IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 31 May 2007 (all times are UTC)
| 00:10 | yanu has quit IRC | |
| 00:23 | PMantis has left #LTSP | |
| 00:25 | Burgundavia has quit IRC | |
| 00:55 | lambda has joined #ltsp | |
| 01:01 | ctyler has quit IRC | |
| 01:32 | lambda has quit IRC | |
| 01:41 | wf_c0d3r has joined #ltsp | |
| 01:41 | <wf_c0d3r> hello guys...can you help me in setting up my ltsp server using ubuntu
| |
| 01:43 | i just followed the instruction in the wiki page...but when connecting to the client an error appear to it pxe-32: TFTP open timeout
| |
| 01:43 | can you help me on this?
| |
| 01:44 | <sep> wf_c0d3r, is your tftp server running and configured preoperly ?
| |
| 01:44 | dean_za_ has quit IRC | |
| 01:46 | <wf_c0d3r> yups but i just dont know if i have a correct setup
| |
| 01:52 | <sep> wf_c0d3r, well double check it ?
| |
| 02:08 | <wf_c0d3r> sep, i get the correct ip on my workstation as what i have setup to my dhcpd.conf
| |
| 02:09 | vagrantc has joined #ltsp | |
| 02:14 | <sep> i did mean to doublecheck you tftp related config
| |
| 02:19 | wf_c0d3r has left #ltsp | |
| 02:19 | MoL0ToV has joined #ltsp | |
| 02:23 | cyberorg has joined #ltsp | |
| 02:24 | bill_c has quit IRC | |
| 02:25 | sahil has quit IRC | |
| 02:47 | Paul_UK has joined #ltsp | |
| 02:48 | <Paul_UK> hey all, got ubuntu server, ltsp is running, i have a client that pxe boots from the server and now im looking at a logon screen. questions are really. whats next? how do i add users, create a "published" desktop, etc, etc. Thanks. Sorry, probably everyone asks this question and its day 1 with ltsp
| |
| 02:57 | <cyberorg> Paul_UK: create accounts on the server
| |
| 02:57 | all the apps and account available on the server are available on the clients
| |
| 02:58 | <Paul_UK> cyberorg: yeah i guessed both of those, but is there a how-to documentation, i'm checking the wiki now and struggling on a few concepts.
| |
| 02:58 | like where is ltspadmin and ltspcfg
| |
| 02:59 | <vagrantc> there is no ltspadmin or ltspcfg
| |
| 02:59 | <Paul_UK> lol :) ok, so im definately looking at the wrong place. if someone can point me to uptodate documentation, would be great
| |
| 03:00 | <vagrantc> !ubuntu
| |
| 03:00 | <ltspbot`> vagrantc: "ubuntu" is a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org. to install ltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
| |
| 03:02 | <vagrantc> Paul_UK: basically there's very little difference between an LTSP server and a desktop machine
| |
| 03:02 | Paul_UK: so you need to look at the ubuntu documentation for creating users, adding programs, etc.
| |
| 03:03 | <Paul_UK> well i looked here : https://help.ubuntu.com/community/UbuntuLTSP?highlight=%28%28UbuntuLTSP%7CLTSPQuickInstall%29%29
| |
| 03:03 | and it only talks about the server and getting it working with a client.
| |
| 03:04 | Administrative tasks, theres no mention. Like adding users and creating the client, as you would want.
| |
| 03:05 | <vagrantc> well, there's nothing LTSP specific about that, other than adding users to the fuse group if you want them to be able to use local devices
| |
| 03:05 | privet has quit IRC | |
| 03:07 | <vagrantc> ogra: should i merge the "controlgen" patches into mainline?
| |
| 03:07 | <Paul_UK> ok, so do i download thinstation and then configure that how i want?
| |
| 03:07 | <vagrantc> thinstation?
| |
| 03:10 | <Paul_UK> ok, i guess what im saying is. im new to all this. i've got a server a client boots via pxe and gets a logon screen. what next? im familiar with citrix, so i know the topology of thin clients, but with ltsp im kinda lost.
| |
| 03:10 | <vagrantc> Paul_UK: do you have a user on the server?
| |
| 03:11 | <Paul_UK> yeah the initial one that i setup the server with. but when i log on with the client, it seems to try to get a desktop, but then i get back to the logon screen.
| |
| 03:12 | <vagrantc> ogra: and thoughts on ltsp-clien-core patches ?
| |
| 03:12 | Paul_UK: any messages in /home/USER/.xsession-errors ?
| |
| 03:14 | <Paul_UK> yeah - warning : xrdb command not found; X resources not merged
| |
| 03:14 | unable to start X session --- no "/home/paul/.xsession" file, no session managers, no windows managers, no terminal emulators found
| |
| 03:15 | <vagrantc> that's more meaningful
| |
| 03:17 | so, you need to install a window manager
| |
| 03:17 | probably install the ubuntu-desktop metapackage
| |
| 03:18 | <Paul_UK> hmm, theres no mention of that in any of the "ubuntu" docs that i have seen
| |
| 03:19 | <vagrantc> did you install ubuntu server or ubuntu desktop ?
| |
| 03:19 | <Paul_UK> server
| |
| 03:19 | <vagrantc> an LTSP server is really more like a desktop machine
| |
| 03:19 | probably better to start off with ubuntu desktop
| |
| 03:19 | <Paul_UK> why would i want desktop?
| |
| 03:19 | <vagrantc> will work on improving all that for the next release
| |
| 03:20 | because an ubuntu server is essentially a "desktop" server.
| |
| 03:20 | it provides desktops to your think clients
| |
| 03:20 | thin
| |
| 03:20 | so the application sets it needs are more similar to a desktop install than a server install
| |
| 03:20 | <Paul_UK> hmm, so shall we take from this that either a) docuementation needs to be completed so that people can get this to work i.e ubuntu server/ltsp + clients or b) ubuntu server + ltsp is broken lol
| |
| 03:21 | <vagrantc> feel free to update the documentation
| |
| 03:21 | apparently, for you, the documentation is broken.
| |
| 03:21 | <Paul_UK> well there is no documentation that i can find
| |
| 03:21 | the quickstart guide is just for the server.
| |
| 03:23 | <vagrantc> the quickstart guide doesn't make it clear that it assumes you have a desktop environment already installed on the server
| |
| 03:24 | privet has joined #ltsp | |
| 03:25 | <Paul_UK> to be honest, frankly doesnt make much clear at all. it was quite a job actually getting the server up and running, but that doesnt have anything with ltsp. anyway im gonna try over at #edubuntu. thanks for your help, but hope you can see its frustrating when ubuntu is saying this is supported and out of the box working and it isnt.
| |
| 03:25 | <vagrantc> Paul_UK: where does it say it's supported out of the box?
| |
| 03:25 | <Paul_UK> on their site
| |
| 03:26 | <vagrantc> where?
| |
| 03:26 | <Paul_UK> im sure you asked me this before and i told you lol
| |
| 03:26 | <vagrantc> in order for it to be fixed, it would be useful to know exactly what you're talking about
| |
| 03:26 | <Paul_UK> well i dont think its a ltsp problem
| |
| 03:27 | http://www.ubuntu.com/products/WhatIsUbuntu/serveredition
| |
| 03:27 | Ubuntu Server edition includes thin client support using LTSP (Linux Terminal Server Project).
| |
| 03:27 | that to me means, if its included. Then it should be working if its installed.
| |
| 03:27 | <vagrantc> ah, yes. ok.
| |
| 03:28 | it appears a little misleading to me, yeah.
| |
| 03:28 | if you want an out-of-the-box LTSP environment, at the moment, edubuntu is the way to go.
| |
| 03:28 | <Paul_UK> from what I recall as well. Edubuntu edition has LTSP actually installed as part of the server install, which Ubuntu Server doesnt. To be honest, they should have had LTSP as an option, like DNS or LAMP.
| |
| 03:28 | <vagrantc> it's planned for the next release
| |
| 03:29 | <Paul_UK> cool. well i'll download and install edubuntu and then see how this fares
| |
| 03:29 | <vagrantc> basically, grabbing the work in edubuntu and generalizing it is all that's needed.
| |
| 03:31 | Thumann has joined #ltsp | |
| 03:37 | <Thumann> Hi, i have two ltsp servers running, i'd like to add "load balancing" and a form of failover, so that if one server fails, they'll be able to use the other. (the dhcp server used is located on a third server)
| |
| 03:37 | i looked into OpenSSI but it's terribly outdated
| |
| 03:38 | <vagrantc> Thumann: we await something better.
| |
| 03:40 | <Thumann> vagrantc: i'm reading up on X Load Balancing on the wiki right now, is that any good?
| |
| 03:41 | http://wiki.ltsp.org/twiki/bin/view/Ltsp/WorkInProgress#X_Load_Balancing
| |
| 03:44 | <vagrantc> honestly, no idea.
| |
| 03:45 | <Thumann> :) it's worth a shit
| |
| 03:45 | .. shot
| |
| 03:45 | :x
| |
| 03:46 | <vagrantc> basically, i'd recommend starting simple with round-robin login servers ... and then proceed from there
| |
| 03:46 | round-robin tftp/NFS servers ... etc
| |
| 03:47 | <Thumann> first come first serve?
| |
| 03:57 | <vagrantc> Thumann: basically, yeah.
| |
| 03:57 | <Thumann> which means i could/would end up with +50 users one machine and 2 users on the other
| |
| 03:57 | <vagrantc> if you've got 52 users, i really doubt round-robin would distribute them that badly.
| |
| 03:58 | more likely the problem of 3 users on one machine and none on the other
| |
| 03:59 | <Thumann> well if it's first come, first serve then aforementioned scenario could happen, no?
| |
| 03:59 | <vagrantc> ok, so not first come first serve, then :P
| |
| 03:59 | round-robin
| |
| 03:59 | <Thumann> :D describe that for me please
| |
| 04:00 | i'm not a native english speaker, so i'm thinking of stealing from the rich and giving to the poor
| |
| 04:00 | <vagrantc> there's many ways to implement it... but the basic idea is when a client makes a request for a server, it gets a random list of servers to choose from.
| |
| 04:00 | <Paul_UK> hehe thats robin hood
| |
| 04:00 | <Thumann> (which kinda sounds like load balancing come to think of it)
| |
| 04:00 | <vagrantc> right ...
| |
| 04:00 | true enough!
| |
| 04:00 | <Thumann> Robin Hood, the first implementation of load balancing
| |
| 04:01 | <vagrantc> Thumann: round-robin is essentially randomly selecting a server
| |
| 04:01 | <Thumann> vagrantc: like described in the link i .. linked before with the x load balancing
| |
| 04:01 | that seemed kinda random aswell
| |
| 04:01 | <vagrantc> maybe i should read the link :)
| |
| 04:01 | <Thumann> o_o
| |
| 04:02 | slacker
| |
| 04:02 | :)
| |
| 04:06 | <vagrantc> Thumann: ok ... so having read that ... the concepts are all there.
| |
| 04:06 | Thumann: if you're using LTSP there's more work to do to get it set up, but the idea's the same.
| |
| 04:07 | <Thumann> vagrantc: aren't "they"(the link) using ltsp?
| |
| 04:07 | <vagrantc> oops
| |
| 04:07 | LTSP 5
| |
| 04:08 | i'm pretty sure those patches are for ltsp 4.something
| |
| 04:08 | <Thumann> which is bad?
| |
| 04:08 | <vagrantc> it's different
| |
| 04:08 | <Thumann> as in bad?
| |
| 04:08 | <vagrantc> as in different
| |
| 04:08 | * Thumann smiles | |
| 04:08 | <vagrantc> you won't be able to follow the instructions exactly
| |
| 04:09 | you would have to adapt the idea
| |
| 04:09 | <Thumann> couldn't i just use v4 ?
