IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 19 August 2008 (all times are UTC)
| 00:42 | juanpaul has quit IRC | |
| 00:41 | juanpaul has quit IRC | |
| 00:55 | dubinsky has joined #LTSP | |
| 00:55 | dubinsky has joined #LTSP | |
| 01:09 | dubinsky has quit IRC | |
| 01:09 | dubinsky has quit IRC | |
| 01:18 | tarzeau has quit IRC | |
| 01:18 | tarzeau has quit IRC | |
| 01:36 | Egyptian[Home] has quit IRC | |
| 01:36 | Egyptian[Home] has quit IRC | |
| 01:55 | Egyptian[Home] has joined #ltsp | |
| 01:55 | Egyptian[Home] has joined #ltsp | |
| 02:02 | ltspbot` has joined #ltsp | |
| 02:02 | ltspbot` has joined #ltsp | |
| 02:03 | Loto_ has joined #ltsp | |
| 02:04 | shogunx has quit IRC | |
| 02:04 | Loto has quit IRC | |
| 02:04 | lts9bot has quit IRC | |
| 02:04 | rauz has quit IRC | |
| 02:03 | Loto_ has joined #ltsp | |
| 02:04 | shogunx has quit IRC | |
| 02:04 | Loto has quit IRC | |
| 02:04 | lts9bot has quit IRC | |
| 02:04 | rauz has quit IRC | |
| 02:04 | C7 has quit IRC | |
| 02:04 | C7 has quit IRC | |
| 02:04 | mu3en has left #ltsp | |
| 02:04 | mu3en has left #ltsp | |
| 02:06 | rauz has joined #ltsp | |
| 02:06 | rauz has joined #ltsp | |
| 02:10 | C7 has joined #ltsp | |
| 02:10 | C7 has joined #ltsp | |
| 02:10 | gonzaloaf_laptop has joined #ltsp | |
| 02:10 | gonzaloaf_laptop has joined #ltsp | |
| 02:16 | bobby_C has joined #ltsp | |
| 02:16 | bobby_C has joined #ltsp | |
| 02:16 | mikkel has joined #ltsp | |
| 02:16 | mikkel has joined #ltsp | |
| 02:23 | sbalneav has quit IRC | |
| 02:23 | sbalneav has quit IRC | |
| 02:24 | shogunx has joined #ltsp | |
| 02:24 | shogunx has joined #ltsp | |
| 02:58 | gonzaloaf_laptop has quit IRC | |
| 02:58 | gonzaloaf_laptop has quit IRC | |
| 03:08 | ltspbot has quit IRC | |
| 03:13 | ltspbot has joined #ltsp | |
| 03:23 | ltspbot_ has joined #ltsp | |
| 03:23 | ltspbot_ has joined #ltsp | |
| 04:10 | jonnor has joined #ltsp | |
| 04:14 | ltspbot_ has quit IRC | |
| 04:14 | ltspbot` has joined #ltsp | |
| 04:14 | ltspbot` has joined #ltsp | |
| 04:18 | ltspbot` has quit IRC | |
| 04:18 | ltspbot` has joined #ltsp | |
| 04:21 | ltspbot` has joined #ltsp | |
| 04:23 | soneyka has joined #ltsp | |
| 04:24 | ltspbot` has quit IRC | |
| 04:24 | sep has quit IRC | |
| 04:24 | ogra has quit IRC | |
| 04:34 | ltspbot_ has joined #ltsp | |
| 04:36 | ltspbot` has joined #ltsp | |
| 04:37 | ltspbot` has joined #ltsp | |
| 04:53 | The_Code has joined #ltsp | |
| 04:53 | The_Code has joined #ltsp | |
| 04:55 | ogra has quit IRC | |
| 04:55 | ogra has quit IRC | |
| 05:27 | ltspbot has joined #ltsp | |
| 05:37 | ogra has joined #ltsp | |
| 05:43 | ogra has quit IRC | |
| 05:44 | ogra has joined #ltsp | |
| 05:57 | J45p3r___ has joined #ltsp | |
| 06:14 | rjune has joined #ltsp | |
| 08:38 | ltspbot has joined #ltsp | |
| 08:38 | <sbalneav> ltspbot: status
| |
| 08:38 | <ltspbot> sbalneav: I am connected to freenode as ltspbot.
