IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 13 February 2012 (all times are UTC)
| 00:20 | monteslu_ has joined IRC (monteslu_!~monteslu@ip68-109-174-213.ph.ph.cox.net) | |
| 00:23 | monteslu has left IRC (monteslu!~monteslu@ip68-109-174-213.ph.ph.cox.net, Ping timeout: 260 seconds) | |
| 00:36 | andy__ has left IRC (andy__!~andy@h247.95.31.71.dynamic.ip.windstream.net, Ping timeout: 240 seconds) | |
| 00:36 | bengoa has joined IRC (bengoa!~bengoa@187-27-63-60.3g.claro.net.br) | |
| 00:36 | bengoa has left IRC (bengoa!~bengoa@187-27-63-60.3g.claro.net.br) | |
| 01:34 | andy__ has joined IRC (andy__!~andy@h247.95.31.71.dynamic.ip.windstream.net) | |
| 01:39 | Yet_another_Bill has joined IRC (Yet_another_Bill!billy@nat/redhat/x-qvxkqnyyleqdsfof) | |
| 02:15 | ba has joined IRC (ba!~ba@airport.kg) | |
| 02:28 | ba has left IRC (ba!~ba@airport.kg) | |
| 02:28 | adrianorg_ has left IRC (adrianorg_!~adrianorg@187.115.110.202, Ping timeout: 252 seconds) | |
| 02:49 | loather-work has joined IRC (loather-work!~khudson@wsip-98-175-250-115.sd.sd.cox.net) | |
| 03:19 | killermike has joined IRC (killermike!~killermik@2.26.100.154) | |
| 03:26 | Kevin` has left IRC (Kevin`!~kevin@router.kwzs.be, Ping timeout: 248 seconds) | |
| 03:27 | sndwrx has left IRC (sndwrx!~sndwrx@unaffiliated/sndwrx, Ping timeout: 248 seconds) | |
| 03:27 | loather has left IRC (loather!~khudson@wsip-98-175-250-115.sd.sd.cox.net, Ping timeout: 247 seconds) | |
| 03:27 | F-GT has left IRC (F-GT!~phantom@ppp121-44-145-103.lns20.syd7.internode.on.net, Ping timeout: 247 seconds) | |
| 03:27 | dberkholz has left IRC (dberkholz!~dberkholz@gentoo/developer/dberkholz, Ping timeout: 247 seconds) | |
| 03:27 | dberkholz has joined IRC (dberkholz!user47699@smtp.gentoo.org) | |
| 03:27 | dberkholz has joined IRC (dberkholz!user47699@gentoo/developer/dberkholz) | |
| 03:27 | sndwrx has joined IRC (sndwrx!~sndwrx@unaffiliated/sndwrx) | |
| 03:28 | Kevin` has joined IRC (Kevin`!~kevin@router.kwzs.be) | |
| 03:28 | loather has joined IRC (loather!~khudson@wsip-98-175-250-115.sd.sd.cox.net) | |
| 03:29 | F-GT has joined IRC (F-GT!~phantom@ppp121-44-145-103.lns20.syd7.internode.on.net) | |
| 04:29 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 04:44 | loather-work has left IRC (loather-work!~khudson@wsip-98-175-250-115.sd.sd.cox.net, Quit: This computer has gone to sleep) | |
| 05:09 | freedomrun has joined IRC (freedomrun!~quassel@89.143.177.168) | |
| 05:17 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.) | |
| 06:15 | loather-work has joined IRC (loather-work!~khudson@wsip-98-175-250-115.sd.sd.cox.net) | |
| 06:24 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 07:01 | sep has joined IRC (sep!~sep@40.211.jostedal.no) | |
| 07:05 | <alkisg> I'm trying to force lockscreen=false for fat clients with the new gsettings system... if anyone knows how to do mandatory settings with dconf, I'd appreciate any help...
| |
| 07:29 | <cyberorg> alkisg, find /usr/share/glib-2.0/schemas/ |grep override
| |
| 07:29 | create your own override in there and run glib-compile-schemas /usr/share/glib-2.0/schemas/
| |
| 07:29 | <alkisg> cyberorg: those are vendor overrides for default values, not mandatory settings
| |
| 07:29 | So the user is still able to change them
| |
| 07:30 | <cyberorg> don't know about mandatory settings
| |
| 07:30 | <alkisg> Thank you though :)
| |
| 07:31 | There's a "lockdown" paragraph in http://live.gnome.org/dconf/SystemAdministrators, but it doesn't seem to apply to precise, e.g.:
| |
| 07:31 | $ dconf update
| |
| 07:31 | fatal: Error opening directory '/etc/dconf/db': No such file or directory
| |
| 08:17 | loather has left IRC (loather!~khudson@wsip-98-175-250-115.sd.sd.cox.net, Quit: This computer has gone to sleep) | |
| 08:20 | killermike has left IRC (killermike!~killermik@2.26.100.154, Read error: Connection reset by peer) | |
| 08:20 | killermike_ has joined IRC (killermike_!~killermik@2.26.91.251) | |
| 08:26 | killermike_ has left IRC (killermike_!~killermik@2.26.91.251, Remote host closed the connection) | |
| 08:26 | killermike_ has joined IRC (killermike_!~killermik@2.26.91.251) | |
| 08:35 | killermike_ has left IRC (killermike_!~killermik@2.26.91.251, Remote host closed the connection) | |
| 08:36 | killermike_ has joined IRC (killermike_!~killermik@2.26.91.251) | |
| 08:40 | <alkisg> Does Ubuntu 11.10 use /etc/nbd-server/config instead of inetd? If so, can someone check if the nbd-server processes only die 2 hours after the client is powered off? https://bugs.launchpad.net/ubuntu/+source/nbd/+bug/696454
| |
| 08:41 | (also I wonder if it wouldn't be possible to only have 1 nbd-server process, it doesn't make sense for ltsp to have one for each user)
| |
| 08:43 | killermike_ has left IRC (killermike_!~killermik@2.26.91.251, Remote host closed the connection) | |
| 08:44 | killermike_ has joined IRC (killermike_!~killermik@2.26.91.251) | |
| 08:47 | toscalix has joined IRC (toscalix!~toscalix@85.137.146.26.dyn.user.ono.com) | |
| 08:48 | dobber has joined IRC (dobber!~dobber@213.169.45.222) | |
| 09:01 | loather-work has left IRC (loather-work!~khudson@wsip-98-175-250-115.sd.sd.cox.net, Quit: This computer has gone to sleep) | |
| 09:32 | killermike__ has joined IRC (killermike__!~killermik@2.26.103.131) | |
| 09:34 | killermike_ has left IRC (killermike_!~killermik@2.26.91.251, Ping timeout: 252 seconds) | |
| 09:36 | ogra_ has joined IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de) | |
| 09:36 | Trixboxer has joined IRC (Trixboxer!~Trixboxer@115.124.115.71) | |
| 09:40 | Gremble has joined IRC (Gremble!~Ben@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com) | |
| 09:55 | adrianorg_ has joined IRC (adrianorg_!~adrianorg@187.115.110.202) | |
| 09:56 | killermike__ has left IRC (killermike__!~killermik@2.26.103.131, *.net *.split) | |
| 09:56 | LoveStorm has left IRC (LoveStorm!Storm@gateway/shell/trekweb.org/x-zeihxpukjlfjvjrw, *.net *.split) | |
| 09:56 | yanu_ has left IRC (yanu_!~yanu@178-117-229-179.access.telenet.be, *.net *.split) | |
| 09:56 | zoobab has left IRC (zoobab!zoobab@vic.ffii.org, *.net *.split) | |
| 09:56 | Hyperbyte has left IRC (Hyperbyte!jan@middelkoop.cc, *.net *.split) | |
| 09:56 | epaphus2 has left IRC (epaphus2!~epaphus@200.122.149.9, *.net *.split) | |
| 09:56 | leio has left IRC (leio!~leio@gentoo/developer/leio, *.net *.split) | |