| |
| 04:10 | <vagrantc> if you would rather
| |
| 04:10 | i can't stop you from doing anything :P
| |
| 04:11 | <Thumann> well since i dont know what the difference is
| |
| 04:11 | <vagrantc> it would be more help to the community to figure out how to get the current version of LTSP working than using older versions.
| |
| 04:11 | <Thumann> true
| |
| 04:11 | though this is due on tuesday
| |
| 04:11 | :o
| |
| 04:12 | <vagrantc> well, it's a little late to start anyways :P
| |
| 04:12 | good luck :P
| |
| 04:12 | <Thumann> >_<
| |
| 04:18 | vanya has joined #ltsp | |
| 04:48 | lambda has joined #ltsp | |
| 05:12 | Avatara has joined #ltsp | |
| 05:18 | Avatara has quit IRC | |
| 05:20 | Avatara has joined #ltsp | |
| 05:28 | lambda has quit IRC | |
| 05:35 | <ogra> jammcq, vagrantc, https://code.launchpad.net/~ltsp-drivers/ltsp/mainline :)
| |
| 05:36 | vagrantc, now all ltsp-drivers have mainline commit rights
| |
| 05:37 | i'll obsolete my old mainline branch and we'll move to that one
| |
| 05:37 | <vagrantc> people gotta be very careful with push/pull not to destroy history
| |
| 05:37 | <ogra> right
| |
| 05:37 | <vagrantc> that's my number 1 complaint about bzr right now
| |
| 05:37 | that and no gpg signatures
| |
| 05:38 | <ogra> well, currently ltsp-drivers are only us four
| |
| 05:38 | <vagrantc> well, no useful way to use gpg signatures
| |
| 05:38 | <ogra> you, me, scottie and jim
| |
| 05:40 | <vagrantc> when i reported the bug, i got the response "it looses history by design!"
| |
| 05:40 | some VCS
| |
| 05:40 | <ogra> heh
| |
| 05:41 | <vagrantc> don't know if any of the other distributed VCS's do any better with that. the nature of distributed vcs makes it hard to sanely preserv history
| |
| 05:45 | <ogra> vagrantc, if you go to the LP page of the branch do you get a note about the upload url ?
| |
| 05:46 | (while being logged in)
| |
| 05:49 | * vagrantc was on an olive break | |
| 05:50 | <vagrantc> gah.
| |
| 05:50 | i need to find a login page that iceweasel remembers.
| |
| 05:53 | ogra: yeah, i get an upload url
| |
| 05:53 | <ogra> yay, great
| |
| 05:53 | then i can obsolete my old mainline branch
| |
| 05:54 | and you can commit as you like ;)
| |
| 05:54 | <cliebow_> ogra:should i keep an eye out for the position you spoke of yesterday??
| |
| 05:54 | <ogra> (yes to both questions from the backlog btw)
| |
| 05:54 | cliebow_, i'll notify you
| |
| 05:54 | * vagrantc forgets which questions | |
| 05:54 | <ogra> <vagrantc> ogra: should i merge the "controlgen" patches into mainline?
| |
| 05:55 | <vagrantc> ogra: and thoughts on ltsp-clien-core patches ?
| |
| 05:55 | <cliebow_> cool!!
| |
| 05:55 | <ogra> vagrantc, and if you like the direct_x stuff
| |
| 05:58 | <vagrantc> i think the directx stuff definitely goes in
| |
| 06:00 | <ogra> yeah
| |
| 06:01 | <vagrantc> so i helped crash someone's vista laptop by letting them try ubuntu feisty today
| |
| 06:02 | ogra: autologin and LDM_SERVER too ?
| |
| 06:06 | <ogra> i dont feel like wanting autologin yet
| |
| 06:07 | also lets not add to much stuff to ldm now that scottie is rewriting the wrapper
| |
| 06:07 | <vagrantc> hrm,
| |
| 06:07 | <ogra> all that needs to be ported
| |
| 06:08 | <vagrantc> LDM_SERVER is a one-liner
| |
| 06:08 | <ogra> in C as well ?
| |
| 06:08 | <vagrantc> no idea
| |
| 06:08 | <ogra> :)
| |
| 06:09 | add it for now
| |
| 06:09 | <vagrantc> but it will be required to support an alternate server to implement the load balancing stuff
| |
| 06:09 | <ogra> reight
| |
| 06:09 | but it will need extra C code ...
| |
| 06:09 | we need to keep that in mind
| |
| 06:10 | <vagrantc> it can't be that bad.
| |
| 06:10 | <ogra> nah
| |
| 06:10 | <vagrantc> more complicated stuff, sure.
| |
| 06:10 | <ogra> but i want to see a working C ldm first ... before we clutter it with features
| |
| 06:11 | * ogra goes for coffee | |
| 06:11 | <vagrantc> while i know it was scottie's pet project, it seemed like the password expiry really was an extra feature ...
| |
| 06:12 | has he published code for the ldm c rewrite yet?
| |
| 06:14 | subir_ has joined #ltsp | |
| 06:15 | nf1 has quit IRC | |
| 06:15 | <vagrantc> ogra: well, i was seemingly able to commit
| |
| 06:23 | <highvoltage> vagrantc: here's an archive of the page I wanted to show you: http://web.archive.org/web/20060513022941/http://www.funroll-loops.org/
| |
| 06:28 | daya has joined #ltsp | |
| 06:28 | <daya> I am getting nfs over tcp failed, any idea,
| |
| 06:29 | what is that's error,
| |
| 06:31 | <vagrantc> what linux distro?
| |
| 06:31 | <daya> vagrantc, debian
| |
| 06:32 | <vagrantc> etch ?
| |
| 06:32 | <daya> vagrantc, yes
| |
| 06:32 | <vagrantc> installed with ltsp-server, or ltsp-utils ?
| |
| 06:32 | dpkg -l '*ltsp*'
| |
| 06:32 | <daya> ltsp-server, ltsp-build-client
| |
| 06:33 | root@basanta:~# dpkg -l | grep ltsp
| |
| 06:33 | ii ltsp-server 0.99debian11 Basic LTSP server environment
| |
| 06:33 | ii ltsp-server-standalone 0.99debian11 Complete LTSP server environment
| |
| 06:33 | <vagrantc> ok.
| |
| 06:34 | so, it boots, downloads the kernel and initramfs image ... and then your error?
| |
| 06:34 | <daya> yes,
| |
| 06:35 | <vagrantc> could you type the error exactly ?
| |
| 06:42 | <daya> vagrantc, NFS over TCP not available from <IP>
| |
| 06:43 | <vagrantc> ok...
| |
| 06:44 | daya: ok... as root, try: exportfs -va
| |
| 06:45 | well ...
| |
| 06:45 | grep ltsp /etc/exports
| |
| 06:47 | ogra has quit IRC | |
| 06:48 | <jammcq> sounds to me like he's got the wrong NFS installed
| |
| 06:48 | on the server
| |
| 06:48 | <vagrantc> like the user one?
| |
| 06:48 | <jammcq> could be
| |
| 06:48 | or NFS v2
| |
| 06:48 | if that's even a package
| |
| 06:48 | <vagrantc> if it's really etch
| |
| 06:49 | daya: dpkg -l nfs-kernel-server
| |
| 06:49 | or better yet, dpkg -l 'nfs-*-server'
| |
| 06:51 | tiagovaz has joined #ltsp | |
| 06:54 | tiagovaz has quit IRC | |
| 06:54 | ogra has joined #ltsp | |
| 06:55 | <daya> vagrantc, ii nfs-kernel-server 1.0.10-6 Kernel NFS server support
| |
| 06:55 | un nfs-user-server <none>
| |
| 06:56 | vagrantc, exportfs -va
| |
| 06:56 | exporting *:/opt/ltsp/i386
| |
| 07:00 | jammcq has quit IRC | |
| 07:00 | tiagovaz has joined #ltsp | |
| 07:02 | <vagrantc> grep ltsp /etc/exports
| |
| 07:06 | bronze has joined #ltsp | |
| 07:06 | <daya> root@basanta:~# cat /etc/exports | grep ltsp
| |
| 07:06 | /opt/ltsp/i386 *(ro,no_root_squash,async,subtree_check)
| |
| 07:06 | root@basanta:~#
| |
| 07:07 | <vagrantc> you could/should drop the i386 part, but that probably won't fix your problem ...
| |
| 07:07 | daya: so ...
| |
| 07:07 | daya: can you paste your /etc/dhcp3/dhcpd.conf to the pastebot ?
| |
| 07:07 | !pastebot
| |
| 07:07 | <ltspbot`> vagrantc: "pastebot" is The LTSP pastebot is at http://pastebot.ltsp.org. Please paste all text longer than a line or two to the pastebot, as it helps to reduce traffic in the channel. A link to the content will be pasted in the channel.
| |
| 07:08 | <daya> ok
| |
| 07:09 | <ltsppbot> "daya" pasted "authoritative; subnet 192.168." (21 lines) at http://pastebot.ltsp.org/151
| |
| 07:09 | <daya> vagrantc, did u get it
| |
| 07:10 | <vagrantc> yeah, let me look at it a bit
| |
| 07:10 | <daya> ok
| |
| 07:10 | <vagrantc> probably want to comment out the example.com line :P
| |
| 07:11 | so ... 192.168.2.3 is the IP address of your nfs server ?
| |
| 07:11 | <daya> yes
| |
| 07:12 | <vagrantc> ok ... mount -o ro,tcp,nolock 192.168.2.3:/opt/ltsp/i386 /mnt
| |
| 07:13 | <ogra> vagrantc, https://code.launchpad.net/~ubuntu-core-dev/ltsp/gutsy-ltsp
| |
| 07:13 | thats the gutsy branch, once you have commit rights to ubuntu-core-dev you can use that one as well
| |
| 07:14 | <daya> mount -o ro,tcp,nolock 192.168.2.3:/opt/ltsp/i386 /mnt
| |
| 07:14 | mount to NFS server '192.168.2.3' failed: possible invalid protocol.
| |
| 07:14 | mount 192.168.2.3:/opt/ltsp/i386 /mnt/
| |
| 07:14 | but 2nd succeed
| |
| 07:14 | Thumann has quit IRC | |
| 07:16 | <vagrantc> daya: you haven't somehow installed some other nfs server on your machine?
| |
| 07:16 | <daya> vagrantc, oh sorry,
| |
| 07:17 | vagrantc, no I think I have done all,
| |
| 07:17 | vagrantc, root@nepalinux:~# dpkg -l | grep nfs
| |
| 07:17 | rc libnfsidmap1 0.16-3 An nfs idmapping library
| |
| 07:17 | ii libnfsidmap2 0.18-0 An nfs idmapping library
| |
| 07:17 | ii nfs-common 1.0.10-6 NFS support files common to client and serve
| |
| 07:17 | ii nfs-kernel-server 1.0.10-6 Kernel NFS server support
| |
| 07:17 | root@nepalinux:~#
| |
| 07:18 | <vagrantc> looks fine to me
| |
| 07:18 | i'm pretty stumped.
| |
| 07:19 | <ogra> you donmt need to give tcp as option
| |
| 07:19 | <daya> vagrantc, I finally dorp to (initrmfs) in client
| |
| 07:19 | <vagrantc> ogra: klibc doesn't support udp nfs mounts
| |
| 07:19 | <ogra> recent nfs-kernel-server will default to it anyway and initramfs nfsmount doesnt know better than iusing tcp
| |
| 07:20 | vagrantc, right, but everything defaults to tcp anway
| |
| 07:20 | and the right option is proto=tcp iirc
| |
| 07:20 | so that might produce the error above
| |
| 07:21 | <vagrantc> but the error is: NFS over TCP not available from <IP>
| |
| 07:21 | <daya> ogra, How to slove it then,
| |
| 07:21 | <ogra> weird
| |
| 07:21 | anything weird in /etc/defaults ? did you tweak any values in there ?