| |
| 08:39 | <ogra> hp, well, in ubuntu we say for a gnome desktop per user you need 128M per session
| |
| 08:39 | plus about 256M to run the server itself
| |
| 08:39 | that should be fine with 2G
| |
| 08:39 | <Q-FUNK> !s
| |
| 08:39 | <ltspbot> Q-FUNK: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
| 08:39 | <ogra> if you have runing desktops that use less resources you will need less
| |
| 08:40 | * ogra hugs sbalneav | |
| 08:40 | <ogra> gracias :)
| |
| 08:40 | <hp> oh thanks
| |
| 08:40 | <sbalneav> NP
| |
| 08:40 | Q-FUNK: Heya
| |
| 08:40 | <Q-FUNK> :)
| |
| 08:43 | <hp> ogra, if i use ubuntu with ltsp 5, default settings mean, even the first thin client I get to boot off the terminal server would be able to detect a usb disk?
| |
| 08:43 | <ogra> yep
| |
| 08:43 | <hp> without me having to configure lts.conf yet?
| |
| 08:43 | <ogra> right
| |
| 08:43 | <hp> only with the printers do I have to modify the lts.conf
| |
| 08:43 | ok I'll try that then
| |
| 08:43 | <ogra> sound, local devices are all set up by default nowadays
| |
| 08:44 | lts.conf is only used for custom setups or to work around bugs
| |
| 08:44 | <hp> wow, I really have to try that
| |
| 08:44 | I'll just follow the howto link you pointed me to :)
| |
| 08:45 | mikkel has quit IRC | |
| 08:47 | <lejo> is there still some development done on ltsp4 ?
| |
| 08:48 | <ogra> not since about 3 years
| |
| 08:48 | <lejo> great ;)
| |
| 08:48 | no better argument to convince people to go to 5 ;)
| |
| 08:50 | <hp> ogra, I'll follow your advice, good night and thank you
| |
| 08:50 | <ogra> :)
| |
| 08:50 | elisboa has joined #ltsp | |
| 08:51 | hp has quit IRC | |
| 09:06 | Q-FUNK has quit IRC | |
| 09:19 | sbalneav has quit IRC | |
| 09:19 | sbalneav_ has joined #ltsp | |
| 09:22 | bobby_C has quit IRC | |
| 09:28 | Gadi has joined #ltsp | |
| 09:33 | <sbalneav_> Hmmm, am I still here?
| |
| 09:35 | <laga> 2no
| |
| 09:36 | cliebow has quit IRC | |
| 09:43 | Q-FUNK has joined #ltsp | |
| 09:49 | <Gadi> !s
| |
| 09:49 | <ltspbot> Gadi: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
| 09:49 | <Gadi> cool, ltspbot is back
| |
| 09:50 | <jonnor> !l
| |
| 09:50 | <ltspbot> jonnor: Error: "l" is not a valid command.
| |
| 09:51 | K_O-Gnom has joined #ltsp | |
| 09:53 | <sbalneav_> Man, I hope the new cablemodem fixes things.
| |
| 09:57 | Q-FUNK has quit IRC | |
| 10:06 | gonzaloaf_laptop has joined #ltsp | |
| 10:07 | Faithful has joined #ltsp | |
| 10:10 | gonzaloaf_laptop has quit IRC | |
| 10:16 | staffencasa has joined #ltsp | |
| 10:17 | gonzaloaf_laptop has joined #ltsp | |
| 10:25 | warren has quit IRC | |
| 10:29 | alekibango has quit IRC | |
| 10:35 | mikkel has joined #ltsp | |
| 10:36 | nicoAMG has joined #ltsp | |
| 10:56 | twinprism has joined #ltsp | |
| 11:13 | rjune has quit IRC | |
| 11:23 | otavio has quit IRC | |
| 11:25 | <_UsUrPeR_> warren: any luck with the core dump I shot you yesterday?
| |
| 11:25 | mccann has joined #ltsp | |
| 11:26 | cliebow has joined #ltsp | |
| 11:33 | cliebow has quit IRC | |
| 11:39 | warren has joined #ltsp | |
| 11:45 | Egyptian[Home] has quit IRC | |
| 11:45 | Egyptian[Home] has joined #ltsp | |
| 11:48 | Lns has joined #ltsp | |
| 11:51 | <Lns> Can I use LTSP over a serial connection to a Commodore 64?