| 09:56 | shogunx has left IRC (shogunx!~shogunx@rrcs-67-79-182-232.se.biz.rr.com, *.net *.split) | |
| 09:59 | killermike__ has joined IRC (killermike__!~killermik@2.26.103.131) | |
| 09:59 | LoveStorm has joined IRC (LoveStorm!Storm@gateway/shell/trekweb.org/x-zeihxpukjlfjvjrw) | |
| 09:59 | yanu_ has joined IRC (yanu_!~yanu@178-117-229-179.access.telenet.be) | |
| 09:59 | zoobab has joined IRC (zoobab!zoobab@vic.ffii.org) | |
| 09:59 | Hyperbyte has joined IRC (Hyperbyte!jan@middelkoop.cc) | |
| 10:00 | epaphus2 has joined IRC (epaphus2!~epaphus@200.122.149.9) | |
| 10:00 | leio has joined IRC (leio!~leio@gentoo/developer/leio) | |
| 10:00 | shogunx has joined IRC (shogunx!~shogunx@rrcs-67-79-182-232.se.biz.rr.com) | |
| 10:09 | freedomrun has left IRC (freedomrun!~quassel@89.143.177.168, Remote host closed the connection) | |
| 10:18 | alexqwesa_ has left IRC (alexqwesa_!~alex@alexo-veto.broker.freenet6.net, Ping timeout: 245 seconds) | |
| 10:18 | alexqwesa_ has joined IRC (alexqwesa_!~alex@109.172.15.11) | |
| 10:37 | epaphus2 has left IRC (epaphus2!~epaphus@200.122.149.9, *.net *.split) | |
| 10:37 | leio has left IRC (leio!~leio@gentoo/developer/leio, *.net *.split) | |
| 10:37 | shogunx has left IRC (shogunx!~shogunx@rrcs-67-79-182-232.se.biz.rr.com, *.net *.split) | |
| 10:39 | epaphus2 has joined IRC (epaphus2!~epaphus@200.122.149.9) | |
| 10:39 | leio has joined IRC (leio!~leio@gentoo/developer/leio) | |
| 10:39 | shogunx has joined IRC (shogunx!~shogunx@rrcs-67-79-182-232.se.biz.rr.com) | |
| 10:44 | alexqwesa__ has joined IRC (alexqwesa__!~alex@109.172.15.11) | |
| 10:44 | alexqwesa_ has left IRC (alexqwesa_!~alex@109.172.15.11, Read error: Connection reset by peer) | |
| 11:04 | Gremble has left IRC (Gremble!~Ben@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com, Quit: I Leave) | |
| 11:35 | artista-frustrad has joined IRC (artista-frustrad!~fernando@200.247.43.2) | |
| 11:53 | Yet_another_Bill has left IRC (Yet_another_Bill!billy@nat/redhat/x-qvxkqnyyleqdsfof, Quit: Good night ^_^) | |
| 11:55 | [GuS] has joined IRC ([GuS]!~MysT@213-117-16-190.fibertel.com.ar) | |
| 11:55 | [GuS] has joined IRC ([GuS]!~MysT@unaffiliated/gus/x-663402) | |
| 11:56 | * [GuS] good morning | |
| 11:57 | Yet_another_Bill has joined IRC (Yet_another_Bill!billy@nat/redhat/x-fnjhqkvoytonlyay) | |
| 12:08 | freedomrun has joined IRC (freedomrun!~quassel@BSN-142-163-63.dial-up.dsl.siol.net) | |
| 12:15 | cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 240 seconds) | |
| 12:17 | cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg) | |
| 12:52 | |GuS| has joined IRC (|GuS|!~MysT@213-117-16-190.fibertel.com.ar) | |
| 12:52 | |GuS| has joined IRC (|GuS|!~MysT@unaffiliated/gus/x-663402) | |
| 12:53 | [GuS] has left IRC ([GuS]!~MysT@unaffiliated/gus/x-663402, Ping timeout: 260 seconds) | |
| 12:54 | |GuS| is now known as [GuS] | |
| 12:57 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Ping timeout: 240 seconds) | |
| 13:02 | <knipwim-work> [GuS]: hey gus
| |
| 13:03 | <[GuS]> knipwim-work: hey! hello
| |
| 13:03 | brunolambert has joined IRC (brunolambert!bruno@nat/revolutionlinux/x-elepbncgjalibqit) | |
| 13:05 | <[GuS]> knipwim-work: i didnt tested yet that, since i dont have right now a thin clien to test. I wil test it later creating a vm machine to boot from net.
| |
| 13:08 | brunolambert has left IRC (brunolambert!bruno@nat/revolutionlinux/x-elepbncgjalibqit, Read error: Connection reset by peer) | |
| 13:13 | veloutin has joined IRC (veloutin!~veloutin@modemcable121.135-59-74.mc.videotron.ca) | |
| 13:13 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 13:27 | brunolambert has joined IRC (brunolambert!bruno@nat/revolutionlinux/x-fdswqrknvalqgjyc) | |
| 13:29 | khildin has joined IRC (khildin!~khildin@ip-80-236-212-243.dsl.scarlet.be) | |
| 13:34 | <knipwim-work> [GuS]: cool, np
| |
| 13:44 | bengoa has joined IRC (bengoa!~bengoa@2001:1291:229:2:216:cbff:feab:6cc9) | |
| 13:44 | mgariepy has joined IRC (mgariepy!mgariepy@ubuntu/member/mgariepy) | |
| 13:45 | <mgariepy> good morning everyone
| |
| 13:58 | <[GuS]> mgariepy: morning
| |
| 14:58 | Gremble has joined IRC (Gremble!~Ben@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com) | |
| 14:58 | killermike__ has left IRC (killermike__!~killermik@2.26.103.131, Remote host closed the connection) | |
| 15:00 | <mgariepy> stgraber, ping
| |
| 15:00 | <stgraber> mgariepy: pong
| |
| 15:01 | <mgariepy> when updating my chroot ,disconnect all NBD devices on stop ?
| |
| 15:01 | why does it ask?
| |
| 15:01 | <stgraber> ?
| |
| 15:02 | not sure I understand the question
| |
| 15:02 | <mgariepy> http://paste.ubuntu.com/840469/
| |
| 15:03 | i'm updating the chroot, is that message normal on precise ?
| |
| 15:03 | <stgraber> well, it's a debconf critical question, so unless you've set all the magic environment variables, it's going to prompt
| |
| 15:03 | it's not new in Precise, it happens everytime you upgrade the nbd package
| |
| 15:03 | which doesn't happen too often
| |
| 15:03 | <mgariepy> ok
| |
| 15:03 | first time i see that ;)
| |
| 15:09 | [GuS] has left IRC ([GuS]!~MysT@unaffiliated/gus/x-663402, Remote host closed the connection) | |
| 15:10 | <alkisg> stgraber, mgariepy: udhcp runs in init-premount, and nbd (which calls the configure_networking function we worry about) in local-top, which comes later
| |
| 15:10 | So I'm thinking we should go for it and completely switch to the new system, letting nbd mount the root for us
| |
| 15:11 | mgariepy: if I write an NBDCLIENT wrapper which calls nbd-proxy, could you test that it works for your needs?
| |
| 15:12 | With that new system, we'll have udhcp + nbd + init-bottom/ltsp (which does the cow magic) in the initramfs. ltsp_nbd won't be there anymore.
| |
| 15:12 | ...and we'll be using the nbdroot=ip:port/path syntax
| |
| 15:16 | [GuS] has joined IRC ([GuS]!~MysT@unaffiliated/gus/x-663402) | |
| 15:19 | <mgariepy> alkisg, yeah i can test no problem
| |
| 15:20 | i just updated my chroot and the system doesn't boot. :S
| |
| 15:20 | <alkisg> mgariepy: precise?
| |
| 15:20 | <mgariepy> yeah
| |
| 15:20 | nbd doesn't mount
| |
| 15:20 | <alkisg> Oh
| |
| 15:21 | <mgariepy> bzr 2025
| |
| 15:21 | <alkisg> After nbd mounts, you'll need init=/usr/share/ltsp/initramfs-scripts in pxelinux.cfg/default too
| |
| 15:21 | But I don't think we've changed anything wrt nbd mounting...
