 | <johnny> it worked for me last time i used gentoo, but packages change, and i have no box.. nor time
| |
12:05 | <[GuS]> johnny: he will try to make it work for sure :P
| |
12:05 | mo matter how many hours to waste :P
| |
12:05 | no*
| |
12:05 | <johnny> how about i remove the files
| |
12:05 | <[GuS]> yeah... i know... besides this slow me down other task in my work...
| |
12:05 | <johnny> i can do that :)
| |
12:05 | tell him i unpublished it
| |
12:05 | <[GuS]> jeje
| |
12:05 | :P
| |
12:06 | gcooper has quit IRC | |
12:13 | <[GuS]> well.. sent an email to him :P will see... but i know that he is capable of spend days trying to make it work (and same with others many problems in gentoo)... which is a waste of time for our clients... but well... his Company.
| |
12:13 | but well... so much offtopic for now :P
| |
12:14 | * [GuS] back to LTSP... | |
12:15 | <[GuS]> also mounting local pendrives does not work... even when i have LOCALDEV=True enabled
| |
12:15 | <johnny> you never have tos et that
| |
12:15 | do you have ltspfs installed in the chroot?
| |
12:15 | <[GuS]> let me check
| |
12:16 | <johnny> it could be that the directory needed to mount has changed in upstream
| |
12:16 | and the packages need adjusted
| |
12:16 | <[GuS]> emerge: there are no ebuilds to satisfy "sys-fs/ltspfs". :O
| |
12:16 | <johnny> it should be putting stuff in /var/run/drives last i remember
| |
12:16 | you do have the overlay in there right?
| |
12:16 | <[GuS]> yep
| |
12:16 | <johnny> prove it
| |
12:16 | <[GuS]> but when i do emerge -s it list it
| |
12:16 | <johnny> then it's not masked properly
| |
12:17 | you had a bad install
| |
12:17 | <[GuS]> wait, let me double check.. at least this morning it was..
| |
12:17 | <johnny> obviously it didn't complete
| |
12:17 | if you don't actually have that
| |
12:17 | it's part of the install, not something that is installed manually
| |
12:17 | <[GuS]> oks
| |
12:18 | johnny: this what i mean http://pastebin.com/XxSHighz now i will check the overlay...
| |
12:18 | <johnny> 0.6 ?
| |
12:18 | hmm.. it should be a bzr version
| |
12:19 | unless knipwim_ added a regular tarball
| |
12:20 | <[GuS]> mm mi remounted everything in the chroot and the same...
| |
12:25 | mmm johnnyi should install inside chroot the ltspfs from overlay using layman?
| |
12:25 | <johnny> i think so.. i have no idea where ltspfs came from otherwise?
| |
12:25 | <[GuS]> from the instalation of the client maybe?
| |
12:26 | <johnny> sure, but it should be a bzr version,
| |
12:26 | <[GuS]> also at this point i have a mess in my head... you can notice that :P
| |
12:26 | <johnny> well sure, but it's likely that it is already installed..
| |
12:28 | <[GuS]> when i tried inside the client, loggued into the ssh i have "FATAL: Module ltspfs not found."
| |
12:29 | <johnny> module ?
| |
12:29 | you need fuse installed .. that's the only module required
| |
12:29 | ltspfs is not really a module
| |
12:33 | <[GuS]> ahh right.. yeah i have fuse
| |
12:34 | jhutchins_lt has joined #ltsp | |
12:49 | hawkI has quit IRC | |
13:01 | dgroos has quit IRC | |
13:05 | ogra_cmpc has quit IRC | |
13:06 | ogra_cmpc has joined #ltsp | |
13:14 | ogra_cmpc has quit IRC | |
13:15 | johnny has left #ltsp | |
13:17 | ogra_cmpc has joined #ltsp | |
13:22 | <Gadi> hey, all
| |
13:22 | who's got the gconftool command to disable user switching in Ubuntu 10.04 handy?
| |
13:22 | anyone?
| |
13:22 | :)
| |
13:24 | <alkisg> moment...
| |
13:25 | desktop/gnome/lockdown/disable_user_switching
| |
13:25 | <Gadi> alkisg to the rescue !