| |
| 11:51 | * Lns ducks | |
| 11:51 | cliebow has joined #ltsp | |
| 12:04 | cliebow has quit IRC | |
| 12:09 | K_O-Gnom has quit IRC | |
| 12:46 | etyack has joined #ltsp | |
| 12:47 | Subhodip has joined #ltsp | |
| 12:50 | Subhodip has quit IRC | |
| 12:52 | Subhodip has joined #ltsp | |
| 12:54 | Subhodip has quit IRC | |
| 12:55 | Egyptian[Home] has quit IRC | |
| 12:57 | |Paradox| has quit IRC | |
| 13:00 | warren has quit IRC | |
| 13:03 | sepski has joined #ltsp | |
| 13:07 | spectra has joined #ltsp | |
| 13:08 | sepone has joined #ltsp | |
| 13:11 | gonzaloaf_laptop has quit IRC | |
| 13:13 | spectra has quit IRC | |
| 13:15 | spectra has joined #ltsp | |
| 13:15 | cliebow has joined #ltsp | |
| 13:16 | spectra has quit IRC | |
| 13:18 | spectra has joined #ltsp | |
| 13:19 | spectra has quit IRC | |
| 13:21 | spectra has joined #ltsp | |
| 13:22 | spectra has joined #ltsp | |
| 13:23 | spective has joined #ltsp | |
| 13:23 | sepski has quit IRC | |
| 13:24 | <spective> I've just upgraded an Edubuntu/LTSP install to Hardy, and now the root filesystem is getting mounted r/o too soon, and so /etc/fstab can't be created to allow /tmp to be mounted r/w... much suffering ensues. Help? Please? :/
| |
| 13:25 | Do I need to remove some package? Reconfigure something?
| |
| 13:25 | sepski has joined #ltsp | |
| 13:26 | <spective> :(
| |
| 13:27 | sepone has quit IRC | |
| 13:28 | <Lns> spective: did you rebuild the client chroot after the upgrade?
| |
| 13:28 | <spective> Lns: no... I was hoping to avoid it, since I custom set up local apps etc...
| |
| 13:28 | heh. It's supposed to just "work". (hahah. sigh.)
| |
| 13:29 | <Lns> spective: it's generally advised to do that after a dist-upgrade...
| |
| 13:29 | well
| |
| 13:29 | not if you upgrade half of your environment ;)
| |
| 13:29 | jammcq has joined #ltsp | |
| 13:29 | <jammcq> hello all
| |
| 13:29 | <Lns> hi jammcq
| |
| 13:29 | <spective> well, I upgraded the client chroot... didn't rebuild it though.
| |
| 13:29 | So technically, I upgraded both halves.
| |
| 13:29 | <Lns> hmm..not sure if that's advisable or not..
| |
| 13:30 | did you ltsp-update-image?
| |
| 13:30 | spectra has quit IRC | |
| 13:30 | <spective> no... was not away of such.
| |
| 13:30 | s/away/aware/
| |
| 13:30 | <Lns> sudo ltsp-update-image - this is necessary
| |
| 13:30 | <spective> running it now...
| |
| 13:31 | * spective crosses 6 fingers. | |
| 13:32 | <etyack> afternoon jammcq
| |
| 13:33 | Egyptian[Home] has joined #ltsp | |
| 13:33 | <jammcq> etyack: hey
| |
| 13:33 | spectra has joined #ltsp | |
| 13:34 | sepone has joined #ltsp | |
| 13:34 | spectra has joined #ltsp | |
| 13:34 | <spective> finished. no effect.
| |
| 13:35 | Pascal_1 has joined #ltsp | |
| 13:36 | spectra has quit IRC | |
| 13:36 | spectra has joined #ltsp | |
| 13:37 | <Lns> spective: i would really recommend rebuilding your chroot
| |
| 13:37 | spectra has quit IRC | |
| 13:37 | <spective> Lns: sigh.... yes.
| |
| 13:37 | How is that best accomplished?
| |
| 13:37 | <Lns> when you dist-upgraded the chroot, did it even pull from the hardy repos?
| |
| 13:38 | <spective> yes.
| |
| 13:38 | ltsp-build-client, yes?
| |
| 13:38 | spectra has joined #ltsp | |
| 13:38 | sepone has quit IRC | |
| 13:38 | Pascal_1 has quit IRC | |
| 13:38 | <Lns> spective: i usually rename my current chroot and then ltsp-build-client, yes
| |
| 13:38 | sepone has joined #ltsp | |
| 13:39 | <Lns> so as to keep your old modifications handy
| |
| 13:39 | <spective> yes.
| |
| 13:39 | running....
| |
| 13:40 | is there a way to specify a diferent mirror?
| |
| 13:40 | I have a nearby one that is faster...
| |
| 13:40 | <Lns> yes
| |
| 13:40 | can't recall, man page has it though
| |
| 13:40 | <spective> ah yes. I see it.
| |
| 13:44 | Much nicer pulling at 800KB/s...
| |
| 13:44 | <Lns> hehe
| |
| 13:45 | local repo?
| |
| 13:45 | <spective> eh, not quite. Just a under-utilized server which a ton upload bandwidth. It's a mirror I run.
| |
| 13:46 | well, it's done downloading and is setting it up now....
| |
| 13:46 | K_O-Gnom has joined #ltsp | |
| 13:47 | sepski has quit IRC | |
| 13:48 | <spective> There was small mod I made to the udev configs... so that they would handle filesystems on devices as well as partitions... which apparently happens with usb drives...