| |
| 15:21 | Do you get an initramfs shell
| |
| 15:21 | ?
| |
| 15:21 | <mgariepy> yes
| |
| 15:21 | <alkisg> nbd-client server-ip -N ltsp_i386
| |
| 15:22 | ...what does that say?
| |
| 15:22 | <mgariepy> wait a sec
| |
| 15:24 | <alkisg> mgariepy: I forgot the device: nbd-client server-ip /dev/nbd0 -N ltsp_i386
| |
| 15:25 | <mgariepy> it works
| |
| 15:25 | <alkisg> mgariepy: what do you have in pxelinux.cfg/default?
| |
| 15:26 | <mgariepy> now : append ro initrd=new/initrd.img init=/usr/share/ltsp/initramfs-scripts nbdname=ltsp_new verbose debug
| |
| 15:26 | i'll test that
| |
| 15:26 | <alkisg> mgariepy: ah, also put: boot=ltsp_nbd
| |
| 15:26 | That's probably the problem
| |
| 15:28 | If we switch to using the local-top/nbd script, that won't be needed anymore
| |
| 15:31 | dead_inside has joined IRC (dead_inside!~dead_insi@76.75.3.174) | |
| 15:32 | <mgariepy> this works : append ro initrd=new/initrd.img nbdname=ltsp_new verbose debug boot=ltsp_nbd
| |
| 15:32 | <alkisg> mgariepy: ok it's on the http://pad.ubuntu.com/ltsp-server-precise notes... we'll fix it soon
| |
| 15:33 | <mgariepy> yeah no problem
| |
| 15:33 | <alkisg> If stgraber doesn't mind, I'll do the ltsp_nbd => local-top/nbd change until tomorrow evening, and then we'll only have bug fixing left :D
| |
| 15:33 | <mgariepy> for nbd-proxy i only use it for /rofs (not for swap)
| |
| 15:34 | <alkisg> I'll make a wrapper for it, using the NBDCLIENT env variable that nbd supports
| |
| 15:34 | <stgraber> alkisg: ok, the sooner the better really, I have quite a bunch of stuff to get done by Thursday ;)
| |
| 15:34 | <alkisg> stgraber: I'd have done those months ago, but I thought we weren't going to switch to initramfs-scripts.d for precise... I think it's good that we'll do, though :)
| |
| 15:34 | Gotta go, bbl, will read logs
| |
| 15:35 | <Hyperbyte> Bye alkisg! :)
| |
| 15:35 | <mgariepy> cya thanks for your help alexqwesa__
| |
| 15:35 | alkisg *
| |
| 15:35 | <alkisg> bb all
| |
| 15:35 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.) | |
| 15:35 | epaphus2 has left IRC (epaphus2!~epaphus@200.122.149.9, Quit: Leaving) | |
| 15:45 | Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com) | |
| 15:59 | laurei has joined IRC (laurei!3a069039@gateway/web/freenode/ip.58.6.144.57) | |
| 16:00 | <laurei> does anyone know the correct groups users need to be in to mount DVD's in fat clients?
| |
| 16:11 | monteslu_ is now known as monteslu | |
| 16:17 | staffencasa_ has joined IRC (staffencasa_!~staffenca@128-193-149-96.oregonstate.edu) | |
| 16:24 | laurei has left IRC (laurei!3a069039@gateway/web/freenode/ip.58.6.144.57, Quit: Page closed) | |
| 16:27 | dead_inside has left IRC (dead_inside!~dead_insi@76.75.3.174, Ping timeout: 244 seconds) | |
| 16:27 | <mmetzger> Any good guides on setting up printing for local apps clients?
| |
| 16:41 | <knipwim> mmetzger: https://help.ubuntu.com/community/UbuntuLTSP/LocalAppsLucidPrinting ?
| |
| 16:43 | ogra_ has left IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de, Read error: Operation timed out) | |
| 16:45 | ogra_ has joined IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de) | |
| 16:46 | <mmetzger> knipwim: Thanks - that's mostly what I've done so far - trying to make sure the printer(s) show up automatically for the clients without changing specifically to the print server (I have a *very* weird environment)
| |
| 16:49 | dobber has left IRC (dobber!~dobber@213.169.45.222, Remote host closed the connection) | |
| 16:56 | khildin has left IRC (khildin!~khildin@ip-80-236-212-243.dsl.scarlet.be, Ping timeout: 276 seconds) | |
| 17:36 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 17:37 | freedomrun_ has joined IRC (freedomrun_!~quassel@BSN-142-163-128.dial-up.dsl.siol.net) | |
| 17:38 | freedomrun has left IRC (freedomrun!~quassel@BSN-142-163-63.dial-up.dsl.siol.net, Ping timeout: 244 seconds) | |
| 17:59 | freedomrun_ has left IRC (freedomrun_!~quassel@BSN-142-163-128.dial-up.dsl.siol.net, Remote host closed the connection) | |
| 18:04 | brunolambert has left IRC (brunolambert!bruno@nat/revolutionlinux/x-fdswqrknvalqgjyc, Quit: brunolambert) | |
| 18:04 | brunolambert has joined IRC (brunolambert!bruno@nat/revolutionlinux/x-ivvhwiwhynnvpxog) | |
| 18:19 | dptech has joined IRC (dptech!~dptech@can06-1-82-242-223-39.fbx.proxad.net) | |
| 18:20 | komunista has joined IRC (komunista!~slavko@adsl-195-098-005-032.dynamic.nextra.sk) | |
| 18:28 | Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Quit: Leaving) | |
| 18:33 | vagrantc has joined IRC (vagrantc!~vagrant@c-76-115-60-19.hsd1.or.comcast.net) | |
| 18:34 | * vagrantc wonders how ltsp-trunk is doing | |
| 18:38 | Gremble has left IRC (Gremble!~Ben@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com, Quit: I Leave) | |
| 18:38 | <alkisg> vagrantc: I'm about to remove ltsp_nbd... :)
| |
| 18:39 | So if you want to do any testing, please wait for some hours
| |
| 18:39 | I'm trying to make the switch to using the upstream nbd script
| |
| 18:40 | <vagrantc> remove ltsp_nbd! awesome!
| |
| 18:41 | alkisg: so you've got initramfs-scripts capable of running as /sbin/init ?
| |
| 18:41 | <alkisg> So the new command line for nbd will look something like: append ro initrd=initrd.img nbdroot=ip:port/path init=/sbin/ltsp-init
| |
| 18:41 | vagrantc: yup, I really think we should rename them to ltsp-init.d
| |
| 18:41 | And move the /usr/share/ltsp/initramfs-scripts to /sbin/ltsp-init
| |
| 18:41 | <vagrantc> should be able to specify nbdroot=dhcp
| |
| 18:41 | <alkisg> Yes Wouter has that covered
| |
| 18:41 | <vagrantc> ok, good.
| |
| 18:42 | <alkisg> Or just nbdroot=/path
| |
| 18:42 | <vagrantc> i run installs on multiple networks from the same chroot, and that's needed.
| |
| 18:42 | good good.
| |
| 18:43 | <alkisg> vagrantc: should I change the name now from initramfs-scripts.d to ltsp-init.d, so that it shows better in the commit messages in the bzr log later on?
| |
| 18:43 | <vagrantc> alkisg: you had some splash screen issues you had to work out to switch it to init ?
| |
| 18:43 | <alkisg> That was a problem in plymouth, easily bypassed
| |
| 18:43 | I.e. it looked for init=whatever in the cmdline, and disabled itself :(
| |
| 18:43 | <vagrantc> alkisg: and does it still work when called from initramfs the "old" way?
| |
| 18:44 | <alkisg> If you pass any parameter to initramfs-scripts, it works the old way
| |
| 18:44 | The only difference is the last line, if it will exec init or not
| |
| 18:44 | <vagrantc> sounds cool.
| |
| 18:45 | alkisg: if i stick with nfs+aufs+tmpfs will i need to wait for your pending changes?
| |
| 18:46 | <alkisg> vagrantc: are you using the "ltsp" command line option to know when to run the initramfs-scripts?