| |
13:25 | <alkisg> Need the whole command?
| |
13:25 | <Gadi> thx
| |
13:25 | if handy
| |
13:25 | if not, not to worry
| |
13:26 | <alkisg> sudo gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.mandatory --set /desktop/gnome/lockdown/disable_user_switching --type boolean "false"
| |
13:27 | * alkisg uses /usr/share/gconf/mandatory/* files instead of gconftool-2... | |
13:27 | <alkisg> ...easier for upgrades etc
| |
13:28 | (logoff/logon needed to take effect)
| |
13:29 | <Gadi> actually, I should set it to true
| |
13:29 | but, yeah, thx
| |
13:29 | <alkisg> Urm yeah sorry
| |
13:41 | johnny has joined #ltsp | |
13:42 | <Gadi> !seen stgraber
| |
13:42 | <ltspbot> Gadi: stgraber was last seen in #ltsp 2 weeks, 0 days, 15 hours, 37 minutes, and 47 seconds ago: <stgraber> you need to use ports.ubuntu.com instead of archive.ubuntu.com/ubuntu
| |
13:44 | <johnny> !seen sbalnaev
| |
13:44 | <ltspbot> johnny: I have not seen sbalnaev.
| |
13:44 | * johnny can't spell | |
13:44 | <johnny> what happened to scott
| |
13:56 | ogra_cmpc has quit IRC | |
14:18 | dgroos has joined #ltsp | |
14:41 | ogra_cmpc has joined #ltsp | |
14:42 | korcan has joined #ltsp | |
14:43 | bernardo0cd0 has joined #ltsp | |
14:45 | <bernardo0cd0> I have LTSP and VirtualBox installed here, when I start the VB client in the LTSP client, the terminal crashes and logout the user (no apparent messages or errors). Someone may have a thought on why?
| |
14:47 | <alkisg> You login on ldm and then it crashes?
| |
14:47 | Or it's before ldm?
| |
14:48 | <bernardo0cd0> after login, with the desktop running
| |
14:48 | dgroos has quit IRC | |
14:48 | <bernardo0cd0> when I open the VB client
| |
14:48 | but with X2Go (NX Session) works normally
| |
14:49 | <alkisg> So you see the desktop, and some minutes later it crashes?
| |
14:49 | <bernardo0cd0> no, the desktop works normally, but when I open the app VirtualBox (to access a VM) the LTSP crashes, and return to the login
| |
14:50 | <alkisg> Ah sorry I thought you meant that you were running an ltsp client in vbox
| |
14:50 | <bernardo0cd0> no, it's an exotic thing, i know :D
| |
14:51 | dgroos has joined #ltsp | |
14:51 | <bernardo0cd0> the strange thing is that isn't even when I open the VM (wich could be then a problem with image transfer or something), but on the VB manager!
| |
14:52 | <alkisg> Check that users ~/.xsession-errors
| |
14:52 | <bernardo0cd0> ok, ty
| |
14:53 | pmatulis has quit IRC | |
14:54 | [GuS] has quit IRC | |
14:54 | <bernardo0cd0> omg, the .xsession-erros have nothing less nothing more than 468MB!
| |
14:56 | that's a lot of errors...
| |
14:58 | dgroos has quit IRC | |
15:09 | <bernardo0cd0> alkisg, those errors where from a problem with ALSA, when the problem happens no message is logged on .xsession-errors
| |
15:13 | any other idea?
| |
15:20 | * alkisg doesn't know if it's possible, never tried it.. | |
15:20 | <alkisg> check that that user can run vbox on the server, and try launching vbox from a terminal, so that you see any error messages there. Also check the system logs while running it
| |
15:24 | mikkel has quit IRC | |
15:27 | grantk has joined #ltsp | |
15:28 | <grantk> Is there any detailed output on ltsp load balancer output, I try and start it but it just fails and the ltsp-cluster-lbserver.log does not move at all
| |
15:31 | Here is my lbsconfig.xml, http://ltsp.pastebin.com/QQSPw710 and I was following these setup instructions. https://help.ubuntu.com/community/UbuntuLTSP/LTSP-Cluster
| |
15:33 | bernardo0cd0 has quit IRC | |
15:49 | <grantk> I see this message in the log files but figured it might be due to the fact the load balancer is not running. ltsp-directory[1193]: ,ip=10.0.2.91,mac=00:50:8B:34:66:1D,display=0.0,code=2, - cannot_set_status: This message is not set in text-en.phpThis is not a valid IP address : appServIp=0.0.0.0
| |
15:50 | alkisg_ has joined #ltsp | |
15:52 | <Gadi> grantk: does your app server have more than one interface?