| |
| 13:48 | _UsUrPeR_ has quit IRC | |
| 13:48 | <spective> w00t. done.
| |
| 13:49 | _UsUrPeR_ has joined #ltsp | |
| 13:49 | alexC has joined #ltsp | |
| 13:51 | <spective> w00t! and we have X!
| |
| 13:51 | now lets see how much works...
| |
| 13:51 | sound works...
| |
| 13:53 | <Lns> nice
| |
| 13:53 | always good to follow the directions. ;)
| |
| 13:55 | <spective> verray.
| |
| 13:55 | Ok, but I have an issue with my ssh keys and local apps...
| |
| 13:55 | <Lns> ssh-keys - try ltsp-update-sshkeys
| |
| 13:55 | <spective> ok.
| |
| 13:55 | <Lns> local apps - you're on your own. never used them. :)
| |
| 13:56 | <spective> they work well. little tricky to set up, but other than that, they're great.
| |
| 13:56 | nice for graphically intensive apps.
| |
| 13:57 | <Lns> spective: did you follow any howtos on localapps or did you do it all yourself?
| |
| 13:57 | <spective> followed howtos, but had to hack a bit on my own.
| |
| 13:57 | <Lns> I'd love to get google earth going
| |
| 13:57 | do you have a URL for the howto you followed?
| |
| 13:57 | <spective> incidentally, ltsp-update-sshkeys, does not fix the local app issue.
| |
| 13:57 | Lns: no. It was a year ago.
| |
| 13:57 | I've forgotten.
| |
| 13:57 | <Lns> ah
| |
| 13:58 | well yeah, ssh-keys won't fix localapps
| |
| 13:58 | just updates ssh keys
| |
| 13:59 | <spective> brb
| |
| 13:59 | ok... lets see. We have ssh keys in three places.
| |
| 14:00 | on the server in /etc/ssh
| |
| 14:00 | on the client image in /etc/ssh
| |
| 14:00 | and on the client image in /root/.ssh/authorized_hosts
| |
| 14:01 | <_UsUrPeR_> Need some help with DVI support in Fedora 9 w/ LTSP5. I am running a brand spanking new TK-3752/1G Dual-output client
| |
| 14:01 | <spective> When we run a local app, we are using ssh on the server to connect to the client's ssh server...
| |
| 14:01 | staffencasa has quit IRC | |
| 14:01 | <spective> so...
| |
| 14:01 | The server's pub keys need to be in the clients /root/.ssh/authorized hosts file. I think.
| |
| 14:02 | <Lns> spective: you got me there. =p
| |
| 14:02 | <spective> lol
| |
| 14:02 | <_UsUrPeR_> I was able to get the same client to work in Ubuntu with a few additions to the lts.conf
| |
| 14:02 | <spective> The thing is, the local apps icon on the server has to make a key-based authentication connection, so it doesn't ask for a password, and just works instead.
| |
| 14:04 | And... I think the server must be in the clients /etc/ssh/known_hosts file....
| |
| 14:05 | <Gadi> other way around
| |
| 14:05 | makghosh has joined #ltsp | |
| 14:05 | <Gadi> oh, no you're right
| |
| 14:06 | actually, no - other way around
| |
| 14:06 | <spective> rofl. Let me know when you decide.
| |
| 14:06 | <Gadi> hehe
| |
| 14:06 | if you are ssh'ing from server to client
| |
| 14:06 | then, the client needs to authorize you
| |
| 14:06 | and the server needs to know who the client is
| |
| 14:07 | so, the client's public key should be in the server's ssh_known_hosts
| |
| 14:07 | and the server should have a key-pair such that whoever is launching the app, has the public key in the client's .authorized_keys file
| |
| 14:08 | man - our new local apps stuff works much better than this
| |
| 14:08 | :)
| |
| 14:08 | silly keys....
| |
| 14:09 | just be sure you don't give your users free reign to run anything they like as root on the client
| |
| 14:09 | :)
| |
| 14:09 | vagrantc has joined #ltsp | |
| 14:10 | * Gadi waves to vagrantc | |
| 14:10 | <spective> In this situation, I don't think that will be too much of a risk...
| |
| 14:10 | * vagrantc waves to Gadi | |
| 14:10 | <spective> given my users... heh
| |
| 14:11 | <score> when using the qemu-ltsp script and trying to login under ldm, i get "this workstation isn't authorized.." can anyone explain?
| |
| 14:11 | <Gadi> ah, thats what they all say
| |
| 14:11 | * vagrantc feels doomed to support tcsh patches for debian unto all time | |
| 14:11 | <score> it appears that it works locally and from thin clients, but i'm trying to test remotely
| |
| 14:11 | <vagrantc> score: i haven't tested that script in a long, long time ... may not really work anymore.