| |
| 18:46 | I.e. are you using nfs-bottom/ltsp ?
| |
| 18:46 | <vagrantc> alkisg: yes.
| |
| 18:46 | <alkisg> OK we need to talk about that
| |
| 18:47 | One option is for you to use the init=/sbin/ltsp-init syntax too
| |
| 18:47 | No "ltsp" magic parameter anymore, only stock parameters
| |
| 18:47 | <vagrantc> i'd like to at least give it a try ...
| |
| 18:47 | might also be easier to use with kfreebsd, if i ever get that insanity going
| |
| 18:48 | <alkisg> Errrr sorry I meant init-bottom/ltsp, not nfs-bottom/ltsp
| |
| 18:48 | Another option is to support both of the methods
| |
| 18:48 | <vagrantc> as of last friday, i was testing with init-bottom/ltsp using the ltsp bootprompt argument.
| |
| 18:48 | <alkisg> I.e. check for either "ltsp", or "init=*ltsp*", or "init=/sbin/ltsp-init", if we're going to hardcode that
| |
| 18:49 | <vagrantc> if it's not too hard to maintain compatibility for both, i kind of like supporting both.
| |
| 18:49 | that way if there are unexpected gotchas with either approach, we can easily switch to the other.
| |
| 18:49 | <alkisg> OK, the only difference is that with init=, we don't want the chroot call
| |
| 18:49 | I'll commit that now, if you want to test
| |
| 18:50 | vagrantc: what about the name change? I basically care about the init=parameter
| |
| 18:50 | Will it be init=/usr/share/ltsp/initramfs-scripts, or init=/sbin/ltsp-init ?
| |
| 18:50 | (even if it is a symlink...)
| |
| 18:51 | (or init=/sbin/init-ltsp, if we want to make it easier for plymouth to know when to NOT disable itself..)
| |
| 18:51 | <vagrantc> alkisg: i like init=/sbin/ltsp-init or init=/sbin/init-ltsp
| |
| 18:52 | keeps the commandline shorter and still reasonable clear what craziness we're implementing
| |
| 18:52 | <alkisg> Cool, thanks, I'll commit the change in 10' or so
| |
| 18:53 | rthomson has joined IRC (rthomson!~rthomson@mars.pet.ubc.ca) | |
| 18:53 | <alkisg> stgraber: do you think you could make plymouth not disable itself if it sees init=/sbin/init* ?
| |
| 18:53 | ...so that we use init=/sbin/init-ltsp, instead of init=/sbin/ltsp-init, and so that we don't have to use the force parameter?
| |
| 18:54 | I can send a patch if it helps...
| |
| 18:54 | <vagrantc> and will we be able to get that into debian...
| |
| 18:54 | alkisg: alternately, we could have /sbin/init be diverted and handle it with some other logic... ?
| |
| 18:55 | <alkisg> vagrantc: you will? Nice, so /sbin/init-ltsp it is. (or with a dot? /sbin/init.ltsp ?)
| |
| 18:55 | <vagrantc> alkisg: dunno why, i prefer the dash...
| |
| 18:55 | <alkisg> Errr sure, that would work too!
| |
| 18:55 | <stgraber> alkisg: you can attach the patch to the bug but I wouldn't count on it being merged by Thursday, I'm horribly busy with other things
| |
| 18:55 | <vagrantc> but that seems more dangerous to divery /sbin/init ... and then we're back to passing "ltsp" as an argument.
| |
| 18:56 | <stgraber> alkisg: as far as I can see, the force parameter only affects this specific bit of plymouth so it should be safe to use for now
| |
| 18:56 | <vagrantc> is plymouth in debian dn ubuntu reasonnably in sync?
| |
| 18:56 | <alkisg> vagrantc: yeah I don't really like that part about passing special kernel parameters
| |
| 18:56 | stgraber: nice
| |
| 18:56 | OK so no hurries about it
| |
| 18:56 | Let's leave the plymouth fix for later on
| |
| 18:57 | We'll use init=/sbin/init-ltsp, and try to push the plymouth fix whenever we can, and use the force parameter in the meantime
| |
| 18:57 | <vagrantc> so plymouth disables itself when init isn't /sbin/init ?
| |
| 18:57 | <alkisg> Yeah, in the changelog it says it gives problems with init=/bin/bash, that's why they disabled it
| |
| 18:57 | https://bugs.launchpad.net/ubuntu/+source/plymouth/+bug/930865
| |
| 18:57 | <vagrantc> we could replace /sbin/init from the initramfs with /sbin/init-ltsp as an alternate approach
| |
| 18:58 | then we're only doing one crazy thing in the initramfs
| |
| 18:59 | [GuS] has left IRC ([GuS]!~MysT@unaffiliated/gus/x-663402, Remote host closed the connection) | |
| 18:59 | <vagrantc> but again, probably would require a custom commandline argument.
| |
| 19:00 | <alkisg> I like the init= approach, as it cleanly states that we want to boot ltsp
| |
| 19:01 | And, here's another crazy idea about it:
| |
| 19:01 | I install a standalone client, then install ltsp-client in it,
| |
| 19:01 | khildin has joined IRC (khildin!~khildin@ip-80-236-212-243.dsl.scarlet.be) | |
| 19:01 | <alkisg> ...which creates a grub entry (like ipxe or xen) for "ltsp booting",
| |
| 19:01 | ...which contains the init=/sbin/init-ltsp parameter, which in turn uses a cow over the local disk
| |
| 19:02 | <vagrantc> love it!
| |
| 19:02 | <alkisg> A very easy way to convert local installations to ltsp fat clients on low bandwidth labs
| |
| 19:02 | We can even have a screen script to "upload the local image to the server"
| |
| 19:02 | For cases like macs
| |
| 19:03 | (or just use rsync, ok :P)
| |
| 19:04 | OK, back to coding...
| |
| 19:05 | <vagrantc> whee.
| |
| 19:10 | komunista has left IRC (komunista!~slavko@adsl-195-098-005-032.dynamic.nextra.sk, Ping timeout: 260 seconds) | |
| 19:14 | <vagrantc> alkisg: if you kill of scripts/ltsp_nbd we'll still need hooks/ltsp_nbd (or rename to "ltsp") in order to ensure the correct modules are there
| |
| 19:15 | <alkisg> vagrantc: yes, but we won't need the BOOT= line anymore
| |
| 19:15 | I think nbd will need to use that modules line itself though, so maybe in the future we won't need the conf.d/ltsp either
| |
| 19:16 | <vagrantc> alkisg: i
| |
| 19:16 | <alkisg> Wait wait
| |
| 19:16 | You were talking about the other file, moment...
| |
| 19:16 | <vagrantc> alkisg: i'm not sure DEBIAN_FRONTEND=passthrough works on debian...
| |
| 19:16 | <alkisg> It's for ubuntu only
| |
| 19:16 | I thought an "if" is better than having 2 separate files
| |
| 19:16 | <vagrantc> shouldn't need conf.d/ltsp at all (i don't think i use it in debian)
| |
| 19:17 | <alkisg> modules=netboot
| |
| 19:17 | Otherwise not all network modules go in the initramfs
| |
| 19:17 | I'll rename hooks/ltsp_nbd to just hooks/ltsp
| |
| 19:17 | <vagrantc> alkisg: i should experiment with passthrough .. it seems like the right thing to do.
| |
| 19:17 | <alkisg> vagrantc: that's only for the alternate cd, right?
| |
| 19:18 | Not for when a sysadmin runs ltsp-build-client manually...
| |
| 19:18 | <vagrantc> alkisg: there's no such thing as an alternate CD in debian
| |
| 19:18 | <alkisg> Yes, but the alternate cd maps to the normal debian cd, somewhat
| |
| 19:18 | <vagrantc> install from an installer CD
| |
| 19:18 | <alkisg> Ah, you don't have an option to install ltsp there, right?
| |
| 19:18 | <vagrantc> vs. install manually
| |
| 19:19 | not yet ...
| |
| 19:19 | <alkisg> Like, in the alternate cd, we have "f4 => install an ltsp server"
| |
| 19:19 | <vagrantc> i should see if i could pull it off for wheezy ...