| |
15:53 | <grantk> no
| |
15:54 | jhutchins_lt has quit IRC | |
15:55 | alkisg has quit IRC | |
15:55 | <Gadi> do you have the agents installed on the app servers?
| |
15:56 | alkisg has joined #ltsp | |
15:57 | <grantk> Yes, that is running on the app server
| |
15:58 | denisesball has joined #ltsp | |
15:58 | <denisesball> hello again everyone. is RCFILE_01 from the manual actually implemented?
| |
15:58 | <alkisg> Yes
| |
15:59 | <denisesball> actually, now that im reading this more closely i dont think its for what i thought it was for
| |
15:59 | "Alternatively, you may have a script you’ve written that you’ve put in the chroot, and want to make sure gets run at startup."
| |
15:59 | does that mean it will run it when a user logs in?
| |
15:59 | <alkisg> No. You can put things in /etc/xdg/autostart for users logon
| |
15:59 | <denisesball> because later it says "Additional RC scripts can be executed by the ltsp-client-setup script."
| |
16:00 | <alkisg> (in the server)
| |
16:01 | <denisesball> /etc/xdg/autostart looks like desktop entries, can i put a script in there?
| |
16:03 | i suppose i could just make a custom .desktop that calls a script
| |
16:05 | <alkisg> right
| |
16:05 | <denisesball> im gonna make a script to find and remove stale firefox lock files when people log in after firfox kills their terms
| |
16:06 | <grantk> denise: I was wondering if there is something like that that can be ran earlier? I know when I have a user crash there is generally a good amount of processes left behind.
| |
16:07 | <denisesball> its very frustrating. someone goes to fark.com and an large image or flash video kills their terminal, then they log back in and cant open firefox because it leaves the .parentlock file in there
| |
16:07 | same with tb
| |
16:07 | firefox and tb do not handle crashing very well
| |
16:07 | <grantk> yep, I have a program in house that stays open too!
| |
16:07 | no not at all, I try to convince my users to use kmail and opera just for that reason.
| |
16:08 | <alkisg> denisesball: check the lts.conf man page for LDM_LIMIT_ONE_SESSION
| |
16:08 | <grantk> alkisg: I have tried that one before but was not sure how it would handle different servers.
| |
16:08 | <alkisg> I'm not sure, does denisesball use different servers?
| |
16:09 | <grantk> If I have a user crash on server a, ...ahhhh ok
| |
16:09 | <denisesball> i have 4 different servers, but the users are assigned to them statically so they never move around
| |
16:09 | <alkisg> That would work then
| |
16:09 | <grantk> ah, yeah that would work, mine float around
| |
16:09 | <denisesball> where is that option from? its not in the manual
| |
16:10 | <grantk> I found it on an opensuse page one day.
| |
16:10 | <denisesball> what does it do exactly?
| |
16:10 | <grantk> from what I read it only allows a user to have one session so if they crash before they log back on all their other processes are killed.
| |
16:11 | <denisesball> yeah but my problem isnt the stale processes, it's the .parentlock files
| |
16:11 | <grantk> http://en.opensuse.org/SDB:KIWI-LTSP_tips_and_tricks
| |
16:11 | <alkisg> LDM_LIMIT_ONE_SESSION => kills all old processes when a user logs in, that would take care of firefox
| |
16:11 | <denisesball> firefox only thinks its running because that file never gets deleted
| |
16:11 | the process isnt actually running
| |
16:11 | <alkisg> LDM_LIMIT_ONE_SESSION_PROMPT => ask the user before the killing
| |
16:11 | denisesball: are you sure? I never saw that problem...