| |
| 07:22 | (like trying to switch to nfs v4 or something ?)
| |
| 07:22 | <daya> let me check
| |
| 07:24 | I didn't get the as such the version, option ,
| |
| 07:24 | /etc/defaults/nfs-kernel-server, and nfs-common
| |
| 07:27 | yopla has joined #ltsp | |
| 07:28 | <yopla> Hello channel!
| |
| 07:29 | subir_ has quit IRC | |
| 07:29 | <yopla> trying to ltsp-build-client --dist sid --arch i386
| |
| 07:30 | <vagrantc> yopla: on a sid system ?
| |
| 07:31 | <yopla> yes
| |
| 07:31 | <vagrantc> how's it going?
| |
| 07:31 | <yopla> but I have an error on mount /proc
| |
| 07:32 | <vagrantc> could you paste the error exactly ?
| |
| 07:32 | <yopla> Ok I'll give a try
| |
| 07:32 | <vagrantc> cut-and-paste, if at all possible
| |
| 07:32 | <yopla> my console seems stuck for now
| |
| 07:34 | a5benwillis has joined #ltsp | |
| 07:34 | Guaraldo has joined #ltsp | |
| 07:37 | <ogra> dont install to an unclean target
| |
| 07:37 | <yopla> Ho
| |
| 07:37 | it doesn't delete files?
| |
| 07:38 | <ogra> rm -rf /opt/ltsp/i386 && ltsp-build-client
| |
| 07:38 | jammcq has joined #ltsp | |
| 07:38 | * vagrantc prefers mv /opt/ltsp/i386 /opt/ltsp/i386.borked | |
| 07:38 | <ogra> in ubuntu it stops and tells you to clean up first, that should also be in the next sid package
| |
| 07:38 | (unless vagrantc dropped it during merge :) )
| |
| 07:38 | <vagrantc> ogra: yeah, i had to fight with that while testing so it didn't re-download packages every time :)
| |
| 07:39 | that's probably proper behavior.
| |
| 07:39 | <yopla> OK now i've got a ssh console :-)
| |
| 07:39 | <vagrantc> yopla: ?
| |
| 07:39 | <ogra> ?
| |
| 07:39 | <yopla> my console was stucked
| |
| 07:41 | Hummm, strange operation seems working now...
| |
| 07:42 | I suspect my local sid mirror wasn't really synced
| |
| 07:42 | tiagovaz has quit IRC | |
| 07:42 | <vagrantc> sid could, in fact, be broken at the moment
| |
| 07:43 | i know debian-live hasn't been able to use sid for weeks now
| |
| 07:43 | <yopla> ouch
| |
| 07:43 | <ogra> https://lists.ubuntu.com/archives/gutsy-changes/2007-May/002482.html there we go
| |
| 07:43 | gutsy is open :)
| |
| 07:44 | <yopla> does ltsp works in feisty?
| |
| 07:45 | <ogra> sure
| |
| 07:45 | Blinny has joined #ltsp | |
| 07:45 | <ogra> its the best ltsp we have :)
| |
| 07:45 | <yopla> lol
| |
| 07:45 | <ogra> but the next upload to sid should have all features as well
| |
| 07:45 | <vagrantc> ogra: you uploaded a package simply to change the distro ? :P
| |
| 07:46 | <yopla> do you konw a way to install it quickly as a xen domU host?
| |
| 07:46 | <vagrantc> ogra: you couldn't have included *some* bugfix or something ?
| |
| 07:46 | <ogra> vagrantc, yes, else my CD builds will fail
| |
| 07:46 | <vagrantc> ogra: why not use lsb_release ?
| |
| 07:46 | ogra: like we do in debian
| |
| 07:46 | :)
| |
| 07:46 | <ogra> vagrantc, thats a traditional thing that my first upload has only this change :)
| |
| 07:46 | vagrantc, for 000-baseconfig ?
| |
| 07:46 | <vagrantc> ogra: yes
| |
| 07:47 | ogra: take a look
| |
| 07:47 | <ogra> since when ?
| |
| 07:47 | <vagrantc> ogra: over a year now
| |
| 07:47 | <ogra> why didnt tell me anyone ?
| |
| 07:47 | <vagrantc> it was going to be one of the first things i changed :P
| |
| 07:47 | i just reviewed the Ubuntu plugins to see what we should grab for debian today.
| |
| 07:48 | i keep forgetting to mention using lsb_release to set DIST
| |
| 07:48 | <yopla> yeah! lts client correctly installed :-)
| |
| 07:48 | <ogra> vagrantc, i'll change that immediately
| |
| 07:48 | vagrantc, that could become common stuff
| |
| 07:49 | <vagrantc> yeah, probably
| |
| 07:49 | <ogra> even though thats only one var
| |
| 07:49 | the rest is dpkg/apt depending
| |
| 07:50 | <vagrantc> yeah, with the controlgen stuff working, we can just have EARLY_PACKAGES=${EARLY_PACKAGES:-"ltsp-client"}
| |
| 07:50 | MIRROR still differs ...
| |
| 07:51 | tiagovaz has joined #ltsp | |
| 07:51 | <vagrantc> i'll think a lot about how we can merge more of the plugins
| |
| 07:52 | <Gadi> morning, guys. does anyone know if there is a way I can tell mkinitramfs to pull in other binaries/dpkgs when it generates the image?
| |
| 07:52 | <vagrantc> Gadi: yes :)
| |
| 07:53 | <yopla> what script does generate tftp ltsp tree?
| |
| 07:53 | <Gadi> care to share with the class, vagrantc?
| |
| 07:53 | <vagrantc> Gadi: look in /usr/share/initramfs-tools/hooks/*
| |
| 07:53 | Gadi: you probably want to make use of the copy_exec function
| |
| 07:54 | <Gadi> ah, so thats the diff between hooks and scripts
| |
| 07:54 | got it
| |
| 07:54 | * Gadi hands vagrantc an olive | |
| 07:54 | <ogra> Gadi, in ubuntu a very good example is to look at all udev scripts in /usr/share/initramfs-tools to get a good overview
| |
| 07:54 | <vagrantc> Gadi: if you make it a package, put it there. otherwise, stick it in /etc/initramfs-tools/hooks
| |
| 07:55 | <Gadi> right
| |
| 07:55 | <vagrantc> Gadi: oh, i've had plenty of olives today
| |
| 07:55 | <daya> vagrantc, in my another testing syste, xdmcp is not running, what should I do,
| |
| 07:56 | <vagrantc> daya: LTSP 5 doesn't use xdmcp by default
| |
| 07:56 | <daya> vagrantc, oh,
| |
| 07:56 | <vagrantc> you do need sshd running, though
| |
| 07:57 | <ogra> well, soon we'll likely get rid of XDMCP completely
| |
| 07:57 | <daya> ogra, :)
| |
| 07:57 | ogra, nice
| |
| 07:57 | <ogra> or do we have any reason to support it
| |
| 07:57 | ?
| |
| 07:57 | now with LDM_DIRECTX :)
| |
| 07:58 | <daya> no, but I am testing ltsp 5. on my system
| |
| 07:58 | now this time I am trying through the tar ball
| |
| 07:58 | image
| |
| 08:04 | <vagrantc> ogra: i really doubt if you'll see xdmcp die completely.
| |
| 08:05 | <ogra> but support from our side for it will
| |
| 08:05 | or do you plan to add features for XDMCP ?
| |
| 08:06 | <vagrantc> no plans on my part.
| |
| 08:06 | gladly leave someone else to maintain the "startx" script.
| |
| 08:06 | <ogra> se should probably switch the startx script to call ldm with LDM_DIRECTX
| |
| 08:06 | <vagrantc> no.
| |
| 08:06 | just no.
| |
| 08:07 | leave it alone, and recommend the alternative. let people decide for themselves.
| |
| 08:07 | <ogra> well
| |
| 08:07 | <daya> vagrantc, any idea root@nepalinux:~# ltsp-update-sshkeys
| |
| 08:07 | /usr/local/sbin/ltsp-update-sshkeys: line 5: ip: command not found
| |
| 08:07 | about it,
| |
| 08:07 | <ogra> its way more insecure than direct_x
| |
| 08:07 | Skarmeth has joined #ltsp | |
| 08:08 | <daya> vagrantc, here i am using tar ball, for installing and mainiting ltsp
| |
| 08:08 | <vagrantc> ogra: i can't think of reasons at the moment, but i am sure there are advantages to xdmcp, and i don't presume to claim we've entirely obsoleted it.
| |
| 08:08 | daya: do you have the ltsp-server package installed?
| |
| 08:08 | nobios has joined #ltsp | |
| 08:09 | <daya> I am using tar ball , that I have extracted in /opt directory
| |
| 08:09 | <yopla> arg! my first ltsp client kernel panics
| |
| 08:09 | <ogra> well, apparently the iproute package is missing in your install (i thought that was essential even in debian)
| |
| 08:10 | <vagrantc> daya: i know you're using the tarball, but do you have the ltsp-server package installed ? you'll want to have all of it's dependencies met.
| |
| 08:12 | <daya> vagrantc, sorry I don't get u for the first time, I have already installed tftpd-hpa, dhcp3, and nfs
| |
| 08:13 | <vagrantc> daya: you may as well install ltsp-server, even if using the tarball.
| |
| 08:13 | <ogra> yeah, that makes sure the deps for the scripts are there
| |
| 08:13 | <vagrantc> daya: that will pull in all the needed dependencies
| |
| 08:14 | <daya> vagrantc, ok
| |
| 08:14 | <yopla> ipconfig: eth0: SIOCGIFISINDEX: No such device
| |
| 08:14 | then kernel panic
| |
| 08:15 | * vagrantc wonders why everything is broken today | |
| 08:15 | <ogra> yopla, looks like you have an unsupported NIC in the client or force a driver on it thats not working for that card
| |
| 08:15 | <yopla> I think so :/
| |
| 08:16 | this is a cheap marvell yukon on a cheap laptop
| |
| 08:16 | <ogra> which distro ?
| |
| 08:16 | <vagrantc> yopla: is it a built-in network card ?
| |
| 08:16 | <yopla> orga: distro sid
| |
| 08:17 | <ogra> hmm, then vagrant is your man
| |
| 08:17 | * vagrantc hasn't actually done any sid installs since etch was released | |
| 08:17 | <yopla> vargantc: duno... I know that it's recognized in 2.6.20
| |
| 08:18 | <ogra> yopla, for a test you could try the feisty tarball
| |
| 08:18 | <vagrantc> yopla: yes, but is it a card you insert into the laptop, like a PCMCIA or USB card?
| |
| 08:18 | <ogra> it uses 2.6.20
| |
| 08:18 | <vagrantc> yopla: or is it built into the laptop case?
| |
| 08:18 | <yopla> orga: nope now it's 2.6.21
| |
| 08:19 | <vagrantc> yopla: and is it wireless?
| |
| 08:19 | <yopla> vargantc: it's onboard ethernet card
| |
| 08:19 | <ogra> yopla, feisty doesnt use 2.6.21
| |
| 08:19 | <yopla> orga: I'm on sid
| |
| 08:19 | Gadi has left #ltsp | |
| 08:19 | <ogra> <ogra> yopla, for a test you could try the feisty tarball
| |
| 08:19 | <ogra> it uses 2.6.20
| |
| 08:19 | <yopla> ok
| |
| 08:20 | is it possible to ltsp-client-build feisty with sid package?