| |
| 19:19 | <alkisg> The passthrough option is only used there
| |
| 19:19 | To update the progress bar etc
| |
| 19:19 | (in the debconf whiptail dialog)
| |
| 19:19 | <vagrantc> would like to have that in debian, yes.
| |
| 19:20 | <alkisg> Nice
| |
| 19:20 | If that "if" causes a problem for you, I can revert the commit
| |
| 19:20 | I just thought it'd be better to have an extra if and a smaller delta
| |
| 19:22 | <vagrantc> alkisg: where does the LTSP_CLIENT_BUILDER variable get set?
| |
| 19:22 | <alkisg> vagrantc: in ltsp-client-builder.postinst, that's not upstream, it's in the ubuntu packaging
| |
| 19:23 | <vagrantc> alkisg: and i surely have my own fork (original) of that
| |
| 19:23 | <alkisg> Hehe
| |
| 19:23 | * alkisg never uses the alternate cd so he doesn't know any details | |
| 19:23 | <vagrantc> which i haven't touched much in years ... it needs a lot of work.
| |
| 19:25 | bakytn has joined IRC (bakytn!~ba@airport.kg) | |
| 19:26 | <vagrantc> alkisg: so should i still wait for your changes?
| |
| 19:26 | <alkisg> vagrantc: just a tiny change in init-bottom/ltsp
| |
| 19:26 | To allow for calling with init=
| |
| 19:27 | But if you won't test with init=, there's no reason to wait...
| |
| 19:27 | vagrantc: in debian, does /run from the initramfs migrate to /root/run ?
| |
| 19:27 | <vagrantc> alkisg: dunno
| |
| 19:27 | <alkisg> Or, in other words, I want to copy /tmp/net-eth0.conf to the real system, where should I put that?
| |
| 19:27 | In case nfs==read only?
| |
| 19:28 | ...so that when init= is used, the initramfs-scripts will get the DEVICE etc
| |
| 19:28 | <vagrantc> alkisg: i can check
| |
| 19:30 | bakytn has left IRC (bakytn!~ba@airport.kg) | |
| 19:32 | <vagrantc> alkisg: are there legitimate cases where /sbin/init would be passed arguments?
| |
| 19:33 | <alkisg> vagrantc: I don't know, but I don't think the kernel syntax allows it. Would you prefer an environment variable?
| |
| 19:33 | NOINIT=true chroot ${rootmnt} /usr/share/ltsp/initramfs-scripts
| |
| 19:34 | * alkisg really thinks we should rename those to init-ltsp.d scripts, even if we support calling them from the initramfs... | |
| 19:34 | <vagrantc> just want to make sure we're not potentially breaking sometnhing
| |
| 19:34 | <alkisg> OK let's use an env var
| |
| 19:34 | <vagrantc> alkisg: yes, i agree with the change to init-ltsp.d
| |
| 19:34 | <alkisg> So wait for 2-3 commits :)
| |
| 19:43 | First push, now renaming...
| |
| 19:44 | <vagrantc> aalso, generally if it applies to multiple distros, prepending the commit with DISTRO: is only misleading...
| |
| 19:45 | alkisg: the dynamic hostname generation appears to be borked, as well...
| |
| 19:45 | alkisg: i get ltsp0 no matter what
| |
| 19:45 | <alkisg> I haven't seen that since Gadi moved it, but I suspect it's because of net*.conf variables do not exist at that point?
| |
| 19:46 | <vagrantc> hadn't really looked deeply into it
| |
| 19:46 | bengoa has left IRC (bengoa!~bengoa@2001:1291:229:2:216:cbff:feab:6cc9, Quit: Leaving.) | |
| 19:47 | <alkisg> vagrantc: some distros don't have initramfs... sorry I haven't done my homework about other distros, I've only tried debian+ubuntu :)
| |
| 19:47 | <vagrantc> alkisg: some of the changes were not specific to the initramfs stuff
| |
| 19:47 | <alkisg> And also many of the scripts I put in Ubuntu/ need to go to Debian/ too, but I haven't tested them there, so..
| |
| 19:48 | <vagrantc> alkisg: regarding the /run question ... at what point in the process do you want to copy a file over?
| |
| 19:49 | alkisg: i tested at the "init" phase from the initramfs, and /run didn't get moved over to /root/run, but /root/run was available and writeable.
| |
| 19:49 | * vagrantc suspects /root would technically be ${rootmnt} or some variable | |
| 19:49 | <alkisg> vagrantc: from init-bottom/ltsp
| |
| 19:50 | We basically only have that file now
| |
| 19:50 | The root mounting is done by the nfs or nbd scripts, not ours
| |
| 19:51 | <vagrantc> alkisg: and you want to copy /tmp/net-* over?
| |
| 19:51 | <alkisg> We'll only need an NBDCLIENT wrapper for nbd-proxy, but it'll only be used for the proxy stuff
| |
| 19:51 | vagrantc: yup
| |
| 19:55 | |dptech| has joined IRC (|dptech|!~dptech@can06-1-82-242-223-39.fbx.proxad.net) | |
| 19:55 | <alkisg> vagrantc: second push... can you push the net-* copy?
| |
| 19:56 | <vagrantc> alkisg: i'm testing the net-* copy stuff
| |
| 19:56 | <alkisg> Nice
| |
| 19:57 | * alkisg tests if it works without ltsp_nbd... | |
| 19:57 | dptech has left IRC (dptech!~dptech@can06-1-82-242-223-39.fbx.proxad.net, Ping timeout: 272 seconds) | |
| 19:58 | Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com) | |
| 19:59 | toscalix has left IRC (toscalix!~toscalix@85.137.146.26.dyn.user.ono.com, Read error: Connection reset by peer) | |
| 19:59 | <vagrantc> so, doing it from init-bottom, you can copy it into /run and it gets moved over
| |
| 20:00 | staffencasa_ has left IRC (staffencasa_!~staffenca@128-193-149-96.oregonstate.edu, Read error: Connection reset by peer) | |
| 20:01 | <alkisg> And it's writable even with nfs, sounds nice
| |
| 20:01 | <vagrantc> i'm thinking it would be best to copy to a sub-dir of /run ?
| |
| 20:01 | i.e. /run/ltsp/ ?
| |
| 20:01 | <alkisg> /run/cache/ltsp ?
| |
| 20:01 | Ah no cache in run
| |
| 20:01 | * alkisg thought it would be the same as /var/run... | |
| 20:02 | <alkisg> Sure, /run/ltsp. Why not put ltsp_config there too?
| |
| 20:02 | <vagrantc> on recent vdebian systems, /var/run is a ssymlink to /run
| |
| 20:02 | <alkisg> How do we handle ltsp_config ?
| |
| 20:02 | (from the initramfs, when nfs is used)
| |
| 20:02 | <vagrantc> might make backporting harder, though :(
| |
| 20:02 | the /run stuff is pretty new in debian
| |
| 20:03 | staffencasa has joined IRC (staffencasa!~staffenca@128-193-149-96.oregonstate.edu) | |
| 20:03 | * alkisg thinks net*.conf should be handled the same way as ltsp_config | |
| 20:03 | <alkisg> Or we can just get the vars we want and only write those to ltsp_config
| |
| 20:04 | <vagrantc> hmmm...
| |
| 20:04 | we've got so many moving parts, it's hard to keep track of what is where and why ...
| |
| 20:05 | <alkisg> At last we'll minimize the initramfs now
| |
| 20:05 | Remove the hostname, the swap code etc
| |
| 20:05 | *at least, not last
| |
| 20:13 | Ouch, there's a bug in the upstream nbd script... (and I'm to blame for the patch I sent... :D) /me tries to find a way around it...
| |
| 20:13 | Or not... with nbdroot=/ltsp_i386, then the name is indeed "/ltsp_i386", not just "ltsp_i386"
| |
| 20:14 | So it's ok
| |
| 20:21 | bakytn has joined IRC (bakytn!~ba@airport.kg) | |
| 20:22 | <alkisg> Is there any reason to use [ltsp_i386] instead of [/opt/ltsp/images/i386.img] in the server configuration?