| |
16:11 | <grantk> denisesball: If firefox recieves a signal 15 will it remove the lock file?
| |
16:12 | <denisesball> alkisg: quite, it happens every few days
| |
16:12 | i'll run in here when it happens again
| |
16:12 | <alkisg> denisesball: so if I kill -9 firefox, then I can't start it again?
| |
16:12 | <grantk> I would think it does, I know when I have a user crash and the process is around killing it fixes the problem
| |
16:12 | -9 is pretty harsh, may not give firefox time to clean up
| |
16:13 | <alkisg> That's the point, to leave the lock file behind
| |
16:13 | litercola has joined #ltsp | |
16:13 | <denisesball> alkisg: when their terminals crash they get logged out. im not sure how firefox dies, but it doesn't clean up its .parentlock file which ties it to a PID that is no longer there and it thinks its still running
| |
16:13 | <alkisg> denisesball: when the terminals crash, firefox is still running...
| |
16:13 | <denisesball> ill check again next time it happens
| |
16:14 | <alkisg> ...along with all their other processes, ssh, gnome-session etc
| |
16:14 | <Gadi> denisesball: is it local firefox?
| |
16:14 | ie localapp
| |
16:14 | <litercola> ive installed ubuntu server, can't i do something like "ltsp-build-client ubuntu-desktop" to build the desktop system for my clients
| |
16:15 | <denisesball> i may be assuming its acting the same as when the server itself would freeze. i had a wonky server in there that kept freezing and even after rebooting it i had this problem. maybe its not doing that currently when just the terminal logs out
| |
16:15 | Gadi: no
| |
16:15 | <alkisg> litercola: not exactly. You can run fat clients, but you can't easily run thin clients on a different chroot.
| |
16:15 | litercola: thin clients run their processes directly on the server, not on a chroot
| |
16:15 | dgroos has joined #ltsp | |
16:18 | * Gadi recalls when you could get a 2-liter cola for the same price | |
16:18 | <Gadi> those were the days....
| |
16:18 | <litercola> lol
| |
16:19 | just like cigs used to be $2
| |
16:19 | <Gadi> don't even get me started on this 1.5 liter bottle business...
| |
16:19 | :)
| |
16:19 | <litercola> i know
| |
16:25 | anyone still in here?
| |
16:28 | <grantk> yeah
| |
16:28 | <litercola> so im very confused...
| |
16:28 | <grantk> I think you want to just install a package like ubuntu-desktop
| |
16:28 | your wanting to use only one server for ltsp right?
| |
16:29 | <litercola> correct
| |
16:29 | <grantk> ok then you should just be able to just install a normal desktop and ldm should pick it up
| |
16:30 | <litercola> will the server then be configured with those packages as well
| |
16:30 | <grantk> yeah, you would be installing them on the server
| |
16:30 | <litercola> so i would be booting up into x, correct?
| |
16:31 | <grantk> if you choose too, I usually disable X so I do not waste the resources
| |
16:31 | <litercola> gotcha, one of the main resons i reinstalled the server version, is that things seemed kind of flaky due to network manager and such
| |
16:32 | <grantk> I want to say you will need to run ltsp-build-client after you install the desktop
| |
16:32 | <litercola> that makes sense, thank you for your help
| |
16:32 | <grantk> yeah, I am not a fan of network manager, you can remove that after you install the desktop. np
| |
16:33 | <litercola> cool, thanks again
| |
16:33 | litercola has quit IRC | |
16:44 | pmatulis has joined #ltsp | |
16:45 | <grantk> ok, I found some socket errors with lbserver,
| |
16:45 | Cannot retrieve node 'Esme', error:<class 'socket.error'>
| |
16:46 | right after the last one of those I get a code 400, message Bad request syntax ('getkey')
| |
16:49 | <Gadi> grantk: is Esme really capitalized like that?
| |
16:50 | can you "ping Esme"
| |
16:50 | <grantk> darn, it is really capitalized like that in the lbsconfig.xml, but not for the hostname or in dns
| |
16:51 | but did not fix it.
