IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 9 September 2010 (all times are UTC)
| 00:10 | japerry_cat has joined #ltsp | |
| 00:14 | japerry has quit IRC | |
| 00:14 | japerry_cat is now known as japerry | |
| 00:39 | cyberorg has quit IRC | |
| 00:47 | cyberorg has joined #ltsp | |
| 01:09 | cyberorg has quit IRC | |
| 01:13 | poff has joined #ltsp | |
| 01:20 | cyberorg has joined #ltsp | |
| 01:42 | shogunx has quit IRC | |
| 02:02 | gnunux has joined #ltsp | |
| 02:06 | RiXtEr has quit IRC | |
| 02:06 | RiXtEr has joined #ltsp | |
| 02:08 | <gnunux> hi
| |
| 02:11 | <alkisg> Good morning
| |
| 02:11 | <muppis> Huomenta.
| |
| 02:12 | alkisg, have you ever played around with touchscreens and ltsp?
| |
| 02:12 | <alkisg> Nope, never had a touchscreen
| |
| 02:13 | <muppis> Darn. I think I got soon call to get those working in thins.
| |
| 02:13 | s/got/gets
| |
| 02:14 | I can get them working in local installation (done that), but ltsp is still whole new world for me.
| |
| 02:16 | <alkisg> What's needed to make them work locally?
| |
| 02:19 | <muppis> I meant standalone-pc, not ltsp related.
| |
| 02:24 | <vmlintu> muppis: we tried some early Asus EeeTop 15" touchscreen model and got it working when required drivers were installed in the chroot..
| |
| 02:25 | This was with Ubuntu 8.04
| |
| 02:25 | <alkisg> muppis: yes, I mean what's needed to make them work on standalone PCs...
| |
| 02:29 | <muppis> vmlintu, ok. Thank you for that.
| |
| 02:29 | alkisg, ok.
| |
| 02:30 | Well, waiting for the call..
| |
| 02:43 | otavio has quit IRC | |
| 02:44 | MrMarkie has joined #ltsp | |
| 02:46 | DaGeek_ has quit IRC | |
| 02:54 | otavio has joined #ltsp | |
| 03:01 | japerry has quit IRC | |
| 03:01 | japerry_cat has joined #ltsp | |
| 03:01 | japerry_cat is now known as japerry | |
| 03:14 | shogunx has joined #ltsp | |
| 03:15 | RicardoPerez has joined #ltsp | |
| 03:20 | <muppis> !compiz
| |
| 03:20 | <ltspbot> muppis: "compiz" :: if compiz is giving you problems, one way to disable it for all users is: sudo gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.mandatory --type string --set /desktop/gnome/session/required_components/windowmanager metacity
| |
| 04:05 | Eghie has joined #ltsp | |
| 04:12 | <RicardoPerez> hi!
| |
| 04:13 | <alkisg> Hello
| |
| 04:13 | <RicardoPerez> well, i'm at work, trying to solve a rather strange problem with my ltsp installation
| |
| 04:13 | the ltsp (fat) clients sometimes refuses to boot and hangs with no evident reason
| |
| 04:14 | i removed the "quiet splash" part of the kernel boot params
| |
| 04:14 | and i see that sometimes the clients hangs in a specific moment, and sometimes hangs in another
| |
| 04:15 | when the client hangs, after a while, i can see many error messages like this:
| |
| 04:15 | INFO: task modprobe:503 blocked for more than 120 secods.
| |
| 04:16 | "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
| |
| 04:16 | INFO: task avahi-daemon:669 blocked for more than 120 secods.
| |
| 04:16 | "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
| |
| 04:16 | and so on...
| |
| 04:16 | dobber has joined #ltsp | |
| 04:17 | <alkisg> You may want to try disabling nbd-proxy... let me find the bug report...
| |
| 04:18 | <RicardoPerez> if i disable nbd-proxy, will the nbd-server work well?
| |
| 04:18 | <alkisg> https://bugs.launchpad.net/ltsp/+bug/589034
| |
| 04:18 | nbd-proxy was only recently added, and is only needed for roaming sessions on ltsp-cluster
| |
| 04:18 | For non-cluster ltsp installations, it only adds overhead...
| |
| 04:19 | vmlintu has quit IRC | |
| 04:19 | <RicardoPerez> alkisg, great, i'll look at the bugreport
| |
| 04:20 | ok, i'm going to apply the patch and i'll tell you :) thank you!
| |
| 04:20 | <alkisg> RicardoPerez: if you can see any other messages while booting, do post them here
| |
| 04:20 | (before the tasks being killed)
| |
| 04:20 | <RicardoPerez> another boot messages:
| |
| 04:21 | Begin: Running /scripts/init-bottom ...
| |
| 04:21 | Done.
| |
| 04:21 | init: gdm main process (446) killed by TERM signal
| |
| 04:21 | udev[431]: worker [444] unexpectedly returned with status 0x0100
| |
| 04:21 | udev[431]: worker [444] failed while handling '/devices/pci0000:00/0000:00:1b.0'
| |
| 04:21 | and so on...