| |
| 14:11 | <spective> Gadi, 6th graders? They get excited about "paint"
| |
| 14:11 | <Gadi> score: sudo ltsp-update-sshkeys in your server vm
| |
| 14:11 | <score> Gadi: i've done that and rebuilt the image
| |
| 14:11 | <Gadi> spective: be careful - one of our best developers is 13
| |
| 14:12 | :)
| |
| 14:12 | <vagrantc> score: the issue will likely be that it's getting a fake ip address for the server, and your ssh keys aren't updated
| |
| 14:12 | <Gadi> score: yeah, make sure your thin client subnet is static IP
| |
| 14:12 | <vagrantc> score: well, the ssh keys will never get updated, because it has no idea it needs to use the fake ip address
| |
| 14:13 | <Gadi> s/subnet/interface/
| |
| 14:13 | <score> vagrantc: in the qemu-ltsp window, it seems to have the ip 10.0.2.15. there aren't any interfaces on the machine with 10. nets
| |
| 14:13 | <spective> odd, it appears that the ltsp-build-client script did not create ssh keys in /etc/ssh
| |
| 14:13 | no... that's probably because openssh-server is not on the clients...
| |
| 14:14 | <score> vagrantc: so, i need to edit the qemu-ltsp script's ip range?
| |
| 14:14 | <vagrantc> score: correct, it's using qemu's virtual ip addreses.
| |
| 14:14 | score: you might have luck by specifying LDM_SERVER=real.ip.of.server in lts.conf
| |
| 14:15 | <spective> Is there a way to prevent the openssh-server postinst script from attempting to restart the openssh server?
| |
| 14:15 | <jonnor> 13 year old ltsp dev? impressive
| |
| 14:15 | <spective> eh?
| |
| 14:15 | oh
| |
| 14:16 | look, if one of my 6th graders hacks the ltsp install, I'll hire him.
| |
| 14:16 | or her.
| |
| 14:17 | <Lns> child labor!
| |
| 14:18 | =p
| |
| 14:18 | <spective> heh
| |
| 14:19 | <Gadi> spective: what are you using for homedirs?
| |
| 14:19 | and what apps are you running locally?
| |
| 14:20 | most apps like users
| |
| 14:20 | :)
| |
| 14:20 | <spective> Gadi: tuxpaint, tuxtype, tuxmath, gcompris, mostly.
| |
| 14:20 | <Gadi> nothing that saves settings?
| |
| 14:20 | or files?
| |
| 14:20 | or prints?
| |
| 14:21 | <spective> I'm using "standard" homedirs. They don't print from any of the local apps...
| |
| 14:21 | don't think they save either.
| |
| 14:21 | <Lns> no saving in tuxpaint? :(
| |
| 14:22 | <Gadi> ok - just be aware - the road you are heading down allows for none of the above
| |
| 14:22 | but, our new local apps support will hopefully address all those issues
| |
| 14:22 | <spective> Gadi: I need none of the above for the local apps I use.
| |
| 14:22 | Gadi, what is the status of the new local app support for a hardy-based system?
| |
| 14:23 | <Gadi> well, as ogra would say "you can request a backport" ;)
| |
| 14:23 | but, I think if anything, it will show up in intrepid
| |
| 14:23 | <spective> targeted at intrepid?
| |
| 14:23 | ah.
| |
| 14:23 | So what approach is being used for the new method?
| |
| 14:24 | <Gadi> it will work out of the box and you won't even know it
| |
| 14:24 | you just need to install apps in the chroot
| |
| 14:24 | (its a mixture of ssh auth + sshfs + Xatom launching + xdg menu magic)
| |
| 14:24 | <spective> wild.
| |
| 14:24 | <Gadi> but it should be mostly transparent
| |
| 14:25 | cyberorg has everything but the xdg menu magic in his opensuse packages atm
| |
| 14:25 | <spective> nice.
| |
| 14:25 | <Gadi> he likes merging the newly minted still in flux code ;)
| |
| 14:26 | <spective> Insanity has its merits.
| |
| 14:26 | elisboa has quit IRC | |
| 14:26 | <Gadi> anyhow, the Xatoms stuff makes your ssh key stuff not necessary
| |
| 14:26 | <spective> delightful.
| |
| 14:26 | <Gadi> which should save you some headaches
| |
| 14:27 | :)
| |
| 14:27 | <spective> yes. since my keys are still not working.
| |
| 14:27 | elisboa has joined #ltsp | |
| 14:28 | <Gadi> ah, right - so to get them to work, you should make a key-pair that all can read on the server
| |
| 14:28 | and then tell your launchers to use that keypair
| |
| 14:28 | and make sure the public key is in: $chroot/root/.ssh/authorized_keys
| |
| 14:29 | (use ssh -i .... in the launcher)
| |
| 14:29 | <spective> pondering.