| |
| 20:22 | I think it's more similar to nfs if we specify nbdroot=/opt/ltsp/images/i386.img in the kernel command line...
| |
| 20:23 | <bakytn> Hey! Any idea why xfreerdp is working only if I am using it through xterm? like SCREEN_07="xterm -e 'xfreerdp ...'" if I set the same but without xterm -e it just showing me a cursor on a black screen
| |
| 20:24 | <vagrantc> alkisg: it's nice to have shorter commandline arguments, even if it's scripted
| |
| 20:24 | <alkisg> vagrantc: for nfs, you're using root-path=/opt/ltsp/i386, right? why not use a real path for nbd too?
| |
| 20:25 | Except for that, I also have a problem with the syntax, because in the nbd patch that I sent I assumed that the name will always start with a slash
| |
| 20:26 | Otherwise, with nbdroot=ltsp_i386, that's a server name, not an export name
| |
| 20:26 | So nbdroot=/ltsp_i386 would work with the current nbd version
| |
| 20:28 | <vagrantc> alkisg: is the lts.conf downloading working for you?
| |
| 20:29 | <alkisg> I haven't reached that part in my tests yet, I'm still to the "ditch ltsp_nbd" part
| |
| 20:29 | A few hours ago it was working
| |
| 20:31 | <vagrantc> it hard-coded NBD_ROOT_HOST=server, before "server" was a resolvable hostname, at least on debian.
| |
| 20:31 | <alkisg> That syntax will work, for specifying an export name that doesn't start with a slash: nbdroot=:ltsp_i386. But it looks silly...
| |
| 20:32 | I think nbdroot=/ltsp_i386 looks better, so maybe we should put [/ltsp_i386] in the nbd-server config file
| |
| 20:32 | <vagrantc> don't think :ltsp_i386 looks any worse than /ltsp_i386 ...
| |
| 20:33 | and if it keeps the syntax more consistant, don't see why not...
| |
| 20:33 | <alkisg> The idea was to resemble the nfs syntax, ip: port / path
| |
| 20:33 | So, path usually starts with a slash, not with a colon...
| |
| 20:34 | <vagrantc> but if you're "breaking" the patch anyways...
| |
| 20:34 | i.e. it's not /ltsp_i386
| |
| 20:34 | it's /opt/ltsp/images/i386.img
| |
| 20:34 | i don't see any reason to kind of sort of mirror it
| |
| 20:35 | <alkisg> Those 3 will work with the current nbd version: nbdroot=/ltsp_i386, nbdroot=/opt/ltsp/images/i386.img, nbdroot=:ltsp_i386
| |
| 20:35 | <vagrantc> either mirror or, or do it's own thing.
| |
| 20:35 | <alkisg> So you'd go with the last one?
| |
| 20:35 | <vagrantc> i'd drop the first one
| |
| 20:35 | artista-frustrad has left IRC (artista-frustrad!~fernando@200.247.43.2, Quit: Leaving) | |
| 20:35 | <vagrantc> what configuration does it affect?
| |
| 20:36 | <alkisg> The export name in /etc/nbd-server/config
| |
| 20:36 | Which currently is [ltsp_i386]
| |
| 20:36 | ...and that works with nbdroot=:ltsp_i386
| |
| 20:36 | Or, we can put /opt/ltsp/images/i386.img everywhere
| |
| 20:36 | <vagrantc> that seems fine to me :)
| |
| 20:36 | er, [ltsp_i386] and nbdroot=:ltsp_i386 seems fine.
| |
| 20:36 | <alkisg> OK, nbdroot=:ltsp_i386 it is :)
| |
| 20:36 | ty
| |
| 20:37 | <vagrantc> probably good to add comments in the code to thhe effect of "yes, the : is really needed in order to properly support syntax ..."
| |
| 20:38 | <alkisg> So the current syntax to boot an nbd client is: append ro initrd=initrd.img nbddev=/dev/nbd0 nbdroot=:ltsp_i386 init=/sbin/init-ltsp quiet splash plymouth:force-splash
| |
| 20:38 | <vagrantc> so you workaround the plymouth bug with commandline options?
| |
| 20:39 | that's not too bad...
| |
| 20:39 | <alkisg> Yes, until hopefully we send a patch for it
| |
| 20:39 | If they can not disable the splash for init=/sbin/init*, then we're ok
| |
| 20:39 | <vagrantc> you explicitly need nbddev ?
| |
| 20:39 | <alkisg> Yes Wouter insisted on that :( (or I misunderstood him)
| |
| 20:40 | I'll check if we can skip that too
| |
| 20:40 | <vagrantc> probably to prevent it from triggering with a disked boot or something
| |
| 20:40 | although you'd think nbdroot would be enough to prevent that
| |
| 20:40 | <alkisg> He accepted root=/dev/nbd0 instead, but not just nbdrooot=x
| |
| 20:40 | <vagrantc> at some point you just go with it :)
| |
| 20:40 | <alkisg> Yeah, it's just a few bytes in the end :D
| |
| 20:41 | <vagrantc> i have had issues with "long" commandlines and grub2...
| |
| 20:41 | <alkisg> How long?
| |
| 20:41 | |dptech| has left IRC (|dptech|!~dptech@can06-1-82-242-223-39.fbx.proxad.net, Quit: When two people dream the same dream, it ceases to be an illusion. KVIrc 3.4.2 Shiny http://www.kvirc.net) | |
| 20:41 | <vagrantc> (not particularly long, really)
| |
| 20:41 | <alkisg> echo "append ro initrd=initrd.img nbddev=/dev/nbd0 nbdroot=:ltsp_i386 init=/sbin/init-ltsp quiet splash plymouth:force-splash" | wc -c
| |
| 20:41 | 120
| |
| 20:42 | I think 200+ are supported even by old pxelinux versions
| |
| 20:42 | <vagrantc> i've been meaning to file a bug about that, but it always comes up when i'm swamped with other things
| |
| 20:42 | <alkisg> (and kernel too)
| |
| 20:48 | <vagrantc> hmmm... translations for ltsp-trunk are falling behind
| |
| 20:50 | khildin has left IRC (khildin!~khildin@ip-80-236-212-243.dsl.scarlet.be, Quit: I'm gone, bye bye) | |
| 20:55 | <alkisg> vagrantc: that line 23 in init-bottom/ltsp works for you?!
| |
| 20:55 | [ -d "${rootmnt}/proc" ] || panic "rootmnt not mounted in init-bottom"
| |
| 20:55 | ...that script mounts rootmnt, why does it check if it's already mounted?!
| |
| 20:56 | And of course that chroot will fail too
| |
| 20:56 | <vagrantc> alkisg: haven't tested recent trunk, but trunk from friday worked
| |
| 20:56 | alkisg: about to test trunk now...
| |
| 20:56 | <alkisg> That file hasn't changed
| |
| 20:56 | Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Quit: Leaving) | |
| 20:56 | <alkisg> vagrantc: wait, I think it's clearly broken, so let's just talk about how to fix it, i.e. tell me how to detect in nfs if you want to do the cow stuff
| |
| 20:57 | if ! chroot ${rootmnt} /usr/bin/test -w "/" ; then
| |
| 20:57 | <vagrantc> alkisg: it's not clearly broken for me...
| |
| 20:57 | <alkisg> Hm. Ah, ok, got it
| |
| 20:57 | It only works for NFS
| |
| 20:58 | As the nfs scripts mount the root to /root
| |
| 20:58 | While nbd has only prepared /dev/nbd0
| |
| 20:58 | <vagrantc> as should the NBD scripts?
| |
| 20:58 | <alkisg> OK give me a minute to check...