| |
| 04:22 | <alkisg> Erm, are those messages from a client that didn't boot?
| |
| 04:23 | Wait, maybe I misunderstood. I thought that the other days you said that you had boot problems, are they boot problems or general hanging problems?
| |
| 04:23 | <RicardoPerez> alkisg, that's right. that messages are from clients that didn't boot
| |
| 04:23 | however, the udev messages doesn't display always
| |
| 04:24 | <alkisg> (12:22:58 PM) RicardoPerez: init: gdm main process (446) killed by TERM signal ==> that's supposed to show up on shutdown, not on boot...
| |
| 04:24 | <RicardoPerez> 99% times the client simply hangs during boot
| |
| 04:24 | <alkisg> So the first problematic messages are the udev ones/
| |
| 04:24 | ?
| |
| 04:24 | <RicardoPerez> the gdm error message is very strange...
| |
| 04:25 | <alkisg> Then it's not nbd-proxy related, it might be upstart related
| |
| 04:25 | <RicardoPerez> i think so, too... maybe upstart isn't able to handle all the signals or something?
| |
| 04:26 | <alkisg> no idea about upstart...
| |
| 04:26 | <RicardoPerez> i'll try with the nbd-proxy patch now, anyway
| |
| 04:27 | it's very strange... my chroot is not customized in any way since the build
| |
| 04:27 | i only followed the steps from https://help.ubuntu.com/community/UbuntuLTSP/FatClients
| |
| 04:28 | curiously, sometimes the clients boots perfectly well, and sometimes hangs during boot (maybe in the next reboot)
| |
| 04:29 | * RicardoPerez waiting for ltsp-update-image to complete | |
| 04:50 | Da-Geek has joined #ltsp | |
| 04:54 | <RicardoPerez> alkisg, it works perfectly!
| |
| 04:54 | <alkisg> RicardoPerez: so it was another problem of nbd-proxy?
| |
| 04:54 | <RicardoPerez> alkisg, i tried the patch against the nbd-proxy and now the clients boots all the time!
| |
| 04:54 | alkisg, i tried almost 20 times and 100% success
| |
| 04:54 | <alkisg> Ugh, do post that under the bug report, but also specify that it affect distribution: "ltsp (ubuntu)" too
| |
| 04:55 | shamino has quit IRC | |
| 04:55 | <RicardoPerez> alkisg, great, i'll do that. thank you very very much!!
| |
| 04:55 | <alkisg> You're welcome
| |
| 04:59 | shamino has joined #ltsp | |
| 05:57 | Da-Geek has quit IRC | |
| 06:27 | <moquist> is there an easy way I can just install ldm on a regular ubuntu system and then point it to an arbitrary ltsp server that has ssh open to the Internet?
| |
| 06:28 | I haven't been able (yet) to figure out how ldm knows which server to use...
| |
| 06:28 | pmatulis has joined #ltsp | |
| 06:29 | RicardoPerez has quit IRC | |
| 06:30 | <cyberorg> moquist, just use ssh -X serverIP
| |
| 06:30 | <moquist> OK, and then run 'X' (or whatever)
| |
| 06:30 | that's what I was going to try now
| |
| 06:30 | installing a vncserver locally
| |
| 06:30 | <cyberorg> moquist, you can start whole session ssh -X serverIP /path/to/desktop-session
| |
| 06:31 | ssh does not need vncserver/viewer
| |
| 06:31 | <moquist> right, but I have a desktop environment locally
| |
| 06:31 | I'd like a fresh X server for the remote desktop
| |
| 06:32 | GodFather has joined #ltsp | |
| 06:35 | <cyberorg> Xnest :1&
| |
| 06:35 | DISPLAY=:1 ssh -X username@serverIP /etc/X11/xdm/Xsession
| |
| 06:35 | ugly, but that works :)
| |
| 06:36 | <moquist> oh, right!
| |
| 06:36 | <cyberorg> or use NX
| |
| 06:36 | * moquist had forgotten Xnest, too | |
| 06:36 | <moquist> so many things forgotten
| |
| 06:39 | MorningSon has joined #ltsp | |
| 06:45 | <moquist> cyberorg: thanks for your help -- this is going great now
| |
| 06:46 | well, it's over the In-ter-net, and as you said, it's ugly, but it's working fine. :)
| |
| 06:46 | <cyberorg> nx works really well over the net
| |
| 07:21 | sid has quit IRC | |
| 07:24 | vvinet has quit IRC | |
| 07:27 | * alkisg votes for neatx too, it works fine over wan | |
| 08:06 | [GuS] has joined #ltsp | |
| 08:10 | rjune has joined #ltsp | |
| 08:34 | mikkel has joined #ltsp | |
| 08:56 | <alkisg> !docs
| |
| 08:56 | <ltspbot> alkisg: "docs" :: For the most current documentation, see https://sourceforge.net/apps/mediawiki/ltsp/index.php?title=Ltsp_LtspDocumentationUpstream
| |
| 08:59 | MrMarkie has quit IRC | |
| 08:59 | reynolds has joined #ltsp | |
| 09:00 | <reynolds> anyone around can help me figure out why my thin clients are saying " Failed to connect to NBD server" on 10.4?