| |
| 14:30 | ok... trying.
| |
| 14:31 | jimjimovich has joined #ltsp | |
| 14:34 | <jimjimovich> can someone give me a tip on where/how to add monitor sync values to lts.conf in LTSP-5? I keep breaking my ubuntu ltsp server trying to do it. I get the "this workstation isn't authorized..." error
| |
| 14:35 | <johnny> that's an ssh error usally..
| |
| 14:35 | of the server IP changing
| |
| 14:35 | <Gadi> the workstaion authorized has nothing to do with sync values
| |
| 14:35 | <spective> jim, does it work without the sync values?
| |
| 14:35 | <jimjimovich> hmmm
| |
| 14:36 | well, it broke (possibly because of IP change). Then I fixed it by rebuilding the ltsp image. then i changed the lts.conf and it broke again
| |
| 14:36 | <Lns> jimjimovich: did you sudo ltsp-update-sshkeys?
| |
| 14:36 | <jimjimovich> yes
| |
| 14:37 | it was working, then i simply changed the lts.conf and it broke again. oh wait, i guess i just got a blinking cursor then
| |
| 14:37 | (sorry, did it hours ago, just nobody was here at that time)
| |
| 14:37 | <Gadi> whats the IP of your thin client interface?
| |
| 14:37 | <Lns> jimjimovich: reversing your changes in lts.conf causes it to work again?
| |
| 14:37 | makghosh has quit IRC | |
| 14:38 | <Gadi> do: grep <that-IP> /opt/ltsp/i386/etc/ssh/ssh_known_hosts
| |
| 14:38 | warren has joined #ltsp | |
| 14:38 | <warren> Is any other distro using nbd-server with the -a timeout option?
| |
| 14:38 | <Gadi> not ubuntu
| |
| 14:39 | <warren> because if a client disappears without disconnecting from nbd-server (the norm with LTSP), then nbd-server is sitting there for 2 hours by default
| |
| 14:39 | <jimjimovich> Gadi: did the grep, got nothing
| |
| 14:40 | <Gadi> jimjimovich: do sudo ltsp-update-sshkeys again
| |
| 14:40 | if your IP changes, then make it not
| |
| 14:40 | :)
| |
| 14:40 | <Lns> =p gadi
| |
| 14:40 | <johnny> and the ltsp-update-image ?
| |
| 14:40 | <Gadi> johnny: indeed
| |
| 14:41 | <jimjimovich> Gadi: the IP changed on the server because of a bad NIC. But after I fixed that, I redid the ltsp image, ssh keys and everything. then it worked. the thing i don't understand is why it broke when i changed lts.conf
| |
| 14:42 | <Gadi> it should not have
| |
| 14:42 | do that grep thing to make sure things are updated
| |
| 14:42 | <jimjimovich> what is the error usually when you just get a flashing cursor right before the login screen?
| |
| 14:42 | what should i see in the grep?
| |
| 14:42 | <Gadi> you should see a line in that file with your IP address
| |
| 14:42 | <jimjimovich> the ip of the server or the client?
| |
| 14:43 | <Gadi> the reason you are not authorizzed is that your server is not a known host
| |
| 14:43 | the server
| |
| 14:43 | <Lns> jimjimovich: blinking cursor before LDM is usually just the monitor/video being detected on TC
| |
| 14:43 | <Gadi> the client needs to "know" your server in the biblical sense
| |
| 14:43 | <jimjimovich> Gadi: :)
| |
| 14:44 | Gadi: actually, i'm sorry, it wasn't the not authorized thing, it was the blinking cursor thing, got confused
| |
| 14:44 | <Lns> Is anyone interested in looking at an introductory document to Ubuntu + LTSP (mainly created for new administrators to understand how it works and common-task solutions) ?
| |
| 14:44 | <jimjimovich> i see nothing but the server IP in the known_hosts file
| |
| 14:44 | <Lns> I just finished it and would appreciate any feedback
| |
| 14:45 | babyhuey has quit IRC | |
| 14:46 | <jimjimovich> is this the correct path for the lts.conf? /opt/ltsp/i386/usr/share/doc/ltsp-client-core/examples/lts.conf
| |
| 14:46 | sorry, wrong path :(
| |
| 14:46 | sysdef has left #ltsp | |
| 14:46 | <jimjimovich> /var/lib/tftpboot/ltsp/i386/lts.conf
| |
| 14:47 | <spective> Odd. It appears that the ssh server is not running on the client.
| |
| 14:47 | <jimjimovich> and would this be a reasonable entry there?