| |
16:52 | Gadi: Does the app server need to be in dns? I did not throw in there because I was just using it for testing right now? I had figured with the ip entered into the lbsconfig.xml I would be ok
| |
16:53 | <Gadi> try adding Esme to the lbserver's /etc/hosts for now
| |
16:53 | or "esme"
| |
16:53 | :)
| |
16:53 | <grantk> yeah, doing that now, I forgot I added my other ltps servers and after the fact decided to make a test copy! ooops
| |
16:54 | ok, I can ping esme now, but still no luck
| |
16:55 | <Gadi> are you starting the lbserver from /etc/init.d
| |
16:55 | <grantk> yes
| |
16:55 | <Gadi> and ps -efw|grep lbs returns nothing
| |
16:56 | <grantk> well, look at that, im sorry , was searching for the wrong process earlier
| |
16:56 | <Gadi> try: wget -O- -q http://localhost:8008
| |
16:56 | <grantk> returned the ip of the server
| |
16:57 | <Gadi> of the app server, right?
| |
16:57 | <grantk> yes
| |
16:57 | <Gadi> then you are good to go
| |
16:57 | if you add more app servers, and more node lines to the lbserver's xml file, that ip should change
| |
16:57 | <grantk> cool, just rebooted the client after restarting the lbserver so I got em crossed!
| |
16:58 | vagrantc has joined #ltsp | |
16:58 | <Gadi> you should also be able to look at the terminal logs in the control center
| |
16:58 | <grantk> ah, the least loaded then
| |
16:58 | <Gadi> and see the App server IP
| |
16:58 | <grantk> darn, still getting no appSerIp error
| |
16:58 | <Gadi> then you have something wrong with your getltscfg-cluster.conf in the chroot
| |
16:59 | or the control center db is messed up
| |
16:59 | <grantk> ok, ill check that out
| |
16:59 | <Gadi> the control center works off pgsql
| |
16:59 | and does not take kindly to problems with the db
| |
16:59 | :)
| |
16:59 | * Gadi had a heck of a time with redundant control centers | |
17:00 | <Gadi> and I am still not happy with the solution
| |
17:00 | :P
| |
17:00 | <grantk> bummer
| |
17:00 | * Gadi counts his grey hairs and dreams of 2-liter cola bottles | |
17:00 | <grantk> do you know what table I should be seeing the app server in the db?
| |
17:00 | jhutchins_lt has joined #ltsp | |
17:01 | <Gadi> grantk: thats not it
| |
17:01 | <grantk> the old kind with the hard plastic bottom?
| |
17:01 | <Gadi> hehe
| |
17:01 | if you get a shell on the thin client
| |
17:01 | <grantk> ok, on to that chroot config file then
| |
17:01 | <Gadi> run: getltscfg-cluster -a
| |
17:01 | if you get errors, that will help explain things
| |
17:02 | Ahmuck-Jr has joined #ltsp | |
17:02 | <grantk> said that one was not installed
| |
17:02 | Ahmuck-Jr has left #ltsp | |
17:02 | <Gadi> do you have a shell as root?
| |
17:02 | <grantk> oh nope
| |
17:03 | can you still set that in lts.conf?
| |
17:03 | <Gadi> yup
| |
17:03 | something like:
| |
17:03 | SCREEN_02=shell
| |
17:03 | SCREEN_07=ldm
| |
17:03 | <grantk> hadn ot looked into it yet, just saw the 07 option on that web interface
| |
17:03 | <Gadi> doh
| |
17:03 | * Gadi slaps head | |
17:03 | <Gadi> right
| |
17:04 | ideally, you get stuff from control center
| |
17:04 | however, if the control center is not working...
| |
17:04 | maybe it is easier if you set a root password for the chroot
| |
17:04 | and then switch to screen 1
| |
17:04 | <grantk> I did do that in the setup, let me try that
| |
17:04 | <Gadi> and log in there
| |
17:05 | <grantk> CLUSTER_CONFIGURED "false"
| |
17:06 | export CLUSTER_CONFIGURED
| |
17:07 | oh, looks like there is no config in the chroot
| |
17:08 | from that script looks like it should be /etc/ltsp/getltscfg-cluster.conf
| |
17:08 | <Gadi> indeed
| |
17:09 | that is what it all hinges upon
| |
17:09 | like a house of cards
| |
17:09 | :)
| |
17:09 | cliebow has joined #ltsp | |
17:09 | <grantk> good to know!