| |
| 09:02 | <alkisg> reynolds: how's your dhcp setup?
| |
| 09:02 | dhcp server == ltsp server?
| |
| 09:02 | <reynolds> I have to Nics on my server and acts as my dhcp server
| |
| 09:02 | two
| |
| 09:03 | <alkisg> So the ltsp clients are on an ISOLATED switch?
| |
| 09:03 | <reynolds> yes
| |
| 09:03 | <alkisg> OK, can you paste your /etc/ltsp/dhcpd.conf ?
| |
| 09:03 | !pastebot
| |
| 09:03 | <ltspbot> alkisg: "pastebot" :: The LTSP pastebot is at http://ltsp.pastebin.com. Please paste all text longer than a line or two to the pastebot, as it helps to reduce traffic in the channel. Don't forget to paste the URL of the text here.
| |
| 09:05 | <reynolds> http://ltsp.pastebin.com/nKzicD8t
| |
| 09:05 | <alkisg> What's your eth1 (the ltsp-facing nic) IP/
| |
| 09:05 | ?
| |
| 09:06 | <reynolds> 192.168.1.98
| |
| 09:07 | <alkisg> The dhcp range is 192.168.0, so that's outside of your subnet?
| |
| 09:07 | MrMarkie has joined #ltsp | |
| 09:07 | <reynolds> 192.168.0.254 sorry
| |
| 09:07 | <alkisg> OK, and the other one is 192.168.1.98?
| |
| 09:07 | <reynolds> yes
| |
| 09:08 | <alkisg> Look good. OK, next step: open /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default
| |
| 09:08 | Ah sorry, before that:
| |
| 09:08 | when you see that "failed to connect to nbd" message, do you get a busybox shell?
| |
| 09:08 | <reynolds> yes
| |
| 09:08 | <alkisg> on the server, run: grep nbd /etc/inetd.conf
| |
| 09:08 | Which port is there? 2000?
| |
| 09:09 | (2000 and 9572 by default)
| |
| 09:10 | <reynolds> yes 2000 and 9572
| |
| 09:10 | <alkisg> OK, on the client, on busybox, run: nbd-client 192.168.0.254 2000 /dev/nbd0
| |
| 09:11 | <reynolds> ok hold on
| |
| 09:13 | error server closed connection
| |
| 09:13 | <alkisg> So it's a server problem
| |
| 09:13 | Can you pastebin the result of: grep nbd /etc/inetd.conf
| |
| 09:13 | or paste it here, it's just 2 lines...
| |
| 09:14 | <reynolds> http://ltsp.pastebin.com/iyFBhEJU
| |
| 09:16 | <alkisg> /opt/ltsp/images/=.img ==> that's not right
| |
| 09:16 | ls -lha /opt/ltsp/images
| |
| 09:16 | What does this give you? ^^^
| |
| 09:16 | <reynolds> total 571M
| |
| 09:16 | drwxr-xr-x 2 root root 4.0K 2010-09-09 08:59 .
| |
| 09:16 | drwxr-xr-x 4 root root 4.0K 2010-09-08 17:26 ..
| |
| 09:16 | -rw-r--r-- 1 root root 571M 2010-09-09 08:59 i386.img
| |
| 09:16 | <alkisg> OK, sudo gedit /etc/inetd.conf and change that line to:
| |
| 09:16 | 2000 stream tcp nowait nobody /usr/sbin/tcpd /usr/sbin/nbdrootd /opt/ltsp/images/i386.img
| |
| 09:17 | Then, save and: sudo invoke-rc.d openbsd-inetd restart
| |
| 09:17 | Then reboot the client
| |
| 09:19 | <reynolds> YOUR THE MAN!!! THANK YOU any idea how it might have become corrupt?
| |
| 09:21 | <alkisg> No idea at all, never seen that kind of corruption
| |
| 09:24 | <reynolds> well thanks for your help,
| |
| 09:26 | <alkisg> You're welcome
| |
| 09:31 | mistik1 has quit IRC | |
| 09:32 | mistik1 has joined #ltsp | |
| 09:32 | jconlon has joined #ltsp | |
| 09:35 | reynolds has quit IRC | |
| 09:36 | <jconlon> I read in the documentation that gnome-watchdog's functionality is embedded in ltsp5, is this correct. Ubuntu documentation said to add a line to ltsp.conf, which I did, but it doesn't seem to be working.
| |
| 09:36 | <alkisg> No, it isn't.
| |
| 09:36 | There are two similar things:
| |
| 09:37 | 1) LDM_LIMIT_ONE_SESSION in lts.conf, and
| |
| 09:37 | 2) xexit, by sbalneav
| |
| 09:38 | <jconlon> I think 1 was what it said to add to lts.conf, but just to make sure, which lts.conf should I edit? /opt/ltsp/i386/lts.conf?
| |
| 09:39 | <alkisg> No
| |
| 09:39 | /var/lib/tftpboot/ltsp/i386/lts.conf
| |
| 09:39 | <jconlon> thanks, thats what I figured. Are you a dev for this project?
