IRC chat logs for #ltsp on irc.freenode.net (webchat)


Channel log from 10 June 2012   (all times are UTC)

00:24uskerine has joined IRC (uskerine!~uske@58.Red-88-1-239.dynamicIP.rima-tde.net)
00:24
<uskerine>
hi, i have a server with two NICs (two IPs)
00:25
will thin clients work if in a failover scenario servers moves from one IP to the other one?
00:36F-GT has joined IRC (F-GT!~phantom@ppp121-44-97-140.lns20.syd6.internode.on.net)
01:04monteslu_ has joined IRC (monteslu_!~monteslu@ip68-109-174-213.ph.ph.cox.net)
01:07monteslu has left IRC (monteslu!~monteslu@ip68-109-174-213.ph.ph.cox.net, Ping timeout: 252 seconds)
01:21uskerine has left IRC (uskerine!~uske@58.Red-88-1-239.dynamicIP.rima-tde.net, Ping timeout: 244 seconds)
01:29vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Ping timeout: 265 seconds)
02:10awilliams has joined IRC (awilliams!mistik1@unaffiliated/mistik1)
02:38andygraybeal_ has left IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net, Ping timeout: 240 seconds)
02:40dgeary2 has joined IRC (dgeary2!~david@c114-77-214-254.rivrw3.nsw.optusnet.com.au)
02:57trubbor has joined IRC (trubbor!~trubbor@173-21-2-181.client.mchsi.com)
03:21andygraybeal_ has joined IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net)
03:53andygraybeal_ has left IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net, Ping timeout: 240 seconds)
04:06adrianorg_ has left IRC (adrianorg_!~adrianorg@177.18.48.65, Ping timeout: 245 seconds)
04:14jordanm has left IRC (jordanm!~jordanm@pdpc/supporter/active/jordanm, Remote host closed the connection)
04:24trubbor has left IRC (trubbor!~trubbor@173-21-2-181.client.mchsi.com, Remote host closed the connection)
04:58alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
05:49* alkisg finally managed to trim the initramfs of 12.04 so that it boots with 64 MB RAM! :)
06:15komunista has joined IRC (komunista!~slavko@adsl-195-168-241-172.dynamic.nextra.sk)
06:34
<alkisg>
Removing FRAMEBUFFER=y gets the initramfs from 20 Mb down to 16 MB. And that's the default on Ubuntu now, so I wonder why we do put that... stgraber any ideas?
06:39dgeary2 has left IRC (dgeary2!~david@c114-77-214-254.rivrw3.nsw.optusnet.com.au, Quit: ĝis la)
06:48
<alkisg>
!learn lowram-initramfs as to get the 12.04 initramfs about 8 MB smaller, so that it loads faster and boots with 64 MB RAM, remove FRAMEBUFFER=y from conf-hook.d/ltsp and comment out "auto_add_modules scsi" from hook-functions
06:48
<ltsp_>
alkisg: The operation succeeded.
06:58mmetzger has left IRC (mmetzger!~mmetzger@99-71-214-196.lightspeed.mdldtx.sbcglobal.net, Ping timeout: 245 seconds)
07:00mmetzger has joined IRC (mmetzger!~mmetzger@99-71-214-196.lightspeed.mdldtx.sbcglobal.net)
07:13vmlintu has left IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi, Remote host closed the connection)
08:02alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
08:09yanu has left IRC (yanu!~yanu@lugwv/member/yanu, Ping timeout: 248 seconds)
08:41yanu has joined IRC (yanu!~yanu@178-117-233-89.access.telenet.be)
08:41yanu has joined IRC (yanu!~yanu@lugwv/member/yanu)
09:06Phantomas has joined IRC (Phantomas!~Phantomas@unaffiliated/phantomas)
09:16andygraybeal_ has joined IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net)
09:23awilliams has left IRC (awilliams!mistik1@unaffiliated/mistik1, Quit: Lost terminal)
09:23awilliams has joined IRC (awilliams!mistik1@69.160.108.17)
09:23awilliams has joined IRC (awilliams!mistik1@unaffiliated/mistik1)
09:45komunista has left IRC (komunista!~slavko@adsl-195-168-241-172.dynamic.nextra.sk, Ping timeout: 260 seconds)
09:58komunista has joined IRC (komunista!~slavko@adsl-195-098-010-254.dynamic.nextra.sk)
10:12komunista has left IRC (komunista!~slavko@adsl-195-098-010-254.dynamic.nextra.sk, Ping timeout: 245 seconds)
10:15lefteris_nik has left IRC (lefteris_nik!~lefteris@ppp141237208190.dsl.hol.gr, Quit: Leaving.)