| |
| 08:20 | <ogra> just move the sid ltsp chroot out of the way, and untar the feisty tarball in /opt/ltsp
| |
| 08:20 | <vagrantc> ogra: if your debootstrap supports feisty
| |
| 08:20 | <ogra> nobios, thats not possible
| |
| 08:20 | s/nobios/no/
| |
| 08:21 | <vagrantc> well, you'll probably run into other problems. just use the tarball
| |
| 08:21 | <yopla> Ok
| |
| 08:21 | where to get it?
| |
| 08:21 | <ogra> !download
| |
| 08:21 | <ltspbot`> ogra: "download" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/DownLoads
| |
| 08:21 | <vagrantc> !tarball
| |
| 08:21 | <ltspbot`> vagrantc: "tarball" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/Ltsp5TarballInstructions
| |
| 08:22 | * ogra hopes that thing is fianlly updated to the final release | |
| 08:22 | <yopla> orga ltsbot: tahnks a lot
| |
| 08:22 | <vagrantc> ogra: i'm guessing it hasn't been
| |
| 08:22 | <ogra> the pre release boots half as fast (about 20secs slower)...
| |
| 08:22 | vagrantc, well, i asked jammcq weekly to update it since the release
| |
| 08:23 | <vagrantc> ogra: me too. pesky upstream.
| |
| 08:23 | <ogra> slackers
| |
| 08:24 | <daya> vagrantc, one simple question
| |
| 08:24 | vagrantc, :D
| |
| 08:25 | <vagrantc> ogra: i even "backported" ltsp 0.99debian12 and ltsp for etch for jammcq to update the chroot.
| |
| 08:26 | <daya> vagrantc, which dhcp file is used by ltsp
| |
| 08:26 | <vagrantc> with dhcp3-server ... /etc/dhcp3/dhcpd.conf
| |
| 08:26 | you can add an "include" line to in to point to /etc/ltsp/dhcpd.conf if you want.
| |
| 08:28 | <daya> and what about that is included with ltsp
| |
| 08:30 | <ogra> /etc/ltsp/dhcpd.conf is included in ltsp ... but it doesnt matter which opne you use
| |
| 08:30 | there should be an example file in /usr/share/doc/ltsp-server/examples/
| |
| 08:31 | <vagrantc> ogra: ok ... so you said mdz is objecting to server-side code ... what about an ldm-server metapackage ?
| |
| 08:31 | ogra: because there are actually some things it would be nice to handle dependency wise.
| |
| 08:31 | <ogra> vagrantc, btw you will get massive problems with ltsp-manager and the dhcp.conf location, python-ltsp doesnt and will never touch /etc/dhcp3-server/
| |
| 08:32 | vagrantc, agreed, at least for ldminfod
| |
| 08:32 | but that should be the ldminfopd package rather :)
| |
| 08:32 | <vagrantc> ogra: yes, but also dependency on x11-common and other stuff
| |
| 08:32 | <ogra> what for a dep on x11-common ?
| |
| 08:32 | <vagrantc> /etc/X11/Xsession
| |
| 08:32 | <ogra> ah, right, i thought because of Xsession.d
| |
| 08:33 | (which i want us to not touch anymore)
| |
| 08:33 | <vagrantc> because that's what LDM uses to start the user's session
| |
| 08:33 | <ogra> i'd prefer to have that solved on the client side
| |
| 08:33 | <vagrantc> so we re-implement Xsession client-side ?
| |
| 08:33 | <ogra> like having a list of possible paths to iterate over
| |
| 08:34 | <vagrantc> that's not what i'm talking about
| |
| 08:34 | <ogra> for the known distros at least
| |
| 08:34 | and a fallback to xterm or so
| |
| 08:34 | <vagrantc> you need a dependency on whatever pulls in the Xsession stuff
| |
| 08:34 | <ogra> no
| |
| 08:34 | you need to be able to find something you can execute in X
| |
| 08:35 | <vagrantc> you'd rather use documentation over dependency resolution?
| |
| 08:35 | <ogra> nobios, i'd rather use a sane fallback
| |
| 08:35 | <vagrantc> what if there is not one single X program installed on the server?
| |
| 08:35 | <ogra> s/nobios/no/
| |
| 08:35 | <vagrantc> there is no sane fallback.
| |
| 08:35 | <ogra> we'll throw an error "please install a desktop session on your server"
| |
| 08:36 | <vagrantc> that's assinine.
| |
| 08:36 | <ogra> we have full communication now with scotties changes there is no need for server side stuff if we can solve it cleaner via a direct link
| |
| 08:36 | and better automated as well
| |
| 08:36 | and distro agnostic
| |
| 08:37 | <vagrantc> have you looked at what Xsession actually does?
| |
| 08:37 | <ogra> i dont care
| |
| 08:37 | <vagrantc> to re-implement that client-side would be insane.
| |
| 08:37 | <nobios> ogra: that why I changed , to . on auto completing nicknames :)
| |
| 08:37 | <ogra> i never said anything about re-implementing
| |
| 08:37 | <vagrantc> ogra: it handles a lot of things many programs expect to be working.
| |
| 08:37 | yes, let's just introduce hundreds of bugs.
| |
| 08:37 | <ogra> right thats why we should use the distros Xsession if its available
| |
| 08:38 | ??
| |
| 08:38 | <vagrantc> much saner than re-implementing everything.
| |
| 08:38 | <ogra> whats so problematic with adding one single case condition to the code ?
| |
| 08:38 | i really dont understand your problem here
| |
| 08:39 | <vagrantc> i think we're talking about different things.
| |
| 08:39 | <ogra> no
| |
| 08:40 | i dont think so
| |
| 08:40 | <vagrantc> well i do.
| |
| 08:40 | :P
| |
| 08:41 | <ogra> i want to have a runtime test that immediately tells the admin to install a desktop session independently of distro or any other stuff, you want to have that solved through deps by adding extra server sided packages instead
| |
| 08:41 | <vagrantc> when LDM logs a user in, it needs to run something...
| |
| 08:41 | <ogra> and if it cant is should advise the admin what to do
| |
| 08:42 | bronze has quit IRC | |
| 08:42 | <ogra> i'm fine with making X11-common an ltsp-server-standalone dep ... but not with adding extra packages just for that
| |
| 08:42 | <vagrantc> i'd rather move it out of ltsp-server-standalone
| |
| 08:42 | or ...
| |
| 08:42 | no
| |
| 08:42 | fine in standalone
| |
| 08:43 | <ogra> especially since that wont solve any of the probs we have with the tarballs
| |
| 08:43 | <yopla> ouch
| |
| 08:43 | same trouble
| |
| 08:43 | bronze has joined #ltsp | |
| 08:43 | <ogra> the tarball stuff can only be solved client side and scotties wrapper will give us all we need to do *everything* only via the client side if we want to
| |
| 08:44 | <vagrantc> yopla: your NFS server is messed up.
| |
| 08:44 | ogra: yeah, i don't object to that.
| |
| 08:44 | ogra: but why shift to documentation what could *easily* be handled by dependencies?
| |
| 08:44 | <yopla> orga: no, I think my pxe kernel is wrong
| |
| 08:44 | <ogra> it will only be handled for us
| |
| 08:45 | but not for speial setups and other distros
| |
| 08:45 | <vagrantc> ogra: that doesn't mean we can't implement the additional error checking you're suggesting
| |
| 08:45 | <ogra> i think both is tha way to go
| |
| 08:45 | <yopla> orga: is there a way to dmesg the client?
| |
| 08:45 | <vagrantc> ogra: both ... as in an ldm-server plus additional client-side error checking?
| |
| 08:45 | <ogra> yopla, once it is running it will write to the servers syslog
| |
| 08:46 | vagrantc, both as in ltsp-server-standalone and error checking ;)
| |
| 08:46 | <vagrantc> ogra: bah.
| |
| 08:46 | <ogra> i want to see more content first before we introduce an ldm-server package
| |
| 08:47 | just for a dependency its not worth to maintain an extra package
| |
| 08:47 | <vagrantc> but you refuse to accept any more content :P
| |
| 08:47 | <ogra> no i push for implementing it on the client side :)
| |
| 08:47 | <vagrantc> how do you implement ldm-server dependencies client-side ?
| |
| 08:47 | <ogra> (where easily< possible at least)
| |
| 08:47 | not at all
| |
| 08:48 | but for that one dep we can use -standalone
| |
| 08:48 | <vagrantc> so, how would you suggest installing additional ldm servers in a load-balanced environment?
| |
| 08:48 | install ltsp-server-standalone on all of them?
| |
| 08:48 | <ogra> not at all ?
| |
| 08:48 | apt-get install openssh-server ?
| |
| 08:49 | <vagrantc> apt-get install openssh-server x11-common *buntu-desktop ....
| |
| 08:49 | ?
| |
| 08:49 | <ogra> didnt our first implementation only talk about an admin maintained list ?
| |
| 08:49 | *buntu-desktop includes x11-common
| |
| 08:50 | if we need server sided code for loadbalancing at some point, i'm fine introducing ldm-server for that if that helps in any way
| |
| 08:50 | but the first implementation wont need that
| |
| 08:51 | <vagrantc> hrm. ubuntu-desktop depends on gdm ...
| |
| 08:51 | <ogra> right
| |
| 08:51 | which is fine
| |
| 08:52 | the desktop packages in debian dont depend on x11-common ?
| |
| 08:52 | not even indirectly ?
| |
| 08:52 | <vagrantc> no, but it pulls in a lot of stuff completely uneccesary to an ldm server
| |
| 08:53 | <ogra> ??
| |
| 08:53 | <vagrantc> for example, gdm
| |
| 08:53 | <ogra> it pulls in a desktop environment
| |
| 08:53 | yes, you need gdm if you want to log in graphically
| |
| 08:53 | you know ?
| |
| 08:55 | <vagrantc> you sure don't if you're logging in via LDM, do you?
| |
| 08:55 | <ogra> but if i log in on the server
| |
| 08:55 | <daya> I have these line in dhcpd.conf file ,tftfilename "/ltsp/i386/pxelinux.0 , option root-path "/opt/ltsp/i386,
| |
| 08:56 | <ogra> daya, looks good
| |
| 08:56 | <daya> I says tftp file not found in client , where should exactly I check
| |
| 08:56 | <ogra> did you run ltsp-update-kernels ?
| |
| 08:57 | <daya> ogra, yes,
| |
| 08:57 | <ogra> it copies the kernels to /var/lib/tftpboot
| |
| 08:57 | which is what the tftp client sees as /
| |
| 08:57 | <daya> root@nepalinux:boot# ltsp-update-kernels
| |
| 08:57 | Skipping invalid chroot: /opt/ltsp/pkg_cache
| |
| 08:57 | Skipping invalid chroot: /opt/ltsp/data
| |
| 08:57 | Updating /var/lib/tftpboot directories for chroot: /opt/ltsp/i386
| |
| 08:58 | <ogra> so do you have an ltsp dir in there ?
| |
| 08:58 | and an i386 dir below
| |
| 08:59 | <daya> yes,
| |
| 08:59 | root@nepalinux:i386# ls
| |
| 08:59 | System.map-2.6.18-4-486 initrd.img nbi.img pxelinux.0 vmlinuz
| |
| 08:59 | config-2.6.18-4-486 initrd.img-2.6.18-4-486 nbi.img-2.6.18-4-486 pxelinux.cfg vmlinuz-2.6.18-4-486
| |
| 09:00 | <ogra> looks ok
| |
| 09:00 | did you tweak your /etc/inetd.conf somehow ?
| |
| 09:00 | <vagrantc> ogra: i guess i'm thinking more like a server administrator with a rack-mount of ldm servers, where having GDM on the system would be a detriment rather than an asset.