| |
17:09 | <Gadi> configure that up and you are golden
| |
17:09 | bbiab
| |
17:09 | <grantk> cool, thanks
| |
17:09 | <cliebow> bye Gadi
| |
17:09 | <Gadi> dont forget to update your image
| |
17:09 | * Gadi waves to cliebow | |
17:09 | <Gadi> cliebow: I was just in Maine (South part)
| |
17:10 | * cliebow cliebow genuflects | |
17:10 | <Gadi> was thinking of you
| |
17:10 | <grantk> ohh, thanks, would have forgot that one!
| |
17:10 | <cliebow> heheh.
| |
17:10 | <Gadi> brb
| |
17:10 | <cliebow> k
| |
17:19 | bobby_C has quit IRC | |
17:23 | johnny has left #ltsp | |
17:24 | dgroos has quit IRC | |
17:25 | korcan has quit IRC | |
17:29 | <grantk> Well, I made a getltscfg-cluster.conf with the following entries
| |
17:29 | http://ltsp.pastebin.com/UyvPcgGg
| |
17:29 | still no luck
| |
17:31 | I see a mention of /etc/ltsp/directory.conf in the getlts-conf script, do I need one of those?
| |
17:38 | * Kyle__ sighs. | |
17:38 | <Kyle__> For some reason, I see a /usr/sbin/in.tftpd --listen being run, even though tftpd is supposed to be off of inetd.
| |
17:39 | Gadi has left #ltsp | |
17:45 | <alkisg> Kyle__: lucid?
| |
17:46 | <Kyle__> alkisg: Yea.
| |
17:46 | <alkisg> Kyle__: it's been changed, it no longer runs from inetd
| |
17:47 | <Kyle__> I'm a bit confused about that, but everything booted fine anyway, so I can ignore it...
| |
17:47 | Really?
| |
17:47 | Humm.
| |
17:47 | <alkisg> (so if you have that in inetd.conf, remove it)
| |
17:48 | <Kyle__> I did have it in there. Fixing now.
| |
17:48 | Wait... no. How did I get confused on this? Nope, it's not being run for inetd.
| |
17:48 | * Kyle__ sighs. | |
17:49 | <Kyle__> Do you need to/does it help to set the address in the tftp config?
| |
17:49 | <alkisg> No
| |
17:49 | <Kyle__> OK. That's the IP the client uses to while pulling the image, right?
| |
17:49 | Or am I confusing it with another boot system?
| |
17:50 | <alkisg> The kernel. The chroot image is server by nbd
| |
17:50 | Can we take this backwards? Do you have a specific problem?
| |
17:50 | *s/server/served/
| |
17:51 | <Kyle__> alkisg: I had a problem, it ended up being unrelated, but I noticed that while figuring it out.
| |
17:53 | Although, now's a good time to ask.... is there a good/standard/not painfull way to bounce all the clients that aren't currently logged in with the new image?