10:16lefteris_nik has joined IRC (lefteris_nik!~lefteris@ppp141237208190.dsl.hol.gr)
10:26komunista has joined IRC (komunista!~slavko@adsl-195-168-233-241.dynamic.nextra.sk)
10:27andygraybeal_ has left IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net, Ping timeout: 240 seconds)
10:59ricotz has joined IRC (ricotz!~rico@p5B2AC161.dip.t-dialin.net)
10:59ricotz has joined IRC (ricotz!~rico@unaffiliated/ricotz)
11:24komunista has left IRC (komunista!~slavko@adsl-195-168-233-241.dynamic.nextra.sk, Ping timeout: 240 seconds)
11:25mikkel has joined IRC (mikkel!~mikkel@80-71-132-15.u.parknet.dk)
11:44komunista has joined IRC (komunista!~slavko@adsl-195-168-233-241.dynamic.nextra.sk)
11:53pranav has joined IRC (pranav!75c1260d@gateway/web/freenode/ip.117.193.38.13)
11:53ogra_ has left IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de, Ping timeout: 260 seconds)
11:54pranav has left IRC (pranav!75c1260d@gateway/web/freenode/ip.117.193.38.13, Client Quit)
12:13adrianorg_ has joined IRC (adrianorg_!~adrianorg@177.18.48.65)
12:28Phantomas has left IRC (Phantomas!~Phantomas@unaffiliated/phantomas, Ping timeout: 260 seconds)
12:33andygraybeal_ has joined IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net)
12:45
<muppis>
!factoids search session
12:45
<ltsp_>
muppis: endsession: Epoptes contains a script that can reboot/shutdown LTSP clients from within the Gnome/KDE/XFCE/LXDE session: http://bazaar.launchpad.net/~epoptes/epoptes/trunk/view/head:/epoptes-client/endsession
12:46
<muppis>
!factoids search gnome
12:46
<ltsp_>
muppis: 'gnome-classic', 'gnome-fallback', 'gnome-panel', 'gnome-shell', 'gnome-wm', and 'lock-gnome3'
12:46
<muppis>
!gnome-shell
12:46
<ltsp_>
muppis: gnome-shell: To select gnome-shell as your default session, put this line in your lts.conf: LDM_SESSION="gnome-session --session=gnome"
12:47
<muppis>
Hoipe that works.
12:47
Hope..
12:55komunista has left IRC (komunista!~slavko@adsl-195-168-233-241.dynamic.nextra.sk, Ping timeout: 248 seconds)
13:15cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Remote host closed the connection)
13:21Phantomas has joined IRC (Phantomas!~Phantomas@unaffiliated/phantomas)
13:23cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg)
13:38khildin has joined IRC (khildin!~khildin@ip-80-236-193-97.dsl.scarlet.be)
13:47bobby_C has joined IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at)
13:50bobby_C has left IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at, Client Quit)
13:52Phantomas has left IRC (Phantomas!~Phantomas@unaffiliated/phantomas, Ping timeout: 260 seconds)
14:03adrianorg_ has left IRC (adrianorg_!~adrianorg@177.18.48.65, Ping timeout: 245 seconds)
14:06Phantomas has joined IRC (Phantomas!~Phantomas@unaffiliated/phantomas)
14:32ricotz has left IRC (ricotz!~rico@unaffiliated/ricotz, Quit: Ex-Chat)
14:38ricotz has joined IRC (ricotz!~rico@p5B2AC161.dip.t-dialin.net)
14:38ricotz has joined IRC (ricotz!~rico@unaffiliated/ricotz)
15:11vagrantc has joined IRC (vagrantc!~vagrant@c-76-115-60-19.hsd1.or.comcast.net)
15:11vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc)
15:20alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
15:22* vagrantc waves
15:22
<alkisg>
Hello :)
15:32
<vagrantc>
alkisg: do you think ltsp-docs is worth uploading in it's current state?
15:32
<alkisg>
vagrantc: I think it needs a bit of fixing
15:32
LDM_SESSION* and RM_*SERVICES
15:32
KEEP_SERVICES etc
15:32
<vagrantc>
alkisg: hm.