| |
| 20:59 | Aaaah ok it's because I'm using btrfs and nbd doesn't know how to mount it
| |
| 20:59 | (while ltsp_nbd does)
| |
| 20:59 | <vagrantc> the NBD script just sets up a device that is then used for root
| |
| 21:00 | trunk still works for me
| |
| 21:01 | it technically mounts it in ${rootmnt} :)
| |
| 21:01 | <alkisg> Eh, no, the nbd script doesn't loop mount /dev/nbd0 anywhere
| |
| 21:01 | root=/dev/nbd0 is a valid block device
| |
| 21:01 | So it doesn't bother with that
| |
| 21:01 | <vagrantc> yes, you'd have to pass root=/dev/nbd0
| |
| 21:01 | and it defaults to using scripts/local
| |
| 21:01 | <alkisg> But still in init-bottom that wouldn't be mounted
| |
| 21:02 | Ah
| |
| 21:02 | Let me check..
| |
| 21:03 | hughessd has joined IRC (hughessd!~hughessd@173-164-117-109-Oregon.hfc.comcastbusiness.net) | |
| 21:04 | <alkisg> Yup, that looks much better! So new command line: append ro initrd=initrd.img root=/dev/nbd0 nbdroot=:ltsp_i386 init=/sbin/init-ltsp
| |
| 21:04 | ...and I got to a tty (not ldm?!) without ltsp_nbd :)
| |
| 21:05 | <vagrantc> this setup should make it really easy to switch between NFS and NBD
| |
| 21:05 | <alkisg> Indeed, just a command line option
| |
| 21:05 | <vagrantc> anyone use the kiosk screen script?
| |
| 21:05 | <alkisg> Some environment variables must be missing. I'll commit the changes so far.
| |
| 21:05 | <vagrantc> i need to get it to work with debian...
| |
| 21:06 | <alkisg> It's a bit broken in ubuntu too, afaik
| |
| 21:06 | It'd be nice to merge those two
| |
| 21:06 | <vagrantc> also wanted to change a few things ... i.e. create the user if it didn't exist annd such
| |
| 21:06 | might also want to look into using nodm
| |
| 21:07 | alkisg: oh, i don't think i tried init=/sbin/init-ltsp
| |
| 21:07 | * vagrantc keeps wanting to try /sbin/ltsp-init | |
| 21:08 | <vagrantc> upstream LXDE fixed shutdown/reboot from LTSP :)
| |
| 21:09 | i think that happened a while ago, but happy to see it now.
| |
| 21:10 | meh. a chain of recommends are pulling in exim4 ... not sure how to conflict with that.
| |
| 21:10 | all i can think of, and it would be wrong, is to conflict/breaks with mail-transport-agent ... but i just want that to be the default, not impossible to do.
| |
| 21:11 | <alkisg> A manual apt-get install --no-install-recommends "all the recommented packages except exim4". Lame :(
| |
| 21:12 | <vagrantc> or i could ship a dummy "ltsp-nomail" package that breaks mail-transport-agent and is recommended by ltsp-client
| |
| 21:12 | although that might get unpredictible.
| |
| 21:13 | unless i could convince maintainers of all the affected packages that recommemnds is too strong, but it seems like an appropriate case of recommends...
| |
| 21:14 | <alkisg> Blacklist it temporarily :P
| |
| 21:14 | <||cw> or, add a depends to nullmailer
| |
| 21:14 | <alkisg> With some apt configuration file
| |
| 21:14 | <||cw> or just install nullmailer
| |
| 21:14 | <alkisg> Then remove that file
| |
| 21:14 | <vagrantc> eeyk.
| |
| 21:14 | mounting with a writeable root ended up re-creating some files in /var/cache/apt/ that eat up 33MB
| |
| 21:15 | <||cw> mount /var/cache/apt/ on a tempfs :)
| |
| 21:15 | I do this on SSD's too
| |
| 21:16 | <vagrantc> it is mounted on a tmpfs, which means the thin-client is wasting ram.
| |
| 21:16 | <||cw> reboot, done?
| |
| 21:16 | <vagrantc> this happened immediately on boot.
| |
| 21:16 | <||cw> does have a "nocache" option?
| |
| 21:16 | or just always run apt-get clean?
| |
| 21:17 | <vagrantc> nocache?
| |
| 21:17 | <||cw> does apt^^
| |
| 21:17 | <vagrantc> i'll figure this out before i upload to debian :P
| |
| 21:17 | * ||cw accidentally a whole word! | |
| 21:18 | <vagrantc> i could mount another aufs on top of it that's read-only... hah.
| |
| 21:18 | <||cw> esmtp-run seems nice, installed size 8KB
| |
| 21:19 | <vagrantc> this is probably some cron job that kicked off
| |
| 21:19 | <||cw> ah, depends on estp, so add another 150K, still a lot smaller than nullmailer
| |
| 21:19 | <vagrantc> now /var/lib/dpkg is also getting ugly.
| |
| 21:20 | although in actuality, i don't even want a mail-transport-agent
| |
| 21:20 | <||cw> I don't think i even allow cron to start on my thins
| |
| 21:20 | <vagrantc> though installing a minimalist one would be better than nothing.
| |
| 21:21 | <||cw> a relay one is good, if you happen to have something that expects it, like cron
| |
| 21:21 | <vagrantc> basically, it all went wrong went i added the dependency on rsyslog, which has a dependency/recommends chain on cron, logrotate, and eventually a mail-transport-agent
| |
| 21:21 | but we need both rsyslog and cron by default in order to support several optional features of LTSP
| |
| 21:22 | * alkisg thinks he's going to break some more stuff before it gets better... removing ltsp_nbd... :) | |
| 21:22 | <||cw> relay sounds great then, so you can know when cron is failing
| |
| 21:23 | <vagrantc> only if there's a sane place to relay to
| |
| 21:23 | and i don't want to have the LTSP server-side packages have a working mail setup.
| |
| 21:24 | or rather, i don't want to require that
| |
| 21:26 | <alkisg> vagrantc: I want to create an ltsp-nbd-client wrapper for those that want to use nbd-proxy. I'll copy_exec that to initramfs's /sbin. But where would I put that in the source tree?
| |
| 21:27 | <vagrantc> alkisg: could just dump it in client
| |
| 21:28 | or client/scripts/
| |
| 21:28 | <alkisg> Thanks :)
| |
| 21:28 | piet has joined IRC (piet!5b60633a@gateway/web/freenode/ip.91.96.99.58) | |
| 21:29 | <vagrantc> alkisg: so init-bottom should copy /tmp/net-*.conf into somewhere the root can get at?
| |
| 21:29 | <alkisg> vagrantc: right
| |
| 21:30 | <vagrantc> and the debate was /run/ltsp vs. /var/cache/ltsp ?
| |
| 21:30 | <alkisg> Yes
| |
| 21:30 | <vagrantc> (or alternately dumping it into /var/cache/ltsp/ltsp_config
| |
| 21:30 | <alkisg> Yup
| |
| 21:30 | <vagrantc> (or ltsp_config_env?)
| |
| 21:30 | <alkisg> No
| |
| 21:30 | That last one gets reset on each lts.conf parsing
| |
| 21:30 | <vagrantc> ah.
| |
| 21:31 | didn't get the difference between the two.
| |
| 21:31 | <alkisg> Yeah Gadi was a bit cryptic there :)
| |
| 21:32 | mgariepy: I'll leave the nbd-proxy-wrapper stuff for tomorrow. Expect something bootable by tomorrow evening, UTC
| |
| 21:32 | <mgariepy> alkisg, great thanks.
| |
| 21:32 | <vagrantc> so, when init=/sbin/init-ltsp is used, all init-bottom/ltsp does is essentially ensurre that ${rootmnt} is writeable?
| |
| 21:32 | <alkisg> stgraber: in a couple of hours, could you upload another ltsp package in your ppa? It'll be broken, but it'll be easier to check what needs fixing...
| |
| 21:32 | <mgariepy> i'll push a working version of nbd-proxy very soon :)
| |
| 21:33 | <alkisg> vagrantc: yes, the cow stuff and the net-eth0.conf stuff
| |
| 21:33 | We have a pretty minimal initramfs code now...
| |
| 21:33 | <vagrantc> and alternately, if init=/sbin/init-ltsp, run the init-ltsp hook.
| |
| 21:33 | er, if not using init= ...
| |
| 21:34 | <alkisg> Right
| |
| 21:34 | <piet> @alkisg: I remember U telling us that U use epoptes. My question: is it possible to run the epoptes-server out of the ltsp image on a fat-client and if so, is there anything special about it? (I got errors . . .)