| |
| 09:40 | <alkisg> A junior dev :)
| |
| 09:40 | There's also LDM_LIMIT_ONE_SESSION_PROMPT=True/False, which presents a yes/no dialog before killing the processes
| |
| 09:40 | <jconlon> you are very helpful. Which is the best option in your opinion?
| |
| 09:41 | <alkisg> No idea, I'm not using any of those
| |
| 09:41 | <jconlon> I have like 20 chrome processed running with 3 users logged in.
| |
| 09:42 | <alkisg> Chrome uses multiple processes, one for each tab, afaik
| |
| 09:43 | poff has quit IRC | |
| 09:43 | <jconlon> If I try option 1, do I need to use a =TRUE with that?
| |
| 09:43 | <alkisg> LDM_LIMIT_ONE_SESSION=True
| |
| 09:44 | LDM_LIMIT_ONE_SESSION_PROMPT=True
| |
| 09:44 | Both of them is a good option, I think...
| |
| 09:44 | <jconlon> One last question, is there any way that I can monitor that this is doing something? Will it write to a log file or something?
| |
| 09:46 | <alkisg> It will present a dialog if you try to login 2 times
| |
| 09:47 | <jconlon> But it should kill old processes that may remain after a user logs out, correct?
| |
| 09:47 | <alkisg> No, only when they try to login again
| |
| 09:47 | xexit does what you said instead
| |
| 09:47 | <jconlon> would it be bad to use both?
| |
| 09:47 | <alkisg> Nope
| |
| 09:49 | <jconlon> Thank you, I will look into xexit as well, you are quite a bad@$$, thanks again for all the help.
| |
| 09:49 | MrMarkie has quit IRC | |
| 09:51 | jconlon has quit IRC | |
| 10:08 | GodFather has quit IRC | |
| 10:11 | staffencasa has joined #ltsp | |
| 10:12 | efra has joined #ltsp | |
| 10:12 | vmlintu has joined #ltsp | |
| 10:20 | gnunux has quit IRC | |
| 10:40 | jhutchins_lt has joined #ltsp | |
| 11:25 | dobber has quit IRC | |
| 11:32 | jhutchins_lt has quit IRC | |
| 11:34 | poff has joined #ltsp | |
| 11:35 | jhutchins_lt has joined #ltsp | |
| 12:39 | dmarkey_ has joined #ltsp | |
| 12:46 | dberkholz has quit IRC | |
| 12:46 | dmarkey has quit IRC | |
| 12:48 | dberkholz has joined #ltsp | |
| 12:59 | Kyle__ has joined #ltsp | |
| 12:59 | <Kyle__> Hello.
| |
| 12:59 | I'm having a pebkac error I'm affraid.
| |
| 12:59 | hiltunen has joined #ltsp | |
| 12:59 | <Kyle__> I tried to set my default session, but my clients say that session doesn't exist.
| |
| 13:00 | anyone have experience setting the default session?
| |
| 13:08 | efra has quit IRC | |
| 13:11 | vagrantc has joined #ltsp | |
| 13:16 | <moldy> hi
| |
| 13:16 | some users on my setup cannot login. other users can login fine from the same workstation. any idea what could be causing this? the users can login fine with plain ssh. i don't get any error message.
| |
| 13:19 | <Kyle__> moldy: do you see anything in the auth.log?
| |
| 13:22 | <alkisg> !compiz
| |
| 13:22 | <ltspbot> alkisg: "compiz" :: if compiz is giving you problems, one way to disable it for all users is: sudo gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.mandatory --type string --set /desktop/gnome/session/required_components/windowmanager metacity
| |
| 13:22 | <alkisg> moldy: that's a frequent cause, worth a try even before troubleshooting.... ^
| |
| 13:23 | <abeehc-_> I guess that defaults to compiz?
| |
| 13:23 | <alkisg> Yup
| |
| 13:23 | <Kyle__> alkisg: Thanks for that compiz note, it's bugging me, but I was going to ignore it until I had everything else happy :)
| |
| 13:23 | <alkisg> No need to restart anything, just try to logon on a client after issuing that
| |
| 13:23 | <abeehc-_> my clients suck at rebooting
| |
| 13:23 | <Kyle__> alkisg: do you need to path the full path to a session with LDM_SESSION/LDM_XSESSION?
| |
| 13:24 | <alkisg> Kyle__: never played with sessions, sorry....
| |
| 13:24 | abeehc-_: meaning?
| |
| 13:24 | <Kyle__> Ah ok. Thanks.
| |
| 13:24 | abeehc-_: clients as in machines or users?
| |
| 13:26 | <abeehc-_> haha at this point my users can't reboot as far as i know
| |
| 13:26 | short of the power button
| |
| 13:26 | but i mean i'm just testing a client at my desk here; ctrl+alt+del on console, results in spew of squash errors, and no reboot
| |
| 13:26 | until the power button is held which is normaly a pain
| |
| 13:27 | i've not played with sessions to this point either, sorry
| |
| 13:28 | <Kyle__> abeehc-_: I know the feeling. I observed a professor _barely_ maintain composure yesterday, when a student asked how to unzip a file in windows. Worst part? It was in the second programming class in the series. The students have all had computer usage courses, as well as another programming course.