| |
| 14:47 | [ws63]
| |
| 14:47 | X_HORZSYNC = 30-85;
| |
| 14:47 | X_VERTREFRESH = 50-160;
| |
| 14:47 | X_MODE_1 = 1024x768;
| |
| 14:47 | <Lns> If anyone is interested in checking my doc out, you can see it here: http://logicalnetworking.net/other/UbuntuLinuxLTSPIntroductoryGuide.pdf
| |
| 14:47 | I'll bbl..lunch time! wooo
| |
| 14:47 | <jimjimovich> Lns: Thanks!
| |
| 14:48 | <Gadi> jimjimovich: thats quite a wide range - are you sure your monitor can handle it?
| |
| 14:49 | <jimjimovich> Gadi: that's what the documentation says ...
| |
| 14:49 | <Gadi> also, is ws63 specified anywhere?
| |
| 14:49 | try by MAC address instead of hostname
| |
| 14:49 | <jimjimovich> yes, in dhcp and in the hosts file
| |
| 14:49 | <spective> anyhow. Time to go.
| |
| 14:49 | cya'll later.
| |
| 14:49 | spective has quit IRC | |
| 14:49 | <jimjimovich> ok
| |
| 14:50 | <Gadi> jimjimovich: then, make a shell on the client and make sure it knows its own hostname
| |
| 14:50 | <jimjimovich> after making changes to that file, does anything need to be done?
| |
| 14:50 | <Gadi> no
| |
| 14:50 | just reboot
| |
| 14:50 | <cliebow> in the biblical sense
| |
| 14:50 | ;-]
| |
| 14:50 | <Gadi> oh my
| |
| 14:51 | <jimjimovich> reboot the client, right?
| |
| 14:51 | <Gadi> right
| |
| 14:51 | <sbalneav_> Lns: is this in addition to the other documentation I wrote? Sure, I'll have a look.
| |
| 14:51 | <jimjimovich> Lns: the file won't open for me on Linux
| |
| 14:51 | Lns: never mind
| |
| 14:51 | fie_wr0k has joined #ltsp | |
| 14:51 | <vagrantc> warren: tried nbd-server with -a timeout, and this was my result: http://bugs.debian.org/409531
| |
| 14:52 | <Lns> sbalneav_: where's your documentation?
| |
| 14:52 | <jimjimovich> does the lts.conf need a [Default] section?
| |
| 14:53 | <sbalneav_> Lns: bzr branch lp:~ltsp-docwriters/ltsp/ltsp-docs-trunk
| |
| 14:53 | <johnny> yes
| |
| 14:53 | <warren> vagrantc: the keepalive value of 7200 seconds is default in the kernel
| |
| 14:53 | vagrantc: you could make that shorter, but then it gets set for all other processes as well, which might not be desired
| |
| 14:54 | <vagrantc> warren: note the post from ronny aasen
| |
| 14:54 | <sbalneav_> I did a bunch of original stuffg on edubuntu, and at the recent hackfest, btilma started up our own doctree.
| |
| 14:54 | If you've got some stuff, lets shoehorn it into there.
| |
| 14:54 | <warren> vagrantc: I see
| |
| 14:54 | * vagrantc waves to sbalneav_ | |
| 14:55 | <vagrantc> warren: basically, it will disconnect you weather you're using it or not. which is not particularly useful.
| |
| 14:55 | <sbalneav_> Het vagrantc!
| |
| 14:55 | <johnny> uggh.. the bts date is during the red and black ball that i have volunteered to help out with .. oops :( .. i guess now i gotta judge whether it's going to end up needing me...
| |
| 14:55 | <Lns> sbalneav_: i'll have a look - thanks.. I created this doc for my (mostly) non-Linux oriented school lab technicians to get a feel for Linux and LTSP
| |
| 14:55 | it's more of an overview than official documentation
| |
| 14:55 | but hopefully a good introduction to how things work
| |
| 14:55 | <sbalneav_> Good, well, lets get it in there anyway. The more the merrirer
| |
| 14:56 | <Lns> i'll bbl
| |
| 14:56 | bobby_C has joined #ltsp | |
| 14:57 | <jimjimovich> hey, thanks guys for the advice. will have to test it in the morning.
| |
| 15:01 | <cliebow> johnny:of Course you need to be ther
| |
| 15:01 | e
| |
| 15:01 | <warren> huh
| |
| 15:01 | my squashfs mount over nbd is failing
| |
| 15:01 | attempt to access beyond end of device
| |
| 15:01 | K_O-Gnom has quit IRC | |
| 15:01 | <warren> nbd0: rw=0, want=459222, limit=131072
| |
| 15:02 | cliebow has quit IRC | |
| 15:03 | <sbalneav_> Hmmm, that's bad.