15:32
<alkisg>
I can do those in 5-6 days, when I've finished my sch-scripts sprint, if you can wait...
15:32
...maybe even sooner
15:33
Well, they shouldn't take long... I may even do them today later on
15:33
<vagrantc>
alkisg: it shouldn't be too bad, although i was wondering if it wouldn't be good to get what's there now uploaded so that any diffs will be smaller, since there was a huge formatting restructuring
15:34
should also try epoptes ... the code changes were small, but there are a lot of translation updates.
15:34
<alkisg>
np, although if you can wait a few hours I'll probably do them today
15:35
<vagrantc>
alkisg: that sounds great :)
15:36
alkisg: any thoughts regarding epoptes?
15:36
highvoltage: any thoughts regarding ldm-themes?
15:36
highvoltage: word on artwork?
15:37
<alkisg>
vagrantc: it's just translation updates, and a removal of a hack to bypass a twisted problem which was fixed in twisted, so it shouldn't need much testing...
15:37
<vagrantc>
alkisg: i think i'll upload epoptes today, then
15:51
<stgraber>
alkisg: FRAMEBUFFER=y was added to get plymouth to show up as early as possible so a normal user wouldn't see the ltsp and dhcp scripts output
15:52
alkisg: on a regular ubuntu system, the initrd pretty much doesn't do anything (we're actually trying to get rid of it entirely), so it's running for so little time that the user doesn't see it
15:52
(and ubuntu doesn't print anything on screen in the initrd)
15:54
<vagrantc>
stgraber: getting rid of the initrd?
15:54
<alkisg>
stgraber: OK, if it's ever needed I'll just override it from some local script. Do you think that FRAMEBUFFER=y can cause any problems? Why was it removed from regular installations in 11.10?
15:54
Just for the bit of boot time overhead it adds?
15:55
<stgraber>
vagrantc: yes, some kernel folks are looking at implementing UUID based rootfs in the kernel, once that's done most of our users won't need the initrd at all. We'd still use an initrd for encrypted disks, lvm, some raid and network root filesystems
15:56
<alkisg>
Cool, omitting the initrd should save some seconds from the boot process...
15:56
<stgraber>
alkisg: yeah, Ubuntu only sets framebuffer=y if the initrd contains some complex disk setup tools (lvm,nfs,cryptsetup,...) that would make the boot slower and so require the splash to show up ASAP
15:57
<alkisg>
OK, ty
15:57
<stgraber>
yeah, on ARM we can apparently cut 30% of the boot time by just not having an initrd (as the bootloader is really slow at reading it)
16:03
<highvoltage>
vagrantc: does the debian freeze affect artwork as well?
16:03* alkisg requests a build for the new ltsp server package that doesn't depend on nfs-kernel-server on ubuntu...
16:04* highvoltage reads a bit more to find context for the previous artwork question
16:04
<vagrantc>
highvoltage: probably less so
16:05
<alkisg>
vagrantc: could you do the same x-ubuntu-recommends trick on ldm too, to recommend ldm-ubuntu-theme | ldm-*buntu-theme on Ubuntu?
16:05
<vagrantc>
alkisg: impossible!
16:06
<highvoltage>
vagrantc: I've been following the lists and there's some artwork proposals for wheezy, but it's not decided yet which one will get in.
16:06
<vagrantc>
highvoltage: well, they'd better allow a freeze exception then, eh? :)
16:07
<alkisg>
bbiab
16:07alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
16:07
<vagrantc>
now that two packages need it, i should probably make a debhelper script or something
16:07
<highvoltage>
vagrantc: heh, yeah. I'm tempted to get the 2 most likely ones ready, since it doesn't take too long anyway
16:08
<vagrantc>
highvoltage: cool :)
16:11
<highvoltage>
so far http://wiki.debian.org/DebianArt/Themes/Joy seems like the closest to a possible candidate
16:11
the rest of them pretty much have weird penguins and logos in them
16:11
sometimes I'm not sure how someone who submits something like http://wiki.debian.org/DebianArt/Themes/penguinSpirit can take themselves seriously :)
16:12* highvoltage doesn't mean to bash
16:20F-GT has left IRC (F-GT!~phantom@ppp121-44-97-140.lns20.syd6.internode.on.net, Ping timeout: 245 seconds)
16:21alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
16:31andygraybeal_ has left IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net, Ping timeout: 240 seconds)
16:37
<vagrantc>
alkisg: were you going to test the kiosk plugin changes i made?