| |
| 09:00 | <ogra> there should be a line like: tftp dgram udp wait root /usr/sbin/in.tftpd /usr/sbin/in.tftpd -s /var/lib/tftpboot
| |
| 09:00 | in 7etc/inetd.conf
| |
| 09:01 | <yopla> can I use a standard vmlinuz in tftp tree?
| |
| 09:01 | <daya> I have done nothing in inetd.conf file
| |
| 09:01 | <vagrantc> yopla: it is a standard vmlinuz
| |
| 09:02 | <yopla> Ok, i'll try another one then
| |
| 09:02 | <ogra> daya, check that the line matches
| |
| 09:02 | <daya> tftp dgram udp wait root /usr/sbin/in.tftpd /usr/sbin/in.tftpd -s /var/lib/tftpboot
| |
| 09:02 | <ogra> ok, looks fine
| |
| 09:03 | daya, are you sure there is no other dhcp server in your network ?
| |
| 09:03 | <daya> ogra, still getting same error
| |
| 09:04 | yes , no other dhcp server running
| |
| 09:06 | <vagrantc> yopla: or ... what do you mean by standard vmlinuz ... by default it uses the standard debian kernel and initramfs image
| |
| 09:07 | yopla: the initramfs is generated to do network boot
| |
| 09:07 | irule has joined #ltsp | |
| 09:07 | <daya> now my client system hangs in this step:
| |
| 09:08 | TFTP Prefix:/ltsp/i386
| |
| 09:08 | Trying to load:pxelinux.cfg/01-00....................
| |
| 09:08 | sbalneav has joined #ltsp | |
| 09:08 | <ogra> !s
| |
| 09:08 | <ltspbot`> ogra: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
| 09:08 | <sbalneav> Morning!!!
| |
| 09:08 | SBNet has joined #ltsp | |
| 09:09 | <daya> root@nepalinux:init.d# ps -aux | grep tftp
| |
| 09:09 | Warning: bad ps syntax, perhaps a bogus '-'? See http://procps.sf.net/faq.html
| |
| 09:09 | root 10254 0.0 0.1 2104 624 ? Ss 19:29 0:00 /usr/sbin/in.tftpd -s /var/lib/tftpboot
| |
| 09:10 | <ogra> thats bad
| |
| 09:10 | <vagrantc> heh
| |
| 09:10 | dpkg -l '*tftp*'
| |
| 09:10 | <ogra> it should be started by inetd (which is the default)
| |
| 09:10 | <vagrantc> daya: dpkg -l '*tftp*'
| |
| 09:11 | <ogra> (or are there tftpd packages not defaulting to inetd ?? i wouldnt know any)
| |
| 09:11 | <daya> vagrantc, root@nepalinux:init.d# dpkg -l | grep *tftp*
| |
| 09:11 | ii tftpd-hpa 0.43-1.1 HPA's tftp server
| |
| 09:11 | root@nepalinux:init.d#
| |
| 09:11 | <vagrantc> defaul for tftpd-hpa is to run from inetd ...
| |
| 09:12 | <daya> root@nepalinux:init.d# cat /etc/default/tftpd-hpa
| |
| 09:12 | #Defaults for tftpd-hpa
| |
| 09:12 | RUN_DAEMON="yes"
| |
| 09:12 | OPTIONS="-l -s /var/lib/tftpboot"
| |
| 09:12 | do i have done something wrong
| |
| 09:13 | <vagrantc> apt-get --purge remove tftpd-hpa ; ap-get install ltsp-server tftpd-hpa
| |
| 09:13 | <daya> vagrantc, I have already installed ltsp-server, when u told me earlier
| |
| 09:14 | ok I will reinstall tftpd-hpa,
| |
| 09:14 | <vagrantc> if you remove tftpd-hpa, it will remove ltsp-server
| |
| 09:14 | <daya> oh,
| |
| 09:18 | tftpd-hpa can be started by the inetd superserver or as a daemon and handle incoming connections by itself. The │
| |
| 09:18 | │ latter is only recommended for very high usage servers. │
| |
| 09:18 | │ │
| |
| 09:18 | │ Should the server be started by inetd?
| |
| 09:18 | sorry ,
| |
| 09:18 | <vagrantc> go with the default.
| |
| 09:18 | <daya> ok
| |
| 09:18 | root@nepalinux:init.d# ./tftpd-hpa restart
| |
| 09:18 | tftpd-hpa disabled in /etc/default/tftpd-hpa
| |
| 09:19 | what should I do
| |
| 09:19 | <ogra> nothing
| |
| 09:19 | be happy and boot a client
| |
| 09:19 | <daya> ogra, hi
| |
| 09:19 | <ogra> the above is what you wanted
| |
| 09:20 | <daya> ogra, but this time I got tftp time out, and client just hangs
| |
| 09:21 | <ogra> make sure your inetd is running
| |
| 09:21 | <vagrantc> bzr st
| |
| 09:21 | <ogra> (ltsp-server should have pulled in openbsd-inetd)
| |
| 09:21 | <daya> root@nepalinux:init.d# ps -aux | grep inetd
| |
| 09:21 | Warning: bad ps syntax, perhaps a bogus '-'? See http://procps.sf.net/faq.html
| |
| 09:21 | root 8915 0.0 0.1 1780 588 ? Ss 19:02 0:00 /usr/sbin/inetd
| |
| 09:22 | <vagrantc> daya: see that warning? that has nothing to do with LTSP. but you should listen and drop the - from your ps commands.
| |
| 09:23 | <ogra> yeah, saves a line in pastes :)
| |
| 09:24 | <daya> vagrantc, o,k :)
| |
| 09:24 | :D
| |
| 09:24 | what is the soultion then, I am feeling late to go to my home, from my off.
| |
| 09:25 | <ogra> bah, crap, gnome-power-manager dropped all hal code ... none of my ltsp patches apply :(
| |
| 09:25 | vagrantc, btw, its probably good to get at least some of them into debian
| |
| 09:26 | the nautilus and gnome-vfs patches as well as the powermanagement ones
| |
| 09:27 | <vagrantc> ogra: are there any bugs filed?
| |
| 09:27 | * vagrantc suspects it won't happen without bugs | |
| 09:27 | <ogra> i dont think so
| |
| 09:28 | <vagrantc> ogra: didn't some of those patches get included upstream?
| |
| 09:28 | <ogra> some went upstream with the latest gnome though
| |
| 09:28 | but even sid is far behind on gnome
| |
| 09:29 | <daya> I got this in my client system PXE-E32-TFTP open timeout , feeling too tired of the error
| |
| 09:31 | boxrock has joined #ltsp | |
| 09:32 | <daya> ogra, vagrantc any way thanks for the help,
| |
| 09:32 | lambda has joined #ltsp | |
| 09:32 | <daya> ogra, vagrantc bye
| |
| 09:33 | <boxrock> can anyone tell me why my thin-client x term hangs after a while esp. at higher resolutions?
| |
| 09:34 | <ogra> boxrock, ram usually
| |
| 09:34 | <boxrock> bad memory , on server?
| |
| 09:35 | tiagovaz has quit IRC | |
| 09:36 | daya has quit IRC | |
| 09:36 | <ogra> boxrock, not enough ram on the client ... do you have any kind of network swapping enabled ?
| |
| 09:39 | <boxrock> ogra, i used a stock 7.04 config
| |
| 09:39 | <ogra> that should have given you NBD_SWAP ...
| |
| 09:40 | so how much ram do your clients have ?
| |
| 09:40 | boxrock has quit IRC | |
| 09:42 | boxrock has joined #ltsp | |
| 09:42 | nobios has left #ltsp | |
| 09:44 | <boxrock> ogra: my client crashed, if you responded after my last msg i missed it. thx
| |
| 09:46 | <ogra> so how much ram do your clients have ?
| |
| 09:50 | <boxrock> it could be as low as 128m
| |
| 09:50 | <ogra> hmm, 128 should be fine
| |
| 09:50 | even without swap
| |
| 09:51 | * highvoltage has never seen a thin client even using half of that | |
| 09:53 | <ogra> well, with many local devices you would eat some ram
| |
| 09:53 | <vagrantc> *many*
| |
| 09:53 | <highvoltage> oh yes
| |
| 09:53 | tiagovaz has joined #ltsp | |
| 09:53 | <vagrantc> ogra: can't we start cdpinger from udev ?
| |
| 09:55 | boxrock has quit IRC | |
| 09:55 | <ogra> we could, does that make any difference ?
| |
| 09:56 | <vagrantc> well, if a cd gets inserted after the ltsp-client init scripts run ...
| |
| 09:56 | er, installed
| |
| 09:56 | <ogra> right
| |
| 09:56 | but you need to catch it on boot as well
| |
| 09:57 | boxrock has joined #ltsp | |
| 09:57 | <vagrantc> ah well. just a passing though.
| |
| 09:57 | t
| |
| 09:57 | <ogra> cdpinger is only to work around the fact that its not possible to monitor the CDrom
| |
| 09:57 | from the kernel
| |
| 09:58 | <vagrantc> our current scripts only run it if the devices are present during the init script run ...
| |
| 09:59 | <ogra> right
| |
| 09:59 | given that udev cant handle more than one /dev/cdrom thats fine for preinstalled CDroms
| |
| 10:02 | boxrock has quit IRC | |
| 10:03 | <ogra> vagrantc, feel free to change it if it doesnt stop working saving ram on non-cdrom clients is always helpful :)
| |
| 10:03 | vagrantc, there is a development team meeting in #ubuntu-meeting btw
| |
| 10:04 | (just starting)
| |
| 10:04 | highvoltage has quit IRC | |
| 10:07 | mhterres has joined #ltsp | |
| 10:07 | <mhterres> morning
| |
| 10:22 | boxrock has joined #ltsp | |
| 10:23 | tiagovaz has quit IRC | |
| 10:25 | boxrock has quit IRC | |
| 10:29 | rasmuson has joined #ltsp | |
| 10:33 | vanya has quit IRC | |
| 10:52 | cyberorg has left #ltsp | |
| 10:55 | meduxa has joined #ltsp | |
| 10:57 | efra has joined #ltsp | |
| 11:01 | Blinny has quit IRC | |
| 11:02 | <yopla> hello again. My nfs server is broken :-(
| |
| 11:03 | hello again. My nfs server is broken :-(
| |
| 11:04 | <sbalneav> Guess you need to fix it then.
| |
| 11:04 | <yopla> lol
| |
| 11:04 | yes bue I don't know where's the trouble
| |
| 11:05 | i've got RPC: failde to contact portmap
| |
| 11:05 | on server
| |
| 11:05 | lambda has quit IRC | |
| 11:05 | <sbalneav> So, looks like your portmap process isn't running.
| |
| 11:05 | ps -ef | grep portmap
| |
| 11:06 | What kind of server are you running? What distro?
| |
| 11:06 | <yopla> daemon 21413 1 0 16:02 ? 00:00:00 /sbin/portmap
| |
| 11:06 | <sbalneav> You got this on the server?
| |
| 11:06 | Or on the client?
| |
| 11:07 | <yopla> oooooops
| |
| 11:07 | it wasn't the correct IP on client
| |
| 11:07 | sorrrry
| |
| 11:07 | <sbalneav> So, your nfs server wasn't broken, your client was broken.
| |
| 11:08 | <yopla> lol you're right
| |
| 11:08 | <sbalneav> Here's a tip that will get you better help, faster
| |
| 11:08 | <yopla> my ltsp client is broken
| |
| 11:09 | <sbalneav> Instead of coming into a channel, and saying "something is broken :(", you might want to say something like this:
| |
| 11:09 | <yopla> i've got running /script/nfs-premount
| |
| 11:10 | then "nfs mount need a path"
| |
| 11:10 | <sbalneav> "Hi, I'm running LTSP version blahdeblah on distro FooFoo, on a server with 4 gigs of ram. When I turn on the client, I get "googoogoo" error. Could anyone tell me what's wrong?