| |
| 13:28 | <abeehc-_> haha
| |
| 13:29 | <Kyle__> abeehc-_: Humm. What clinet OS are you using? Maybe it needs some tweaks for proper shutdown on that hardware?
| |
| 13:29 | RicardoPerez has joined #ltsp | |
| 13:29 | * Kyle__ has actually seen that on modern hardware again... not just a thing of the past. | |
| 13:29 | <alkisg> abeehc-_: distro/version?
| |
| 13:30 | <abeehc-_> pretty sure it has something to do with nbd
| |
| 13:30 | ubuntu 9.10
| |
| 13:30 | seen the same stuff on 9.04 with non-sgraber-ppa
| |
| 13:30 | <ltsplogbot> I don't know who the is.
| |
| 13:30 | <Kyle__> Hum. OK that's one distro that I've rarely seen that issue with.
| |
| 13:30 | anivair has joined #ltsp | |
| 13:32 | * Kyle__ ponders | |
| 13:32 | <Kyle__> abeehc-_: are you using OpenVZ or somefin?
| |
| 13:33 | <anivair> So the new thin Clients are picking up dhcp now (magically) but here's the thing: it takes about 30 seconds for it to grab the dhcp offer. So I watch the little line pinwheel for about 30 seconds before it does anything. Anyone know why that might be?
| |
| 13:33 | <alkisg> abeehc-_: in lucid it's handled by upstart so it works better with alt+ctrl+delete
| |
| 13:33 | abeehc-_: otherwise you either need to reboot/shutdown from ldm, or with poweroff -fp
| |
| 13:34 | <RicardoPerez> hi! I only want to say that I've another issue regarding user's home exporting using nfs. the client sometimes doesn't mount /home. looking at the error messages I see that the statd daemon doesn't run well during boot all the times. I fix the issue hacking the ltsp-client-setup script adding a "service statd start" line just before the "configure_home || true" line. Hope this could be helpful for someone ;)
| |
| 13:35 | <alkisg> RicardoPerez: I'm using -nolock to bypass that, but we should file a bug against it, it's broken on lucid...
| |
| 13:35 | NFS_HOME_OPTIONS="defaults,nolock"
| |
| 13:35 | <RicardoPerez> alkisg: that sounds good, i'll try it tomorrow... and yes, a bug report should be filed :)
| |
| 13:36 | <alkisg> RicardoPerez: wait, I was having stad problems on the server, are you having problems on the client too?
| |
| 13:36 | <RicardoPerez> thank you very much, again :)
| |
| 13:36 | <alkisg> (not on ltsp, on statd, whatever package that is in)
| |
| 13:36 | * alkisg guesses it's an upstart problem | |
| 13:36 | <RicardoPerez> alkisg: yes, sometimes the ltsp client doesn't mount the /home dir via nfs
| |
| 13:37 | alkisg: so the client regenerate a new /home dir for the user which is log in
| |
| 13:37 | <alkisg> RicardoPerez: yes, I was having the same symptoms because statd wasn't started on the server
| |
| 13:37 | So, if I ran "start statd" on the server, then the clients booted...
| |
| 13:37 | <RicardoPerez> do yo do so on the server instead of the client?
| |
| 13:37 | <alkisg> On the server
| |
| 13:37 | Yes
| |
| 13:38 | I haven't tried on the client, to see if it makes any difference, I started using NFS_HOME_OPTIONS instead
| |
| 13:38 | <abeehc-_> you'd think pastebin would allow image posting
| |
| 13:38 | at least i would think that
| |
| 13:38 | <alkisg> imagebin.org
| |
| 13:38 | <abeehc-_> ty
| |
| 13:38 | <RicardoPerez> mmmmmmmm
| |
| 13:38 | well, I did the following:
| |
| 13:39 | I hack the ltsp-client-config, configure_home() function, in the line that says:
| |
| 13:39 | if ! mount -t nfs ...; then
| |
| 13:39 | I added "2> /tmp/output" to the mount command
| |
| 13:40 | then, after the client boots, I do a "cat /tmp/output"
| |
| 13:40 | and then I see error messages concerning statd
| |
| 13:40 | <alkisg> I did about the same, but I thought it was refering to the server statd
| |
| 13:40 | So I started statd on the server, and it worked...
| |
| 13:41 | <RicardoPerez> alkisg: interesting, I'll try it, then
| |
| 13:41 | so, what do you think I should do: NFS_HOME_OPTIONS or statd running on the server?
| |
| 13:41 | <alkisg> Do you have that output file handy? Could you pastebin it? Maybe someone who knows more about nfs can help us...
| |
| 13:42 | <RicardoPerez> no, sorry, the problem is at work, and i'm on home, now...
| |
| 13:42 | <alkisg> I think the "correct way" (TM) is to make sure statd is running on the server,
| |
| 13:42 | ...but I went for NFS_HOME_OPTIONS to get some speed with no locks, because I know I don't need them...