16:37
oh yes, and there's devilspie2 to look into ... ugh.
16:37
<alkisg>
vagrantc: nope, I never needed the kiosk mode so I never tested it in the past either
16:37* alkisg is updating lts.conf.xml...
16:38
<vagrantc>
ok.
16:39* vagrantc swears someone offered to test it on ubuntu
16:40komunista has joined IRC (komunista!~slavko@adsl-195-168-249-156.dynamic.nextra.sk)
16:59
<alkisg>
vagrantc: done with the docs
17:00
<vagrantc>
alkisg: rockin!
17:02
<alkisg>
!alkisg-todo
17:02
<ltsp_>
alkisg: alkisg-todo: (#1) ltsp-update-image.symlinks, or (#2) review RM_*_SERVICES lists, or (#3) put LDM_PRINTER_LIST example in lts.conf, or (#4) package a newer version of ltsp-docs in the Greek schools PPA, or (#5) upload fixed nbd-server package in the PPA
17:02
<alkisg>
!forget alkisg-todo 2
17:02
<ltsp_>
alkisg: The operation succeeded.
17:03
<alkisg>
stgraber: could you give an LDM_PRINTER or LDM_PRINTER_LIST example offhand, for me to put it in the example lts.conf ?
17:03
<vagrantc>
alkisg: so i'll try to build ltsp-docs and upload, then.
17:03
<alkisg>
vagrantc: yup, nice
17:03
<vagrantc>
alkisg: or wait for more?
17:04
<alkisg>
No, I'd like to test LDM_PRINTER before putting it to lts.conf.xml
17:04
I don't know its syntax
17:04* alkisg will test with multiple printers and labs etc in 2 weeks
17:05viggy_prabhu has joined IRC (viggy_prabhu!~viggy@117.216.129.62)
17:08
<alkisg>
!learn alkisg-todo as implement LDM_[PRE|IN|POST]SESSION_CMD
17:08
<ltsp_>
alkisg: Error: "PRE" is not a valid command.
17:08
<alkisg>
!learn alkisg-todo as `implement LDM_[PRE|IN|POST]SESSION_CMD`
17:08
<ltsp_>
alkisg: The operation succeeded.
17:08
<alkisg>
!alkisg-todo
17:08
<ltsp_>
alkisg: alkisg-todo: (#1) ltsp-update-image.symlinks, or (#2) put LDM_PRINTER_LIST example in lts.conf, or (#3) package a newer version of ltsp-docs in the Greek schools PPA, or (#4) upload fixed nbd-server package in the PPA, or (#5) implement LDM_[PRE|IN|POST]SESSION_CMD
17:12
<vagrantc>
alkisg: hrm. epoptes isn't showing the desktop thumbnail images
17:14
<alkisg>
vagrantc: hmmm run epoptes from the console on the server to see any error messages, and from an xterm on the client
17:15
<vagrantc>
alkisg: epoptes-client isn't running as the user.
17:15
<stgraber>
alkisg: comma separated list of printers to allow
17:15
<alkisg>
stgraber: where can one see the printer names to use?
17:16
<stgraber>
alkisg: lpstat -a
17:16
<alkisg>
ty
17:16
<vagrantc>
alkisg: it's trying to connect to "server" but there's no hosts alias for "server" on the server.
17:17
<alkisg>
vagrantc: epoptes-client should connect to localhost for thin clients
17:18
<vagrantc>
alkisg: it's running on the thin client itself, but seems to be failing to run from the server at login time
17:20
<alkisg>
vagrantc: it runs from /etc/xdg/autostart/epoptes-client.desktop
17:20
...LXDE does run the scripts there, doesn't it?
17:20
<vagrantc>
alkisg: it thinks it's a fat client
17:20
<alkisg>
Hrm
17:20
<vagrantc>
alkisg: oh, i bet if i uninstall ltsp-client* from the server it will not.
17:21
<alkisg>
Let me check the code...
17:21
<vagrantc>
yup, that did it.
17:22
so it doesn't appear to work correctly if using the ltsp-pnp type stuff
17:22
<alkisg>
Right, it needs fixing, to check the xsessions dir
17:22
Or /var/cache/ltsp/ltsp_config
17:22
<vagrantc>
who uses ltsp-pnp anyways? :)
17:23
<alkisg>
Doh so I'll need an epoptes upload too for the workshop :(
17:23* vagrantc happy to report bugs
17:24
<vagrantc>
alkisg: you're doing a workshop soonish?