| |
| 11:10 | OK, so check your dhcpd.conf
| |
| 11:11 | <yopla> sbalneav: you're right.. bit me :)
| |
| 11:11 | <sbalneav> Anywho, I have to go to a meeting.
| |
| 11:11 | I'll be back later.
| |
| 11:13 | <yopla> ok
| |
| 11:13 | I have to go home.
| |
| 11:13 | I'll test again later
| |
| 11:13 | bye channel
| |
| 11:13 | yopla has quit IRC | |
| 11:16 | vagrantc has quit IRC | |
| 11:17 | pablo_ has quit IRC | |
| 11:18 | pablo__ has joined #ltsp | |
| 11:20 | Gadi has joined #ltsp | |
| 11:28 | Paul_UK has quit IRC | |
| 11:33 | ohmbr has joined #ltsp | |
| 11:34 | <ohmbr> hi, how do i know with client ltsp am I?
| |
| 11:38 | I got it.. echo $DISPALY...
| |
| 11:40 | gu has quit IRC | |
| 11:40 | gu has joined #ltsp | |
| 11:41 | ohmbr has quit IRC | |
| 12:05 | <vmlintu> sbalneav: Hello! Got a second to check ltspfs behaviour? I've been trying to investigate why ltspfs mounts sometimes hang and for example 'df' -command blocks when trying to show ltspfs information.
| |
| 12:07 | What I found out was that in file ltspfs.c the ltspfs_read() method never calls pthread_mutex_unlock() if there was an error while reading and instead just returns.
| |
| 12:09 | I checked both the cvs version and the sources for the ubuntu ltspfs package version 0.4.3 and they both do the same.
| |
| 12:14 | What happens is that if the local device being used has e.g. corrupted filesystem, it returns an error to ltspfs when trying to read a file. Because ltspfs received an error, it returns from the ltspfs_read() method without freeing the lock and so nautilus/konqueror/what ever is hanging forever until the ltspfs process is killed.
| |
| 12:17 | jstraw is now known as Lumiere | |
| 12:18 | <vmlintu> Is there any reason why pthread_mutex_unlock() is not called in case of an error or is a bug? Changing it to free the lock in case of an error did fix the problem for me, but I'm wondering what problems this might bring..
| |
| 12:22 | nobios has joined #ltsp | |
| 12:30 | gu has quit IRC | |
| 12:35 | nf1 has joined #ltsp | |
| 12:49 | gu has joined #ltsp | |
| 12:49 | MoL0ToV has quit IRC | |
| 12:57 | ctyler has joined #ltsp | |
| 13:00 | ohmbr has joined #ltsp | |
| 13:02 | <ohmbr> hi, I'm trying to install a printer on ltsp client... I changed the lts.conf and ther client shows up on the network printer information...
| |
| 13:02 | but it isn printing...
| |
| 13:02 | what may I doing wrong?
| |
| 13:03 | ohmbr has quit IRC | |
| 13:06 | pablo__ has quit IRC | |
| 13:17 | rasmuson has quit IRC | |
| 13:19 | mhterres has quit IRC | |
| 13:23 | ogra has quit IRC | |
| 13:23 | ogra has joined #ltsp | |
| 13:26 | <ogra> grmbl, vagrant ....
| |
| 13:31 | vagrantc has joined #ltsp | |
| 13:34 | <ogra> vagrantc, grmbl
| |
| 13:34 | i didnt check the deps you added ...
| |
| 13:35 | now i just reintroduced esound in ubuntu :P
| |
| 13:35 | <vagrantc> ogra: yeah, well, you should've seen that last merge.
| |
| 13:36 | sepski has joined #ltsp | |
| 13:36 | <ogra> why do you depend on esound-utils ? i thought you merged all the pulse stuff
| |
| 13:36 | so drop te deps to nas and esound
| |
| 13:36 | *the
| |
| 13:37 | <vagrantc> sure
| |
| 13:38 | <ogra> libasound2-plugins, libgstreamer-plugins-pulse0.10-0 of -standalone should suffice
| |
| 13:38 | <vagrantc> ogra: the switch to controlgen will make that much easier to keep track of in the future. but the merging was challenging.
| |
| 13:38 | <ogra> indeed
| |
| 13:38 | actually only libasound2-plugins is really needed
| |
| 13:39 | libgstreamer-plugins-pulse0.10-0 is rather gnome specific, dunno if you want that in debian
| |
| 13:39 | <vagrantc> well, that's the beauty. we don't need to have the same deps in debian.
| |
| 13:40 | <ogra> well, we should probably have ltsp-server-sound depending on ltsp-sound-gnome | ltsp-sound-kde | ltsp-sound-native (whatever native might be)
| |
| 13:40 | and ltsp-server-sound bringing libasound2-plugins
| |
| 13:41 | |esound |nas for the dinosaurs
| |
| 13:41 | <vagrantc> does it include any code, or just a metapackage?
| |
| 13:41 | <ogra> hng ... this powermanager changes are evil
| |
| 13:41 | only a metapackage
| |
| 13:42 | * vagrantc stabs ogra with ldm-server | |
| 13:42 | * ogra curses while poking around in haö code | |
| 13:42 | <ogra> *hal
| |
| 13:42 | vagrantc, *sigh* ok, go for it ....
| |
| 13:43 | even though i want that case condition in the wrapper as well :P
| |
| 13:43 | <vagrantc> i just don't see what an ltsp-sound-server is ok where ldm-server was not :P
| |
| 13:43 | <ogra> yeah
| |
| 13:45 | * ogra tries sudo ltsp-build-client --chroot gutsy | |
| 13:45 | <vagrantc> ok, so dropping esound-clients | nas-bin from ltsp-server
| |
| 13:45 | <ogra> lets see what still works :)
| |
| 13:45 | yep
| |
| 13:46 | we have what we need in -standalone
| |
| 13:46 | its pretty cool that you dont need the daemon on the server with pulse
| |
| 13:46 | <vagrantc> leave libasound2-plugins in -standalone ... drop libgstreamer-plugins-pulse0.10-0
| |
| 13:47 | nice!
| |
| 13:47 | pulseaudio-esound-compat ?
| |
| 13:47 | <ogra> hmm, why does debootstrap hang ...
| |
| 13:47 | <vagrantc> currently a recommends for -standalone
| |
| 13:47 | <ogra> yep
| |
| 13:47 | you can use it if you want
| |
| 13:48 | <vagrantc> ok.
| |
| 13:48 | i only played with pulse briefly once
| |
| 13:48 | <ogra> we still set ESPEAKER so apps using esd can forward through it to pulse if you install pulseaudio-esound-compat
| |
| 13:48 | <vagrantc> right
| |
| 13:48 | nice bonus
| |
| 13:48 | <ogra> but usually all apps should use alsa, it works fine without the compat stuff
| |
| 13:49 | yep
| |
| 13:49 | <vagrantc> very cool
| |
| 13:49 | <ogra> the only drawback is flash
| |
| 13:49 | <vagrantc> how hard is it to set up the alsa stuff ? i mean, i know it's default in ubuntu ... just wondering for debian
| |
| 13:49 | we've go an open bug "document sound"
| |
| 13:49 | flash is always a drawback, yes.
| |
| 13:49 | <ogra> there is a bug in flash that adobe doesnt cloes ... it opens pulse connections over and over without closing them
| |
| 13:50 | * vagrantc can't wait for the next multimedia-takes-too-long-to-be-supported-in-free-software phenominon | |
| 13:50 | <ogra> so someone wrote a lib called libflsahsupport thats makes your firefox unstable and is strangely licensed but asdds a workaround for the flash bug
| |
| 13:50 | *libflashsupport
| |
| 13:51 | <vagrantc> ok, ltsp-client ... for ubuntu ... i have pulseaudio-esound-compat | esound | nas ... i suspect you want something different.
| |
| 13:51 | <ogra> we need to package that this release
| |
| 13:51 | nobios has quit IRC | |
| 13:52 | SBNet has quit IRC | |
| 13:53 | lambda_ has joined #ltsp | |
| 13:53 | <ogra> the compat code gives us the socket for ESPEAKER and additionally it depends on pulseaudio so we get that for free
| |
| 13:53 | s/socket/socket and port/
| |
| 13:55 | so the dep is correct
| |
| 13:58 | hm, i'm curious what will happen if i disable the X detection code completely once that chroot is built
| |
| 13:58 | <vagrantc> heh.
| |
| 13:58 | <ogra> xorg is supposed to just work ... but i havent see that yet
| |
| 13:59 | xrandr is pretty cool to adjust all the stuff we do in the initscript atm at runtime
| |
| 14:00 | so it would suffice if it comes up with a standard resolution and some default values an we can use xrandr to adjust it at runtime
| |
| 14:01 | hmm, we could even start X from rcS already i guess
| |
| 14:07 | Guaraldo has quit IRC | |
| 14:21 | Guaraldo has joined #ltsp | |
| 14:23 | Guaraldo has left #ltsp | |
| 14:25 | Guaraldo has joined #ltsp | |
| 14:45 | Avatara has quit IRC | |
| 14:47 | bobby_C has joined #ltsp | |
| 14:49 | nobios has joined #ltsp | |
| 14:49 | nobios has left #ltsp | |
| 14:52 | <vagrantc> ogra: or would you rather just depend on pulseaudio-esound-compat or ltsp-client ?
| |
| 14:52 | <ogra> pulseaudio-esound-compat should be a dep of ltsp-client
| |
| 14:53 | it puls in pulse and gives us the esd socket and port
| |
| 14:53 | meh
| |
| 14:53 | <vagrantc> pulseaudio-esound-compat | esound | nas
| |
| 14:53 | is what it is now
| |
| 14:53 | * ogra builds te third gutsy chroot now | |
| 14:53 | <ogra> yeah, thats fine
| |
| 14:53 | <vagrantc> i was wondering if i should change it to just pulseaudio-esound-compat
| |
| 14:53 | <ogra> ah
| |
| 14:54 | well, as a dep it should offer alternatives, if people really urgently want esd or nas they should be able to
| |
| 14:54 | <vagrantc> if they want the alternatives, they can install ltsp-client-core
| |
| 14:54 | <ogra> hmm, right
| |
| 14:54 | <vagrantc> :)
| |
| 14:55 | <ogra> yeah so drop it there
| |
| 14:55 | * ogra curses usplash | |
| 14:55 | <vagrantc> in debian i'd probably keep all the alternatives ... but in ubuntu it seems like the "one of anything" applies
| |
| 14:55 | sort of
| |
| 14:56 | <ogra> well the first one should be priorized and we pull it in from EARLY packages anyway
| |
| 14:56 | <vagrantc> well, i'd like to drop as much as possible from EARLY_PACKAGES
| |
| 14:56 | EARLY_PACKAGES="ltsp-client"
| |
| 14:56 | <ogra> hmm
| |
| 14:57 | hairy
| |
| 14:57 | <vagrantc> rather than having hidden RC bugs?
| |
| 14:57 | <ogra> well
| |
| 14:57 | rather than having a package with really logn crapy dep lines
| |
| 14:57 | *crappy
| |
| 14:57 | <vagrantc> having the dependencies all in a package means that we can more easily spot breakages on a distro-wide level
| |
| 14:58 | <ogra> right but still the dep list will be horribly ugly
| |
| 14:58 | anyway
| |
| 14:58 | seems sane
| |
| 14:58 | just doesnt look sane :)
| |
| 14:58 | <vagrantc> well... we'll sit on that for a while.