| |
| 13:42 | * alkisg doesn't know enough about nfs to be sure, though | |
| 13:42 | <RicardoPerez> alkisg: what about the two solutions together?
| |
| 13:43 | <alkisg> Sure, that'd work too, I think...
| |
| 13:43 | <RicardoPerez> great, i'll investigate it tomorrow, then...
| |
| 13:43 | mmmm..... i think the /etc/init/statd.conf file should be hacked...
| |
| 13:43 | more specifically, the "start on ..." line
| |
| 13:44 | i.e. the starting event
| |
| 13:44 | <abeehc-_> wow that service failed me pretty hard.. would you be able to say this is consistent with what you mentioned about upstart doing a better job earlier, alkisg
| |
| 13:44 | <RicardoPerez> I wonder if putting "start on runlevel [S]" could be an enhancement
| |
| 13:44 | <abeehc-_> http://conan.tripleblack.net/~mmeija/IMG00430-20100909-1133.gif
| |
| 13:45 | or is ctrl+alt+del at the console bad practice
| |
| 13:45 | <alkisg> No, it's bad ltsp implementation :D Fixed on Lucid, though...
| |
| 13:46 | <abeehc-_> hah damn
| |
| 13:46 | even using ltsp ppas?
| |
| 13:47 | <alkisg> Not sure, I think it'd be fixed with ppas. Is that on boot or on shutdown? I don't see any "sending term messages..."
| |
| 13:47 | japerry has quit IRC | |
| 13:47 | <abeehc-_> it's on reboot in theory; I did ctrl+alt+del from the console
| |
| 13:48 | japerry has joined #ltsp | |
| 13:48 | <alkisg> Does it work ok if you reboot from ldm?
| |
| 13:48 | bobby_C has joined #ltsp | |
| 13:48 | <alkisg> RicardoPerez: i've seen quite a few problems with upstart scripts (/etc/init/*.conf), maybe that's one too, it's worth filing a bug...
| |
| 13:49 | loather-work has joined #ltsp | |
| 13:49 | <Vonor> is it possible to have an auto-login for a specific client?
| |
| 13:49 | <RicardoPerez> alkisg: about the statd on the server, interestingly, I had the following behavior: a client boots without nfs'ed /home, I then reboot it (without touching the server in any way) and now the client boots with the /home served perfectly good
| |
| 13:49 | <abeehc-_> hmm I don't know how to reboot via ldm I guess
| |
| 13:50 | for the most part, my ctrl+alt+del works as I would expect
| |
| 13:50 | <alkisg> abeehc-_: click on the button on the bottom left corner, and select reboot
| |
| 13:50 | <abeehc-_> sometimes I can shell into these hung machines and see there's no filesystem
| |
| 13:50 | yep that worked perfectly
| |
| 13:50 | anivair has quit IRC | |
| 13:50 | <abeehc-_> what's that trigger? maybe I can improve my crappy reboot shell script
| |
| 13:50 | <alkisg> abeehc-_: let me find you the upstart script for rebooting with alt+ctrl+delete...
| |
| 13:51 | <abeehc-_> thank you
| |
| 13:51 | <alkisg> $ pastebinit /opt/ltsp/i386/etc/init/ltsp-reboot.conf
| |
| 13:51 | http://pastebin.com/4F9iHdaB
| |
| 13:51 | $ pastebinit /opt/ltsp/i386/etc/init/ltsp-shutdown.conf
| |
| 13:51 | http://pastebin.com/zpjuYiN6
| |
| 13:51 | Did you say 9.04 or 9.10?
| |
| 13:52 | I think 9.10 uses upstart, so I believe it would work there, but not on 9.04
| |
| 13:53 | <abeehc-_> 9.10 with them ppas
| |
| 13:54 | thanks that should get me in the right direction
| |
| 13:54 | <alkisg> np
| |
| 13:55 | <RicardoPerez> alkisg: did you had some good /home nfs mounting from a client when the statd isn't running on the server?
| |
| 14:00 | <moldy> Kyle__: nothing in auth.log
| |
| 14:01 | alkisg: i will try disabling compiz, thank you
| |
| 14:01 | <Kyle__> Humm.
| |
| 14:02 | <moldy> Kyle__: well, to be precies, auth.log contains a successfull login
| |
| 14:03 | <alkisg> RicardoPerez: with NFS_HOME_OPTIONS="defaults,nolock", yeah, it always works because it doesn't use statd
| |
| 14:03 | Without that, no, it never boots without statd on the server
| |
| 14:04 | moldy: the symptoms you describe are the same that are caused by compiz, so yeah try that
| |
| 14:04 | <Kyle__> moldy: Suffessful login? Toss your lts.conf on pastebin. lets see if there's anything odd there...
| |
| 14:04 | <moldy> Kyle__: i'll try compiz first, give me a minute :)
| |
| 14:04 | <Kyle__> OK.
| |
| 14:10 | <alkisg> abeehc-_: btw, in my ppa there are packages that enable ltsp clients to reboot/shutdown from within the gnome session
| |
| 14:12 | <RicardoPerez> alkisg: thank you very much! i'll try all the options tomorrow ;)
| |
| 14:12 | <alkisg> np
| |
| 14:14 | <RicardoPerez> alkisg: btw, may it be some risk if we don't use statd?