17:25
<alkisg>
In June 21th, about 80 teachers will come to see ltsp-pnp and about 20 of them will also bring their school servers for installation
17:26
<vagrantc>
whee!
17:26
<alkisg>
vagrantc: since you already have a booted client, do you see LTSP_FATCLIENT anywhere in /var/cache/ltsp/* ?
17:26
<vagrantc>
alkisg: i've purged the install
17:26
<alkisg>
Ah, or I could just grep /proc/cmdline for init=/sbin/init-ltsp
17:27
<vagrantc>
i can reinstall.
17:27
<alkisg>
That'd do it
17:27
No I have a fat chroot ready I can test with that
17:35
<vagrantc>
alkisg: i think it's the check for /usr/bin/getltscfg
17:35
alkisg: that's what triggers the LTSP_FATCLIENT=True
17:35
<alkisg>
That's what I wanted to replace with the grep /proc/cmdline, but I wonder if that whole paragraph is needed
17:36
. ltsp_config above should read the previously stored LTSP_FATCLIENT variable...
17:36
<vagrantc>
alkisg: is it added to support older versions of ltsp ?
17:36
<alkisg>
Could be...
17:37
Ah no the . ltsp_config above is only for root
17:37
That's not correct, it should also do it for fat client user sessions too
17:39
Anyway since ltsp-pnp is only for newer ltsp versions, grepping /proc/cmdline should do it
17:39
<vagrantc>
that export clause is a very clever hack :)
17:39
<alkisg>
Hrm. Can't reproduce the problem.
17:40
vagrantc: you tested with a thin client booted from an ltsp-pnp image, right?
17:41
<vagrantc>
alkisg: . /usr/share/ltsp/ltsp_config will not work well with a non-root user ... as it tries to write to /var/cache/ltsp/*
17:41
alkisg: no, a thin client booted from a non-ltsp-pnp image with ltsp-client* installed.
17:41
alkisg: removing ltsp-client* from the server fixes it.
17:42
<alkisg>
OK at least I was able to reproduce the problem of seeing the thin client as fat
17:42
<vagrantc>
ltsp-client installed on the server.
17:42
<alkisg>
So let me check if the grep fixes it...
17:42
<vagrantc>
alkisg: it's line 80 that's triggering the problem
17:43
alkisg: LTSP_CLIENT is set, LTSP_FATCLIENT is not, UID -gt 0, -x /usr/bin/getltscfg
17:44
er, not line 80 ... hrm.
17:45
87: # LTSP_FATCLIENT may not be available in user sessions, autodetect it
17:45
if [ -n "$LTSP_CLIENT" ] && [ -z "$LTSP_FATCLIENT" ] &&
17:45
[ "$UID" -gt 0 ] && [ -x /usr/bin/getltscfg ]
17:45
<alkisg>
vagrantc: this works for me:
17:45
if [ -n "$LTSP_CLIENT" ] && [ -z "$LTSP_FATCLIENT" ] &&
17:45
[ "$UID" -gt 0 ] && egrep -qs 'ltsp|nfs|nbd' /proc/cmdline
17:45
...the silly egrep there is for older ltsp versions :(
17:47
<vagrantc>
alkisg: and that should work for older ltsp versions ... might be some corner cases where it gets triggered otherwise... (i.e. server with nfs or nbd root)
17:47
alkisg: so drop the getltscfg check?
17:48
<alkisg>
Yup
17:48
vagrantc: so you can't test if it fixes it for you, right? Should I just commit it?
17:48
<vagrantc>
alkisg: no, i can test it.
17:48
alkisg: er, did test it
17:49khildin has left IRC (khildin!~khildin@ip-80-236-193-97.dsl.scarlet.be, Read error: Operation timed out)
17:50
<vagrantc>
alkisg: could actually leave the getltscfg check in there
17:50
<alkisg>
Hmm, yes you're right
17:51
<vagrantc>
it would provide a little more protection in some corner cases...
17:51
<alkisg>
This should make it a bit even harder to hit a corner case
17:51
So, commit?
17:52
<vagrantc>
alkisg: your commit or mine? :)
17:52
<alkisg>
Fell free, it was your bug anyway :)
17:52
Take all the fame and glory!
17:52
<vagrantc>
heh!