| |
| 21:34 | <alkisg> piet: it's much easier to use ltsp-remoteapps for that
| |
| 21:35 | I.e. to still run epoptes on the server, and run the GUI with ltsp-remoteapps from the fat client
| |
| 21:35 | <piet> that is an alternate soft?
| |
| 21:35 | <alkisg> It's an lts.conf option
| |
| 21:35 | !ltsp-remoteapps
| |
| 21:35 | <ltsp`> alkisg: Error: "ltsp-remoteapps" is not a valid command.
| |
| 21:35 | <alkisg> Meh, dumb bot
| |
| 21:36 | klausade has left IRC (klausade!~klaus@84.215.157.180, Ping timeout: 248 seconds) | |
| 21:36 | <alkisg> You put REMOTE_APPS=True in lts.conf
| |
| 21:36 | And you launch epoptes with: ltsp-remoteapps epoptes
| |
| 21:37 | And you can also create a wrapper script in /usr/share/local/bin/epoptes that launches ltsp-remoteapps /usr/bin/epoptes. This way you'll be able to launch epoptes from the menu, instead of using the terminal.
| |
| 21:38 | * alkisg needs to write that in the wiki page... piet, file a question about that, in launchpad, use the epoptes help menu to find the link | |
| 21:38 | <piet> I didnt get that totally. What You just explained sound like the management for the client. Am I right?
| |
| 21:39 | Isnt there a difference between Clients and server?
| |
| 21:39 | <alkisg> piet: your objective is to be able to run the epoptes gui from a fat client, right?
| |
| 21:39 | <piet> yes
| |
| 21:39 | server
| |
| 21:40 | <alkisg> The server (daemon, no user interface) will still be in the ltsp server
| |
| 21:40 | <stgraber> alkisg: do you want me to wait for something or should I upload one now?
| |
| 21:40 | <alkisg> The GUI, the epoptes window, will be launched from the fat client
| |
| 21:40 | stgraber: please wait a couple of hours, I want to remove ltsp_nbd first
| |
| 21:40 | <stgraber> alkisg: ok, ping me on IRC when you want one, it takes me 30s to push one ;)
| |
| 21:40 | <alkisg> stgraber: thanks :)
| |
| 21:41 | piet: I'm in the middle of something and can't take the time to write the steps properly. Please file a question in launchpad about it, so that I can answer it with a wiki page.
| |
| 21:41 | https://answers.launchpad.net/epoptes
| |
| 21:41 | <piet> thanx
| |
| 21:42 | <alkisg> Or use the ltsp or edubuntu mailing lists
| |
| 21:43 | <piet> isnt there a wiki where its already documented, or is it that brand new?
| |
| 21:43 | <vagrantc> GAH!
| |
| 21:44 | the fact that simply running ltsp-update-image overwrites the configuration to use NBD drives me crazy ... i've got to fix that.
| |
| 21:45 | <stgraber> it should be changed to using the new .d directory anyway
| |
| 21:45 | overwriting an ltsp file in there instead of using my python script to override the main configuration file (which is really ugly but we didn't really have a choice back then)
| |
| 21:46 | <alkisg> piet: the wiki is in http://www.epoptes.org/documentation, I'll write another page there about running epoptes from a fat client in 2 days, if you remind me with that question or mail in the mailing list.
| |
| 21:46 | <vagrantc> it's that it tweaks the pxelinux configuration that ltsp-update-image breaks things ... the configuration changes are fine.
| |
| 21:47 | i.e. setting it up to offer nbd images isn't nearly as damaging as actually forcing it before it's ready.
| |
| 21:48 | <alkisg> vagrantc: so, how should I save the contents of net-*.conf?
| |
| 21:48 | Should I just put them to ltsp_config and be done with it?
| |
| 21:48 | <vagrantc> alkisg: just tested a patch to copy to /var/cache/ltsp/net-*.conf
| |
| 21:49 | <alkisg> Ah, cool, please commit it when you think it's ready
| |
| 21:49 | (remember to pull first, I just removed ltsp_nbd... :))
| |
| 21:50 | <piet> @alkisg: I got that - remind U by question - no hurry - thank U, pls dont feel disturbed, cause I realy appreciate Ur help.
| |
| 21:51 | <vagrantc> alkisg: bzr-rebase is really useful for that :)
| |
| 21:52 | <alkisg> Ehhh my capacity for new commands is limited to 10 per month, now I learned about madison, syncpackage and a lot of others, leaving that one for March... :D
| |
| 21:52 | *rmadison
| |
| 21:52 | brunolambert has left IRC (brunolambert!bruno@nat/revolutionlinux/x-ivvhwiwhynnvpxog, Quit: brunolambert) | |
| 21:58 | <vagrantc> alkisg: well, i've got to tackle some other things for a bit, but pushed the changes
| |
| 21:58 | <alkisg> vagrantc: I think we're missing the $SERVER in the init-ltsp.d dir
| |
| 21:58 | I'll have a look tomorrow too
| |
| 21:59 | <vagrantc> alkisg: it's because of 00-getltsconfig
| |
| 21:59 | <alkisg> I guess it's pretty broken now, but I hope tomorrow it'll be bootable
| |
| 21:59 | <vagrantc> alkisg: it hard-codes NBD_HOST_SERVER=server, which ends up setting SERVER=server
| |
| 21:59 | * vagrantc was hoping to upload today :( | |
| 21:59 | <alkisg> vagrantc: I don't see any SERVER= lines anywhere
| |
| 22:00 | <vagrantc> alkisg: 50-hostname sets server, and the hard-coding of NBD_ROOT_HOST breaks 50-hostname
| |
| 22:01 | <alkisg> vagrantc: 50-hostname writes $SERVER to /etc/hosts, but where does that environment variable get its value?
| |
| 22:02 | <vagrantc> oh, weird.
| |
| 22:02 | it was in there somewhere...
| |
| 22:02 | it took me a while to find it ...
| |
| 22:03 | client/ltsp_config.d/10ltsp-server
| |
| 22:04 | <alkisg> Ah, set_lts_var, that's why I couldn't find it with grep SERVER= :D
| |
| 22:04 | <vagrantc> yup
| |
| 22:04 | <alkisg> Hey! set_lts_var?!
| |
| 22:04 | That will be emptied on the next lts.conf evaluation...
| |
| 22:04 | hughessd has left IRC (hughessd!~hughessd@173-164-117-109-Oregon.hfc.comcastbusiness.net, Quit: hughessd) | |
| 22:04 | <vagrantc> it's a maze of stuff
| |
| 22:05 | <alkisg> ...we need to clean that up too some time :(
| |
| 22:06 | vagrantc: err that 10ltsp-server is awful
| |
| 22:06 | All that just to get the server where nbd-client or nfs client connected to?!
| |
| 22:09 | For nbd, we can save the parameters from the initramfs, along with net-*.conf
| |
| 22:09 | or parse the command line, but I don't see why we need to parse fstab..
| |
| 22:12 | Time to get some sleep. stgraber could you please upload a package to your ppa?
| |
| 22:18 | <stgraber> alkisg: ok
| |
| 22:22 | alkisg: uploading now
| |
| 22:23 | <alkisg> Ty, good night all
| |
| 22:25 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.) | |
| 22:32 | hughessd has joined IRC (hughessd!~hughessd@173-164-117-109-Oregon.hfc.comcastbusiness.net) | |
| 22:37 | adrianorg_ is now known as adrianorg | |
| 22:38 | piet has left IRC (piet!5b60633a@gateway/web/freenode/ip.91.96.99.58, Quit: Page closed) | |
| 22:39 | bobby_C has joined IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at) | |
| 22:43 | mistik1 has left IRC (mistik1!~mistik1@unaffiliated/mistik1, Ping timeout: 276 seconds) | |
| 23:50 | bobby_C has left IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at, Read error: Operation timed out) | |
| 23:54 | Trixboxer has left IRC (Trixboxer!~Trixboxer@115.124.115.71, Quit: "Achievement is not the end, its the beginning of new journey !!!") | |