| |
| 14:14 | <alkisg> Afaik yes, if both users try to write on the same file
| |
| 14:14 | * alkisg doesn't have that use case in his classrooms | |
| 14:14 | <alkisg> *simultaneously
| |
| 14:15 | <RicardoPerez> alkisg: oh, well, i'll take it in mind... my classroom is a bit anarchic about that...
| |
| 14:15 | thanks again!
| |
| 14:22 | <Kyle__> Well it doesn't solve any of my real problems :) but that line to disable compiz really does make the ubuntu sessions run smoother.
| |
| 14:23 | RicardoPerez has quit IRC | |
| 14:23 | <elias_a> Good evening, dudes...
| |
| 14:23 | Good evening, fellowes...
| |
| 14:24 | Could you be so kind and check m$ page http://www.microsoft.com/uk/education/schools/products/learning-gateway/test-drive-learning-gateway.aspx
| |
| 14:25 | Links under Teaching and Learning work in Windoze but not in Linux...
| |
| 14:25 | Could someone confirm this?
| |
| 14:27 | <Kyle__> elias_a: Looks like it could be a CSS blunder on their part.
| |
| 14:27 | <alkisg> elias_a: there's an <h3> overlay on top of the links, so clicking goes to h3 instead of the link
| |
| 14:28 | RicardoPerez has joined #ltsp | |
| 14:28 | efra has joined #ltsp | |
| 14:28 | vagrantc has quit IRC | |
| 14:28 | <elias_a> alkisg: Well - that does not happen if one uses Windoze...
| |
| 14:29 | <alkisg> elias_a: you mean it doesn't happen neither with firefox nor with chrome on windows?
| |
| 14:29 | Or do you mean IE?
| |
| 14:29 | <RicardoPerez> alkisg: well, i can confirm that statd *should* be running *in client* in order to get the "mount -t nfs ... /home" working in the client
| |
| 14:29 | <elias_a> alkisg: I do not touch IE :)
| |
| 14:29 | <alkisg> RicardoPerez: thanks, I'll keep that in mind
| |
| 14:29 | <elias_a> Works on FF3
| |
| 14:29 | <RicardoPerez> if statd isn't running in the client, i get the following error message:
| |
| 14:29 | mount.nfs: rpc.statd is not running but is required for remote locking.
| |
| 14:30 | mount.nfs: Either use '-o nolock' to keep locks local, or start statd.
| |
| 14:30 | alkisg: you're welcome :)
| |
| 14:30 | it should be, obviously, an issue in upstart (or /etc/init/*.conf scripts)
| |
| 14:31 | <alkisg> RicardoPerez: so, if statd is running on the client, but not on the server, does mounting work?
| |
| 14:31 | <Kyle__> elias_a: Well, it really looks like a CSS issue, from what I can see. So either it was written for ie6 or quirks mode on 7/8, or they purposely wrote css that doesn't work properly in firefox/chrome (pretty unlikely).
| |
| 14:31 | <RicardoPerez> alkisg: YES!
| |
| 14:31 | <elias_a> Kyle__: It does not run on FF & Ubuntu. It runs on FF3 on Windoze
| |
| 14:32 | <RicardoPerez> alkisg: statd is not running on server, is running on client only, and the "mount -t nfs ... /home" command on client works OK
| |
| 14:32 | <Kyle__> Very odd.
| |
| 14:32 | Doesn't run under chrome in linux either.
| |
| 14:32 | <alkisg> RicardoPerez: nice, thanks
| |
| 14:32 | <RicardoPerez> alkisg: you're welcome again :) obviously a problem in upstart, then
| |
| 14:33 | <alkisg> elias_a: well, if you delete that <h3> it works on linux too
| |
| 14:34 | <elias_a> alkisg: Really?
| |
| 14:34 | <alkisg> <h3>Management and administration</h3> <== that one
| |
| 14:34 | Yes, it floats over the links and intercepts all the clicks
| |
| 14:34 | I'd bet it's just bad css
| |
| 14:34 | <elias_a> So are you getting a page that says "Education Test Drive"?
| |
| 14:35 | <alkisg> Yup
| |
| 14:35 | <elias_a> Bad is the keyword in the case of this company.
| |
| 14:36 | <alkisg> You can work around it with a greasemonkey script or a custom css or something similar...
| |
| 14:36 | <elias_a> alkisg: I am not going to make a workaround.
| |
| 14:37 | <alkisg> Nice :)
| |
| 14:37 | <elias_a> I am going to kick their butt!
| |
| 14:37 | <alkisg> ...even better :D
| |
| 14:37 | Time to go, bye alll.
| |
| 14:37 | <abeehc-_> bye ty again
| |
| 14:37 | * Kyle__ hums to himself as all the stations start syncing new data on boot. | |
| 14:37 | <Kyle__> This is nice.
| |
| 14:37 | <RicardoPerez> bye! cheers!
| |
| 14:37 | <elias_a> alkisg: Night!