| |
| 14:59 | <ogra> nah, why ?
| |
| 14:59 | ohmbr has joined #ltsp | |
| 14:59 | <ogra> just moving them over from the variable shouldnt be to hard
| |
| 15:00 | <vagrantc> ok.
| |
| 15:00 | <ogra> for a start 1:1 should be fine
| |
| 15:00 | <vagrantc> i'll finish up the sound cleanup and then add more of that in.
| |
| 15:04 | a5benwillis has quit IRC | |
| 15:05 | <vagrantc> ogra: ok .. .so move all EARLY_PACKAGES as hard dependencies on ltsp-client?
| |
| 15:06 | <ogra> yes
| |
| 15:06 | <ohmbr> I have ltsp 4.2 on kubuntu and I want to upgrade to 5... i'll find a problem to do that?
| |
| 15:06 | <ogra> even though what you are doing there should have been called ubuntu-thin-client :)
| |
| 15:06 | <vagrantc> ogra: would you rather call it that?
| |
| 15:06 | <ogra> thats why we had added it back then
| |
| 15:06 | i just wonder
| |
| 15:07 | thats the historical mane of it ... a tribute to mdz .... not necessarily needed to keep though
| |
| 15:07 | <vagrantc> do you still want ldm in EARLY_PACKAGES ... i remember you needed it for a specific reason... CDs maybe ?
| |
| 15:07 | <ogra> no i need the dep for Cds
| |
| 15:08 | <vagrantc> right. i'll leave it.
| |
| 15:08 | any others shold be left in?
| |
| 15:08 | <ogra> yep
| |
| 15:08 | i'll sort that for ubuntu later
| |
| 15:08 | * ogra -> dinner | |
| 15:08 | <vagrantc> so, i shouldn't commit that stuff then?
| |
| 15:09 | <ogra> sure, do it ... i'll sort whats broken i meant :)
| |
| 15:09 | <vagrantc> ok.
| |
| 15:13 | lambda_ has quit IRC | |
| 15:16 | lambda has joined #ltsp | |
| 15:16 | sep__ has joined #ltsp | |
| 15:19 | lambda has quit IRC | |
| 15:20 | lambda has joined #ltsp | |
| 15:21 | meduxa has quit IRC | |
| 15:30 | Guaraldo has quit IRC | |
| 15:31 | <ace_suares> ogra!
| |
| 15:31 | hi vagrantc!
| |
| 15:31 | <vagrantc> greetings
| |
| 15:31 | <ace_suares> Can anyone write a spec on LP ?
| |
| 15:31 | * ace_suares bows to vagrantc | |
| 15:33 | <ogra> woah
| |
| 15:33 | on the t-1220 X works without config file
| |
| 15:33 | <vagrantc> nice :)
| |
| 15:33 | <ogra> but then thats using vesa anyway
| |
| 15:34 | * ogra gets the stopwatch | |
| 15:34 | Guaraldo has joined #ltsp | |
| 15:35 | bronze has quit IRC | |
| 15:35 | <ogra> 45 sec to X
| |
| 15:35 | <vagrantc> compared to ... ?
| |
| 15:36 | <ogra> 77 sec to ldm
| |
| 15:36 | well, i tried feisty will all the speedups today on the same machine with over 90 sec
| |
| 15:36 | seems the kernel is faster
| |
| 15:37 | ok, now without udevsettle
| |
| 15:38 | <ace_suares> ogra, vagrantc, if you have a sec. Can anyone write a spec on edubuntu ? I.e. me, and will this attract comments on the spec ?
| |
| 15:38 | <ogra> (thats all still with the feisty ldm btw)
| |
| 15:39 | hmm same numbers
| |
| 15:39 | so udevsettle doesnt gain anything on that machine
| |
| 15:39 | ace_suares, anyone with an LP account can write specs
| |
| 15:42 | <ace_suares> ogra: and is ruby accepted within edubuntu or has everything to be in python ?
| |
| 15:43 | Q-FUNK has joined #ltsp | |
| 15:45 | * vagrantc can't wait to rip apart the kiosk plugin | |
| 15:45 | <ogra> yay
| |
| 15:45 | e2300 witout X config starts as well
| |
| 15:45 | but ldm is a joke ...
| |
| 15:46 | takig 2/3 of the boottime
| |
| 15:46 | lol, 3 min 12 sec
| |
| 15:46 | Q-FUNK has quit IRC | |
| 15:46 | <ogra> ace_suares, well pythin is surely preferred
| |
| 15:46 | i'm not even sure ruby is in main
| |
| 15:46 | *python
| |
| 15:47 | ah, 1.8 is in main
| |
| 15:48 | <vagrantc> ogra: you realize, in feisty, that update-sshkeys gets run once as an Ubuntu plugin and once as a common plugin ?
| |
| 15:49 | hmmm...
| |
| 15:49 | <ogra> oh207, well, we'll drop it then
| |
| 15:49 | <vagrantc> maybe our pattern matching should strip off the number part somehow ...
| |
| 15:49 | <ogra> it doesnt do any harm to run it twice
| |
| 15:49 | <vagrantc> yeah, i know
| |
| 15:49 | which is probably why you didn't catch it
| |
| 15:50 | <ogra> yup
| |
| 15:50 | Q-FUNK has joined #ltsp | |
| 15:52 | <vagrantc> if you want to "move" the order of a plugin from common, you essentially have to put an empty place-holder plugin in your directory
| |
| 15:52 | one of the flaws of the design...
| |
| 15:54 | <ogra> woah xorg 7.3 rocks so hard :) all our probs are solvd it seems :)))))
| |
| 15:54 | <vagrantc> what you're saying is X configuration seemed to be the primary drag ?
| |
| 15:54 | <ogra> i just heard it can even read partial xorg.conf files so we could use it for keymaps etc
| |
| 15:54 | <vagrantc> or is it faster beyond that, even ?
| |
| 15:54 | <ogra> its faster beyond that
| |
| 15:55 | <vagrantc> oh yeah, that was my first question... partial configuration files
| |
| 15:55 | <ogra> i never managed to boot the e2300 in 3 minutes with an unmodified feisty chroot
| |
| 15:55 | but we wont need any config files, xmodmap and xrandr at runtime will do
| |
| 15:56 | <vagrantc> xrandr to set the resolution?
| |
| 15:56 | <jammcq> ogra: so, how fast is it now?
| |
| 15:56 | <ogra> so we can bring up a greyed out ldm with hourglass cursor and entertain the user by setting up resolution etc
| |
| 15:56 | jammcq, 3 mins atm
| |
| 15:56 | but with the old ldm and no speedups at all
| |
| 15:56 | <jammcq> what was the speed just before changing?
| |
| 15:56 | <ogra> plain feisty chroot
| |
| 15:56 | sell with all my poking 2:13 or so
| |
| 15:57 | standard feisty was way over 4 min iirc
| |
| 15:57 | <jammcq> i'm looking for a comparison between autoconfig and no-autoconfig
| |
| 15:57 | <ogra> ldm taking 1.5 min of that
| |
| 15:57 | jammcq, oh, ok
| |
| 15:57 | on the t-1220 it gains 20 seconds
| |
| 15:57 | 77 vs 97 secs
| |
| 15:57 | <jammcq> i'm guessing you ripped out the code that does the detection
| |
| 15:58 | whoa
| |
| 15:58 | nice
| |
| 15:58 | <ogra> i ripperd out all X related code
| |
| 15:58 | but gutsy feels faster overall
| |
| 15:58 | <jammcq> is it just turned off? or did you remove all the code?
| |
| 15:58 | <ogra> i turned it off for now
| |
| 15:58 | <jammcq> so you could turn it back on, eh?
| |
| 15:58 | <ogra> but its a script so the code isnt even in ram or anywhere
| |
| 15:58 | right
| |
| 15:59 | oh207, the 2:13 on the e2300 were with a static xorg.conf ...
| |
| 15:59 | <jammcq> i'm just curious about the speed difference caused by Xorg7.3 only
| |
| 15:59 | everything else being equal
| |
| 16:00 | * vagrantc loves ogra's funny irc client completions | |
| 16:00 | <jammcq> :)
| |
| 16:00 | <ogra> 7.3 seems to start up slower, it takes 2.3 secs to get to the X cursor compared to ! or less sec
| |
| 16:00 | <vagrantc> uh-oh. lights are turning off.
| |
| 16:00 | <ogra> s/\!/1
| |
| 16:01 | <vagrantc> they's gonna kick me out soon.
| |
| 16:01 | <ogra> *2-3
| |
| 16:01 | where are you ?
| |
| 16:01 | not in your hostel ?
| |
| 16:01 | <vagrantc> hostal closes common spaces at 11
| |
| 16:02 | <ogra> and you get no wlan outside the ommon space i gues
| |
| 16:02 | s
| |
| 16:02 | <vagrantc> not sure. not enough electrical outlets
| |
| 16:02 | and no battery life
| |
| 16:02 | <ogra> ah
| |
| 16:02 | meh
| |
| 16:02 | time for new HW
| |
| 16:02 | <vagrantc> well, newer :)
| |
| 16:03 | <ogra> (yes yes ... new refurbished HW)
| |
| 16:03 | :)
| |
| 16:03 | * vagrantc makes one quick push | |
| 16:04 | racter has joined #ltsp | |
| 16:04 | <vagrantc> ogra: well, take a look at the mess i made of mainline when you get a chance :)
| |
| 16:04 | <ogra> i will, for sure :)
| |
| 16:05 | <vagrantc> i think the controlgen stuff could be improved, for sure. but it's orders of magnitude better than what we were doing.
| |
| 16:06 | * vagrantc wants to see DIRECTX in action | |
| 16:07 | <ogra> its cool
| |
| 16:07 | but you really only see a difference in playing movies etc
| |
| 16:07 | <vagrantc> and server load
| |
| 16:07 | when you've got 20 some clients
| |
| 16:07 | <ogra> indeed
| |
| 16:08 | * vagrantc was given the freindly reminder to clear out | |
| 16:08 | * sutula wonders what DIRECTX is in this context...pointer? | |
| 16:08 | <ogra> unencrypted X traffic via ldm
| |
| 16:08 | <vagrantc> ah hell. i subscribed to mainline and launchpad is spamming me with all my commits
| |
| 16:09 | <ogra> oh207, it does that now ?
| |
| 16:09 | doesnt seem to wrk for me
| |
| 16:09 | <sutula> ogra: Darn...I'm already doing that
| |
| 16:09 | <ogra> sutula, how ?
| |
| 16:10 | <sutula> ogra: Logins via XDMCP instead of through ssh
| |
| 16:10 | <ogra> thats crap
| |
| 16:10 | <vagrantc> well
| |
| 16:10 | it's normal
| |
| 16:10 | <ogra> i said "unencrypted X traffic via *ldm*" ;)
| |
| 16:10 | <vagrantc> but LDM is better :)
| |
| 16:11 | * sutula agrees w/r/t security, but wonders whether there are other issues | |
| 16:11 | <sbalneav> Speaking of LDM, I got the ssh subsystem to change a password last night :)
| |
| 16:11 | <ogra> yay
| |
| 16:11 | <sbalneav> Hey, one thing I've run into.
| |
| 16:11 | <vagrantc> sutula: the cool thing is that the login happens over SSH, but the traffic is not encrypted... so should be comparable to XDMCP in speed
| |
| 16:11 | <ogra> sbalneav, did you notice ? we dont need any X detection anymore :D
| |
| 16:11 | <sbalneav> I modified an edgy LDM here to do the setting DISPLAY, as per gadi's patch.
| |
| 16:12 | gnome-screensaver stopped working.