17:53* vagrantc didn't know that && tests didn't need \ line continuation
17:55
<alkisg>
Ah, and I can use a the recipe to upload epoptes to the PPA after you tag, phiew
17:57
...and the reason it worked for me so far and it didn't for you, is that I have server=127.0.1.1 in my LTSP server's /etc/hosts...
17:59
<vagrantc>
alkisg: aha
18:08viggy_prabhu has left IRC (viggy_prabhu!~viggy@117.216.129.62, Quit: Ex-Chat)
18:11
<vagrantc>
alkisg: tagged, pushed and uploaded.
18:11
epoptes
18:33andygraybeal_ has joined IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net)
18:36vmlintu has joined IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi)
18:40
<alkisg>
vagrantc: ty, I'll upload to the ppa too
18:51bobby_C has joined IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at)
19:01
<vagrantc>
we should really get in the habit of checking what version an lts.conf variable was introduced in.
19:02
i.e. LDM_FORCE_SESSION ldm 2.2.10
19:02adrianorg_ has joined IRC (adrianorg_!~adrianorg@189.58.183.236.dynamic.adsl.gvt.net.br)
19:02
<vagrantc>
and including that in ltsp-docs
19:03
<alkisg>
We could probably write a script to get the ltsp versions where each var was introduced
19:04
<vagrantc>
iterate through the tags and check for presence of each variable?
19:04
<alkisg>
Yup
19:07
<vagrantc>
it would be a good project, that way if someone found newer docs on the internets, they would at least have some clue if it actually applied to them ... that was one of the problems with the old ltsp wiki ... lots of pages, no idea what was relevent.
19:12
!vagrant-todo
19:12
<ltsp_>
vagrantc: vagrant-todo: (#1) pxelinux configuration, or (#2) pxemenus, or (#3) ltsp-tool.conf.d
19:13
<vagrantc>
that #1 intimidates me, for some reason.
19:13
(and #1 is a dependency for #2, really)
19:14Lns has left IRC (Lns!~Lns@pdpc/supporter/professional/lns, Ping timeout: 244 seconds)
19:15andygraybeal_ has left IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net, Remote host closed the connection)
19:26
<alkisg>
vagrantc: I think we had it all sorted out though, didn't we?
19:26
...the pxelinux/menus design...
19:27* alkisg hopes him and vagrantc don't forget what they settled on, and have to re-design it in the future... :P
19:27
<vagrantc>
alkisg: i'm looking at the pad, and i see a bunch of variables and a touch for pseudocode
19:27
<alkisg>
vagrantc: no no
19:27
Those are old notes
19:28
<vagrantc>
oh dear
19:28
<alkisg>
We settled on something more elegant...
19:28
<vagrantc>
did we?
19:28
<alkisg>
Yeah
19:28* alkisg tries to remember...
19:28* vagrantc too
19:28* vagrantc chuckles
19:28
<alkisg>
OK it was the includes stuff
19:28
I.e. we generate some pxelinux.cfg/* files and include them from default,
19:28
<vagrantc>
the includes stuff seems like an implementation detail
19:28
<alkisg>
and we only regenerate the ltsp one
19:28
No, because then we don't need a configuration file
19:28
<vagrantc>
ah, right.
19:29
<alkisg>
We tell the user to directly edit the pxelinux files himself
19:29
Header, footer, everything go in the default
19:29
Then we talked about the ordering of the kernels
19:30
Let me find the logs...
19:30
<vagrantc>
but we do regenerate the included stuff, yes?
19:30
<alkisg>
Only the ltsp and another one, not all of them
19:31
<vagrantc>
so we regenerate the ltsp menu as we see fit
19:32
<alkisg>
vagrantc: http://irclogs.ltsp.org/?d=2012-06-02 => 06:32 and beyond
19:33
<vagrantc>
makes my head spin
19:38
alkisg: so with all those INCLUDES= we were talking about the client-side update-kernels, or the server-side ltsp-update-kernels?
19:39
<alkisg>
vagrantc: client side
19:39
Give me a few minutes to finish some things and I'll try to sum it up
19:46ricotz has left IRC (ricotz!~rico@unaffiliated/ricotz, Quit: Ex-Chat)
19:48
<alkisg>
erm actually it was from 21:40 and below :(
19:50
<vagrantc>
hah. my head spun then too
19:52
<alkisg>
Ah and then we got ifcpu in the mix and settled on PAE_LIST_KERNELS etc
19:52
OK let's go to pad
19:52
!pad
19:52
<ltsp_>
alkisg: pad: http://pad.ubuntu-uk.org/ltsp
19:59Parker955_Away is now known as Parker955
20:02
<vagrantc>
alkisg: so, they would modify the version in the chroot, or in the tftp dir?