| |
| 14:37 | alkisg has quit IRC | |
| 14:37 | <RicardoPerez> bye everybody!ç
| |
| 14:38 | RicardoPerez has quit IRC | |
| 14:38 | <elias_a> No - really guys... we'd better get organised...
| |
| 14:48 | vagrantc has joined #ltsp | |
| 14:50 | [GuS] has quit IRC | |
| 14:55 | pmatulis has quit IRC | |
| 15:14 | Vonor has quit IRC | |
| 15:24 | vagrantc has quit IRC | |
| 15:26 | sid has joined #ltsp | |
| 15:27 | <sid> is there anyone around that can help me figure out what is causing some of my client displays to be flipped
| |
| 15:35 | <Kyle__> sid: Are they all on the same hardware?
| |
| 15:47 | <sid> no
| |
| 15:49 | the two offending machines. I belive use a intel chipset
| |
| 15:52 | <Kyle__> sid: I hate to say it, but it could be one of those cases where you just need to specific some X specifics for those boxes :/
| |
| 15:52 | Once you figure out what they are.
| |
| 15:54 | loather-work has quit IRC | |
| 15:56 | mikkel has quit IRC | |
| 16:01 | <sid> ok thanks
| |
| 16:08 | Sorinan has joined #ltsp | |
| 16:15 | <Sorinan> jetpipe dies on client everytime I try to print something, exists any log that I can check why?
| |
| 16:16 | Selveste1 has quit IRC | |
| 16:25 | <abeehc-_> Sorinan: if you poke around I think there's a way you can change the script to not fork
| |
| 16:28 | <Sorinan> abeehc-_: fork? no, the script simply crash
| |
| 16:29 | poff has quit IRC | |
| 16:31 | loather-work has joined #ltsp | |
| 16:32 | Mava has joined #ltsp | |
| 16:35 | Selveste1 has joined #ltsp | |
| 16:37 | pmatulis has joined #ltsp | |
| 17:18 | bobby_C has quit IRC | |
| 17:38 | alkisg has joined #ltsp | |
| 17:42 | <alkisg> sid: nvidia server and intel clients?
| |
| 17:43 | !compiz
| |
| 17:43 | <ltspbot> alkisg: "compiz" :: if compiz is giving you problems, one way to disable it for all users is: sudo gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.mandatory --type string --set /desktop/gnome/session/required_components/windowmanager metacity
| |
| 17:43 | <alkisg> sid: that should fix it ^^^
| |
| 17:56 | japerry_cat has joined #ltsp | |
| 17:57 | japerry_cat has quit IRC | |
| 17:57 | japerry_cat has joined #ltsp | |
| 17:58 | japerry has quit IRC | |
| 17:58 | japerry_cat is now known as japerry | |
| 18:08 | efra has quit IRC | |
| 18:22 | alkisg has quit IRC | |
| 18:25 | vagrantc has joined #ltsp | |
| 18:32 | RiXtEr has quit IRC | |
| 18:33 | RiXtEr has joined #ltsp | |
| 18:42 | markit has joined #ltsp | |
| 18:43 | <markit> hi, kubuntu 10.04 here. I've installed ldm-kubuntu-theme, but hte ldm login screen has not changed. Any hint?
| |
| 18:43 | (I've also rebooted ltsp server...)
| |
| 18:45 | mmm maybe they are additional themes for the desktop, not for the ltsp login lol
| |
| 18:46 | jhutchins_lt has quit IRC | |
| 18:46 | <markit> no, I was right "Kubuntu theme for the LTSP Display Manager"
| |
| 18:48 | maybe I need to install it in chroot?
| |
| 18:58 | nothing changed :(
| |
| 19:23 | vagrantc has quit IRC | |
| 19:47 | ogra_cmpc has quit IRC | |
| 20:00 | ogra_cmpc has joined #ltsp | |
| 20:06 | korcan has joined #ltsp | |
| 20:31 | irule has joined #ltsp | |
| 20:37 | ogra has quit IRC | |
| 20:41 | <markit> ok, I solved
| |
| 20:42 | had do add a specific key in lts.conf
| |
| 20:42 | thought was a matter of packaging and select the right "alternatives", but was unable to figure out
| |
| 20:42 | not the right solution, but a decent one ;)
| |
| 20:42 | sleep time (3.44am here)
| |
| 20:43 | markit has quit IRC | |
| 20:43 | ogra has joined #ltsp | |
| 20:47 | japerry_cat has joined #ltsp | |
| 20:51 | japerry has quit IRC | |
| 20:51 | japerry_cat is now known as japerry | |
| 20:52 | <elias_a> 04:53 here. Yet already awake...
| |
| 20:58 | M4gic5t0rM has joined #ltsp | |
| 21:20 | pianolender has quit IRC | |
| 21:27 | M4gic5t0rM has quit IRC | |
| 21:46 | pmatulis has quit IRC | |
| 22:41 | Sorinan has quit IRC | |
| 23:08 | pianolender has joined #ltsp | |
| 23:09 | japerry has quit IRC | |
| 23:19 | japerry has joined #ltsp | |
| 23:23 | japerry has joined #ltsp | |