| |
| 16:12 | <ogra> yea, its cool, isnt it ?
| |
| 16:12 | oh
| |
| 16:12 | right
| |
| 16:12 | <sutula> vagrantc: k, I was wishing for more speed (and that's not going to happen) but the better security would be good
| |
| 16:12 | <ogra> i'll check that,can you file a bug ?
| |
| 16:12 | <sbalneav> Sure.
| |
| 16:12 | I need to check and see if it does it via XDMCP
| |
| 16:12 | theoretically, it shouldn't work there either.
| |
| 16:13 | <ogra> sbalneav, so with your 20 secs from ldm and my 15-20 secs for X detection we should be near the 4.2 boottime :)
| |
| 16:13 | <vagrantc> damn, that'd be nice.
| |
| 16:13 | <sbalneav> Sexy.
| |
| 16:13 | <ogra> sbalneav, might be an xauth issue
| |
| 16:13 | or debus
| |
| 16:13 | <sbalneav> aaaaaahhhhhh
| |
| 16:13 | <vagrantc> without maintaining a custom kernel, libc, or X :)
| |
| 16:13 | <ogra> likely dbus
| |
| 16:13 | <sbalneav> yeah, maybe
| |
| 16:13 | hm
| |
| 16:14 | we'll have to figure that one out.
| |
| 16:14 | <ogra> vagrantc, well, that doesnt speed up the e2300 still ..
| |
| 16:14 | sbalneav, consolekit should help
| |
| 16:14 | <sbalneav> buh?
| |
| 16:14 | Consolekit?
| |
| 16:14 | <ogra> we need to provide conslekit integration with ldm anyway
| |
| 16:14 | <sbalneav> pointer please
| |
| 16:14 | <ogra> i gave you the link before UDS when you asked about it
| |
| 16:14 | <sbalneav> yeah, think I still have it? :)
| |
| 16:15 | <vagrantc> oh yeah
| |
| 16:15 | night all, or day, or afternoon, or other appropriate parting greeting
| |
| 16:15 | <ogra> the thing that defines sessions as seats for dbus and hal usage
| |
| 16:15 | vagrantc has quit IRC | |
| 16:15 | lynn_ has joined #ltsp | |
| 16:15 | <sbalneav> I'm lucky if I remember to put my pants on before my shoes :)
| |
| 16:15 | Night
| |
| 16:15 | <ogra> it doesnt have a homepage ... must be anywhere on f.d.o ... i'm diggin
| |
| 16:16 | <sbalneav> ok
| |
| 16:16 | yeah, googles got me there.
| |
| 16:16 | <lynn_> has anybody gotten a powerpc thin-client to work?
| |
| 16:16 | <ogra> http://gitweb.freedesktop.org/?p=ConsoleKit.git
| |
| 16:17 | framework for defining and tracking users, sessions and seats
| |
| 16:17 | ConsoleKit is a system daemon for tracking what users are logged into the system and how they interact with the computer (e.g. which keyboard and mouse they use).
| |
| 16:17 | thats from the debian experimental package
| |
| 16:17 | hal and friends will depend o it at soe point
| |
| 16:17 | *some
| |
| 16:19 | Egyptian[Home] has joined #ltsp | |
| 16:24 | sbalneav has quit IRC | |
| 16:27 | Gadi has left #ltsp | |
| 16:38 | lambda_ has joined #ltsp | |
| 16:40 | Burgundavia has joined #ltsp | |
| 16:41 | sepski has quit IRC | |
| 16:54 | Guaraldo has left #ltsp | |
| 16:54 | nicoAMG has joined #ltsp | |
| 16:55 | joebaker has quit IRC | |
| 16:56 | joebaker has joined #ltsp | |
| 16:56 | Q-FUNK has quit IRC | |
| 16:57 | lambda_ has quit IRC | |
| 16:57 | Sunborne has joined #ltsp | |
| 16:59 | <Sunborne> Hello everybody. Sunborne here - I'm looking for some advice on auto-login. I'm running MueKow on Kubuntu.
| |
| 16:59 | jammcq has quit IRC | |
| 16:59 | <Sunborne> My LTSP Server auto-logs in but clients don't.
| |
| 17:00 | In my xdmrc I have Xdmcp enable=true.
| |
| 17:00 | nicoAMG has quit IRC | |
| 17:01 | nicoAMG has joined #ltsp | |
| 17:01 | yanu_ is now known as yanu | |
| 17:02 | efra has quit IRC | |
| 17:02 | <Sunborne> I also have section like: [X-ws001:0-Core] NoPassEnable=true NoPassUsers=ltspguest AutoLoginEnable=true AutoLoginUser=ltspguest
| |
| 17:04 | Also strange - when client boots up my server's console goes blank.
| |
| 17:05 | Be right back - gonna go do some more googling on this problem...
| |
| 17:14 | Egyptian[Home1 has joined #ltsp | |
| 17:16 | lambda_ has joined #ltsp | |
| 17:23 | Egyptian[Home] has quit IRC | |
| 17:24 | <Sunborne> I'm back.
| |
| 17:26 | ohmbr has quit IRC | |
| 17:33 | nicoAMG has quit IRC | |
| 17:36 | irule has quit IRC | |
| 17:37 | jabular has left #ltsp | |
| 17:38 | lambda_ has quit IRC | |
| 17:42 | Sunborne has left #ltsp | |
| 17:51 | muh2000_ has joined #ltsp | |
| 17:51 | muh2000 has quit IRC | |
| 17:55 | bobby_C has quit IRC | |
| 17:55 | <ltsppbot> "racter" pasted "/etc/ltsp/dhcpcd.conf" (28 lines) at http://pastebot.ltsp.org/152
| |
| 18:11 | muh2000_ is now known as muh2000 | |
| 18:14 | Egyptian[Home1 has quit IRC | |
| 18:28 | <ltsppbot> "racter" pasted "/var/lib/tftpboot/ltsp/powerpc/yaboot.conf" (8 lines) at http://pastebot.ltsp.org/153
| |
| 18:29 | GodFather has joined #ltsp | |
| 19:07 | cliebow_ has quit IRC | |
| 19:08 | cliebow_ has joined #ltsp | |
| 19:27 | jammcq has joined #ltsp | |
| 20:02 | gepatino has joined #ltsp | |
| 20:04 | tessier_ has quit IRC | |
| 20:12 | sbalneav has joined #ltsp | |
| 20:12 | <sbalneav> Evening all
| |
| 20:18 | ogra has quit IRC | |
| 20:19 | <jammcq> !s
| |
| 20:19 | <ltspbot`> jammcq: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
| 20:35 | <jammcq> sbalneav: ping
| |
| 20:37 | <sbalneav> Pong
| |
| 20:38 | <jammcq> hey, did you see that guy that supplied a patch for ltspfs ?
| |
| 20:38 | on the ltsp-developer list
| |
| 20:38 | <sbalneav> No
| |
| 20:38 | When was this?
| |
| 20:39 | <jammcq> today
| |
| 20:39 | if the media is bad, the read returns an error, and exits from the routine without releasing a mutex
| |
| 20:39 | yer on the ltsp-developer list, aren't you?
| |
| 20:39 | <sbalneav> Yeah, but I haven't looked at it for a few days
| |
| 20:39 | <jammcq> well......
| |
| 20:40 | <sbalneav> Yeah, I see it.
| |
| 20:41 | <jammcq> pretty cool, eh?
| |
| 20:42 | <sbalneav> Well, he's only patched _read, I'll have a look through and see if theres anywhere else where there's a mutex not being freed
| |
| 20:42 | <jammcq> yeah. at least _read fixes something
| |
| 20:43 | <sbalneav> If I forgot it there, I probably forgot it in a few other places.
| |
| 20:43 | I'll have a looksee tonight.
| |
| 20:43 | * jammcq is ordering a Dell desktop machine with Ubuntu pre-loaded | |
| 20:44 | pimpministerp has quit IRC | |
| 20:53 | <cliebow_> 8~)
| |
| 20:54 | gonzaloaf has quit IRC | |
| 20:56 | gepatino has quit IRC | |
| 21:00 | GodFather has quit IRC | |
| 21:27 | a5benwillis has joined #ltsp | |
| 21:51 | amadeu has joined #ltsp | |
| 22:06 | sutula has quit IRC | |
| 22:06 | viking-ice has quit IRC | |
| 22:07 | sutula has joined #ltsp | |
| 22:07 | viking-ice has joined #ltsp | |
| 22:08 | ltsppbot has quit IRC | |
| 22:08 | oh207 has quit IRC | |
| 22:08 | parsley has quit IRC | |
| 22:08 | <mistik1> jammcq: WoW!!!
| |
| 22:11 | parsley has joined #ltsp | |
| 22:11 | oh207 has joined #LTSP | |
| 22:12 | ltsppbot has joined #ltsp | |
| 22:13 | <jammcq> mistik1: hey buddy
| |
| 22:13 | <mistik1> hey man
| |
| 22:13 | <jammcq> you sitting on the edge of your seat?
| |
| 22:13 | <mistik1> This is the best game i've seen a while
| |
| 22:13 | <jammcq> yeah
| |
| 22:14 | <mistik1> I think its safe to say that cleveland has arrived
| |
| 22:14 | <jammcq> they are definately playing some good ball
| |
| 22:15 | <mistik1> In the words of Rasheed, "Both teams played hard" ;-)
| |
| 22:21 | ChinaForge has joined #ltsp | |
| 22:23 | <ChinaForge> Hello,How to get LTSP 5.0 rdesktop screen script?
| |
| 22:24 | <sbalneav> Hmmm, doesn't look like we have one in there yet.
| |
| 22:24 | Probably just need to port it across from 4.2
| |
| 22:24 | <ChinaForge> :(,Thank you.
| |
| 22:25 | * sbalneav adds it to the todo list | |
| 22:25 | <jammcq> sbalneav: Gadi's prolly already got one. that's fairly important to him
| |
| 22:25 | <sbalneav> Good point. I'll check with him next time I see him on
| |
| 22:26 | <jammcq> good plan.
| |
| 22:32 | mistik1: i'm getting a bad feeling
| |
| 22:38 | <mistik1> this may go double
| |
| 22:38 | icredible
| |
| 22:38 | <jammcq> jeeeeeeeeez
| |
| 22:38 | <mistik1> incredible even
| |
| 22:39 | jammcq: this is one game you should have baught a ticket for
| |
| 22:41 | Burgundavia_ has joined #ltsp | |
| 22:42 | Burgundavia has quit IRC | |
| 22:44 | Burgundavia_ is now known as Burgundavia | |
| 22:52 | <jammcq> mistik1: can you say "3T" ?
| |
| 22:52 | <mistik1> All I can say is WOW!!!
| |
| 22:54 | Its a game of attrition
| |
| 22:56 | jammcq: what exactly are we seeing?
| |
| 22:57 | brings sharply to mind another #23
| |
| 22:59 | <jammcq> and then..... depression sets in
| |
| 23:00 | <mistik1> sorry bro, but you must admit this is one for the history books
| |
| 23:00 | <jammcq> that's what they're all saying
| |
| 23:00 | * jammcq misses the old team | |
| 23:01 | <mistik1> I think part of the key to this game was that ejection
| |
| 23:01 | I thought the same of the phoenix suspensions
| |
| 23:02 | gonzaloaf_work has quit IRC | |
| 23:03 | gonzaloaf_work has joined #ltsp | |
| 23:06 | tiagovaz has joined #ltsp | |
| 23:12 | joebaker has left #ltsp | |
| 23:22 | sbalneav has quit IRC | |
| 23:27 | tiagovaz has quit IRC | |
| 23:39 | daya has joined #ltsp | |
| 23:57 | ctyler has quit IRC | |