20:02
alkisg: for the "generate once" files?
20:03
<alkisg>
vagrantc: update-kernels writes to CHROOT/boot/pxelinux.cfg/*
20:04
<vagrantc>
ltsp-update-kernels just blindly copies those files to the tftp dir...
20:04
<alkisg>
Then, ltsp-update-kernels copies all of them (multiple chroots etc), modifies the path to match the tftp subdir, and writes them to TFTP
20:04
And, writes a master pxelinux.cfg/ltsp file that sources all of them
20:06
<vagrantc>
we still have the problem of the pxelinux.cfg/memtest file being editable in two places ...
20:06
if they modify the chroot, should it overwrite the tftp dir, and if they modify the tftp dir, tthe chroot shouldn't overwrite it.
20:06
<alkisg>
ltsp-update-kernels can comment the copied files appropriately
20:06
<vagrantc>
both of which say it's ok to edit
20:07
<alkisg>
# This file was copied from this chroot:
20:07
# if you change it, it'll take effect immediately, but it will be overwritten etc...
20:08
<vagrantc>
so ltsp-update-kernels needs to get significantly smarter.
20:08
<alkisg>
A bit, yeah... the idea is for the chroot to not need to know its name
20:09
<vagrantc>
and still be able to override rootpath provided from dhcp and such?
20:09
<alkisg>
And the server (ltsp-update-kernels) should only do basic stuff, like commenting a few files, changing the path for kernel and initrd=, and including them to the master menu
20:09
(11:09:27 μμ) vagrantc: and still be able to override rootpath provided from dhcp and such? ==> you mean nbdroot=xxx? Sure
20:11
<vagrantc>
right.
20:11
or nfsroot
20:12
<alkisg>
But for that, the user is responsible, so we don't need to care about what path he puts there
20:12
ltsp-update-kernels will only change the kernel and initrd path in the tftp, nothing more
20:13
<vagrantc>
ok.
20:13
<alkisg>
And finally, we'll need ifcpu64 on the master pxelinux.cfg/ltsp too
20:14
<vagrantc>
alkisg: so ltsp-update-kernels will generate a higher-level pxelinux.cfg ... like /var/lib/tftpboot/ltsp/pxelinux.cfg/default vs. /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default ?
20:14
<alkisg>
vagrantc: right, or even higher, tftpboot/pxelinux.cfg/default
20:14Parker955 is now known as Parker955_Away
20:14
<alkisg>
Since it's user editable, so he can put other stuff there too
20:15
<vagrantc>
sure
20:15* vagrantc wants to focus on the chroot-side update-kernels first
20:16
<alkisg>
ltsp-update-kernels will try to match the chroot name with the arch, but the user will again be able to specify his own list for "which chroot to select for 64, pae, 32 bit"
20:16
OK
20:17
<vagrantc>
and just to keep myself distracted, i wonder how PXE for arm will work...
20:17
since it's apparently possible after all
20:18
<alkisg>
Maybe pxelinux will grow a bigger "ifcpu" module
20:19* vagrantc was pretty sure pxelinux is actually x86 code
20:19
<vagrantc>
even if PXE isn't
20:27
alkisg: ok, grabbed your list_kernels function from that paste way back when
20:27
and can generate a menu entry for NFS, NBD and FOOBAR...
20:28
for each kernel installed
20:28andygraybeal_ has joined IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net)
20:36
<alkisg>
I wrote some stuff to the pad for the server side too
20:53
<vagrantc>
alkisg: you haven't abandoned the CMDLINE_LINUX CMDLINE_LINUX_DEFAULT and CMDLINE_NFS CMDLINE_NBD idea, have you?
20:53
or BOOT_METHODS?
20:54* alkisg tries to remember... damn multitasking too much for a single-core mind... :(
21:00
<vagrantc>
for whatever reason, this particular TODO item swiftly overwhelms my poor mind.
21:02
<alkisg>
vagrantc: if you leave it for post-wheezy, I'll be able to help, but I'll be too busy with sch-scripts for the next couple of weeks to help till then...