| |
18:12 | johnny has joined #ltsp | |
18:18 | Faithful has joined #ltsp | |
18:41 | staffencasa has quit IRC | |
18:50 | ogra has quit IRC | |
19:05 | ogra has joined #ltsp | |
19:12 | vagrantc has quit IRC | |
19:12 | jhutchins_lt has quit IRC | |
19:21 | pmatulis has quit IRC | |
19:30 | alkisg has quit IRC | |
19:46 | ogra_cmpc has quit IRC | |
19:54 | grantk has left #ltsp | |
19:58 | pmatulis has joined #ltsp | |
19:59 | ogra_cmpc has joined #ltsp | |
20:20 | mae_tae has joined #ltsp | |
20:20 | pmatulis has quit IRC | |
20:21 | <mae_tae> how do i resolve this "cannot bind to ipv4 address is in use", its TFTP matter
| |
20:35 | <johnny> don't run 2 tfpt servers/
| |
20:35 | you probably tried to start one manually or added to your init scripts unnecessarily
| |
20:35 | usually one is automatically spawned on demand
| |
20:35 | pointing to the right place
| |
20:35 | but if another one is in place, it won't start
| |
20:44 | <mae_tae> johnny, how do i check that i run 2 tftp servers? i've done this ps -aux | grep tftp or netstat -nap | grep 69, but i dont find running tftp
| |
20:45 | but if you type tftp localhost it will give me tftp> command line
| |
20:45 | johnny, if i type netstat -a | grep tftp, there is a result
| |
20:47 | johnny, what im doing is setting up THIN client,
| |
20:47 | mae_tae has quit IRC | |
20:48 | mae_tae has joined #ltsp | |
20:52 | <mae_tae> people, if i type ps -aux | grep tftp or netstat -nap | grep 69, i couldnt find any tftp port in there or tftp but if i type netstat -a | grep tftp, there is a result
| |
20:53 | <johnny> sure.. you don't want that
| |
20:53 | turn it off
| |
20:53 | remove it from your init script
| |
20:53 | ifi you have one
| |
20:55 | <mae_tae> how i stop it?
| |
21:02 | <johnny> well first kill it and try to boot a client
| |
21:02 | to make sure that is the problem
| |
21:05 | <mae_tae> how do i kill when i cannot seen in command ps -aux | grep tftp, it can only be seen in netstat -a | grep tftp, in netstat does it has that process id?
| |
21:09 | <johnny> it should have a process id in netstat
| |
21:09 | but really you should use -l
| |
21:09 | that looks for a listening process
| |
21:10 | and check to see what process it is
| |
21:10 | <mae_tae> johnny will i do this netstat -al | grep tftp
| |
21:10 | is that the one? ill try it
| |
21:13 | japerry has quit IRC | |
21:14 | mae_tae has quit IRC | |
21:19 | johnny has left #ltsp | |
21:22 | cliebow has quit IRC | |
21:25 | mae_tae has joined #ltsp | |
21:27 | <mae_tae> johnny, if you already had reply, pls post it again, im disconnected
| |
21:32 | intelliant has joined #ltsp | |
21:39 | FGXR6 has quit IRC | |
21:46 | johnny has joined #ltsp | |
22:14 | mae_tae has quit IRC | |
22:43 | intelliant has left #ltsp | |
22:44 | intelliant has joined #ltsp | |
22:47 | <intelliant> hi!
| |
22:47 | How do i persist changes to group.conf ? The changes I make for the ltsp users seem to be lost every now and then. ... may be whenever i run apt-get or on reboot
| |
23:12 | Appiah: hi!
| |
23:12 | is there anything wrong in having each of the appserver's installed with sshd
| |
23:12 | ?
| |
23:19 | how can we make chabges to /etc/security/groups.conf persistent
| |
23:19 | they seem to be lost in the appservers
| |
23:36 | Aug 26 10:06:43 ltsp-appserve01 ltsp-cluster-accountmanager: Killing user "shrenik.bhura"
| |
23:36 | ^ can anyone please explain this ?
| |
23:37 | found this in syslog of the appserver in an ltsp-cluster setup
| |
23:37 | this throws out this user consistently
| |
23:38 | sweetpi has joined #ltsp | |
23:42 | johnny has left #ltsp | |
23:42 | johnny has joined #ltsp | |
23:44 | <intelliant> does anyone know how exactly the ltsp-cluster-accountmanager work ?
| |
23:45 | for approx 10-12 mins of an appserver reboot, it did not allow any user to persistently login
| |
23:45 | i am testing the logins using ssh
| |
23:46 | it used to throw out the user within a few seconds of login
| |
23:46 | with this in syslog ltsp-appserve01 ltsp-cluster-accountmanager: Killing user
| |
23:46 | after 10-12 mins the user is finally in
| |
23:51 | stgraber: need help on ltsp-cluster-accountmanager
| |
23:58 | sweetpi has quit IRC | |