21:02
<vagrantc>
sure
21:07bobby_C has left IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at, Ping timeout: 248 seconds)
21:14
<alkisg>
No I still think those are a good idea
21:14
It can still work with the same layout, even if it needs... triple loops :P
21:14
They'd all go inside the ltsp-old submenu
21:15mikkel has left IRC (mikkel!~mikkel@80-71-132-15.u.parknet.dk, Quit: Leaving)
21:57komunista has left IRC (komunista!~slavko@adsl-195-168-249-156.dynamic.nextra.sk, Quit: Leaving.)
22:05gentgeen__ has left IRC (gentgeen__!~kevin@c-98-236-71-64.hsd1.pa.comcast.net, Read error: Operation timed out)
22:08alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
22:13lefteris_nik has left IRC (lefteris_nik!~lefteris@ppp141237208190.dsl.hol.gr, Quit: Leaving.)
22:13lefteris_nik has joined IRC (lefteris_nik!~lefteris@ppp141237208190.dsl.hol.gr)
22:25Lns has joined IRC (Lns!~Lns@pdpc/supporter/professional/lns)
22:32
<vagrantc>
hmmmm... include files ain't working out so well.
22:55andygraybeal_ has left IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net, Quit: Ex-Chat)
22:56instalador_deses has joined IRC (instalador_deses!~instalado@Wimax-Cali-190-0-0-222.orbitel.net.co)
22:56
<instalador_deses>
Hi
22:56
i have a problem with my thin clients
22:56
they refuse to log in randomly
22:56
it is almost half of the time
22:56
<vagrantc>
you get a login screen, but it fails to login?
22:57
or they fail to boot to the login screen?
22:57
<instalador_deses>
i get the ldm screen
22:57
then a password prompt
22:57
and then i get the ubuntu wheel going forever
22:57
if i log into the server and pkill all the users processes i can log in
22:58
the processes for the user that tried to log in
23:00andygraybeal_ has joined IRC (andygraybeal_!~andy@h156.214.213.151.dynamic.ip.windstream.net)
23:01
<instalador_deses>
any idea of where to look?... could this be related to nbd? ...
23:01
i am using ubuntu 12.04
23:01
maybe dbus?
23:13
could it be ubuntu 12.04? ... i am running 10.04 in other server and it rocks
23:16
<vagrantc>
instalador_deses: there may be errors in the user's ~/.xsession-errors file that indicate what the problem is
23:16
a likely candidate is compiz:
23:16
!disable_compiz
23:16
<ltsp_>
vagrantc: disable_compiz: To disable compiz for all users, run: sudo gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.mandatory --type string --set /desktop/gnome/session/required_components/windowmanager metacity
23:17
<instalador_deses>
i am using gnome-session --session=ubuntu-2d
23:19
<vagrantc>
using LDM_SESSION ?
23:19
<instalador_deses>
yes
23:22
vagrantc, xsession-errors at http://notepad.cc/qogeasu59
23:23
and then it pauses
23:24
vagrantc, i updated http://notepad.cc/qogeasu59 with the output from ps -ef | grep username
23:27F-GT has joined IRC (F-GT!~phantom@ppp121-44-97-140.lns20.syd6.internode.on.net)
23:29
<vagrantc>
instalador_deses: something is calling xhost, which is a really bad idea...
23:31
<instalador_deses>
vagrantc, do you think this could be the problem?
23:33
<vagrantc>
instalador_deses: is the user logged in from the local console as well as logging into LTSP ? a user can only really be logged in from one place at a time, unfortunately.
23:34
<instalador_deses>
no, the user is logged in just from one place
23:35
i log in, then i log out of the server
23:35
when the process halts i run pkill -u username to force the user out
23:35
then i am able to log in
23:36
i just pkill the processes and the login attemp failed
23:45tarksosin has joined IRC (tarksosin!~tarksosin@wsip-24-234-29-30.lv.lv.cox.net)
23:45instalador_deses has left IRC (instalador_deses!~instalado@Wimax-Cali-190-0-0-222.orbitel.net.co, Quit: Saliendo)
23:48
<tarksosin>
Howdy folks, got a question that's probably going to have way too much information in it, but here goes.
23:50
I initially played with LTSP about two years ago in my wife's school's computer lab. It didn't suit their needs, so we ended up installing directly to the clients. This summer, I want to give LTSP another shot. I've installed Edubuntu 12.04 with the LTSP option checked onto what's going to be her server. However, googlefu is failing me and I'm not really finding any documentation for getting past that point. Anyone have any pages
23:50
I can read, videos I can watch, or anything of that nature?