| |
| 15:03 | All the copyright notices have been stripped out of the manual.
| |
| 15:03 | jimjimovich has left #ltsp | |
| 15:06 | <warren> has anyone noticed nbd has limits before?
| |
| 15:07 | pimp^air has joined #ltsp | |
| 15:08 | <pimp^air> hi
| |
| 15:08 | i'm looking for a way to install my ubuntu onto a high-power server at the company and use it via ltsp from home or when i'm on the road
| |
| 15:09 | <warren> 131072 seems to be a special number
| |
| 15:09 | <pimp^air> however i only find information about using ltsp on the local network
| |
| 15:09 | warren: its a power of 2
| |
| 15:09 | ist 128M ram:)
| |
| 15:09 | remember the counter
| |
| 15:09 | <Gadi> warren: you are limited by the file size being exported
| |
| 15:09 | thats the limit
| |
| 15:10 | it is exporting 131072 blocks
| |
| 15:10 | <pimp^air> so my question: is there a way to connect to ltsp via a network connections and ssh
| |
| 15:10 | <Gadi> warren: and you are trying to read 459222
| |
| 15:10 | <pimp^air> i did not install ltsp yet, so a "yes or no" would help
| |
| 15:10 | <warren> Gadi: yeah, something is confused here
| |
| 15:10 | pimp^air: ltsp is exclusively for super high bandwidth local networks
| |
| 15:10 | pimp^air: you want something else like VNC or NX for over the internet
| |
| 15:10 | pimp^air: which have nothing to do with LTSP
| |
| 15:11 | <pimp^air> ok, i know those two... vnc is to slow
| |
| 15:11 | what is super high bandwidth?
| |
| 15:11 | especially the latency would have to be small i guess
| |
| 15:11 | at home i have 16mbit
| |
| 15:11 | <warren> pimp^air: like 10mbit/sec minimum
| |
| 15:11 | <Gadi> pimp^air: use NX
| |
| 15:11 | <pimp^air> but only downstream
| |
| 15:11 | ok, i'll have a look at NX
| |
| 15:12 | i tried it once or twice, but without looking further into it
| |
| 15:12 | <Gadi> it is your best bet
| |
| 15:12 | <pimp^air> however i finally want to get rid of local software installation
| |
| 15:12 | i really hate my computer for their noise and for braking kaput
| |
| 15:12 | e
| |
| 15:12 | <warren> Gadi: Creating little endian 3.1 filesystem on client-fedora-9-i386.img, block size 131072
| |
| 15:12 | Gadi: this is making little sense..
| |
| 15:13 | <Gadi> sounds like thats the size of your squashfs image
| |
| 15:13 | <warren> the squashfs image is 351367168 bytes
| |
| 15:14 | <Gadi> thats bytes not blocks
| |
| 15:15 | <warren> block size 131072?
| |
| 15:15 | Gadi: what does mksquashfs say when you're creating the nbd image on ubuntu?
| |
| 15:15 | Egyptian[Home] has quit IRC | |
| 15:18 | <Gadi> same thing
| |
| 15:18 | block size 131072
| |
| 15:18 | pimp^air has quit IRC | |
| 15:18 | <Gadi> but, that makes sense
| |
| 15:18 | default block size is 1024M
| |
| 15:18 | er, 1024k
| |
| 15:19 | er
| |
| 15:19 | I dunno what Im saying
| |
| 15:20 | <warren> 1024k == 131072?
| |
| 15:20 | staffencasa has joined #ltsp | |
| 15:21 | <Gadi> 128*1024 = 131072
| |
| 15:21 | <warren> where is 128 coming from?
| |
| 15:21 | <Gadi> so, it seems that mksquashfs uses 128k blocks
| |
| 15:21 | <alexC> pimp^air: you could have a local LTSP run network, a server with two NICs and SSH to your server to monitor the network
| |
| 15:22 | <Gadi> So, your image should be 2681 blocks
| |
| 15:22 | of 128kbytes each
| |
| 15:22 | warren: is this all on a virtual client?
| |
| 15:23 | <warren> yes
| |
| 15:23 | mount -t squashfs /dev/nbd0 -o ro /sysroot
| |
| 15:23 | switchroot /sysroot
| |
| 15:30 | captain_1agnus has joined #ltsp | |
| 15:30 | babyhuey has joined #ltsp | |
| 15:31 | <warren> Gadi: how big are the ubuntu nbd images?
| |
| 15:32 | <Gadi> about 173M
| |
| 15:32 | or so
| |
| 15:37 | ltspbot has joined #ltsp | |
| 15:42 | captain_magnus has quit IRC | |
| 15:54 | otavio has joined #ltsp | |
| 15:56 | captain_magnus has joined #ltsp | |
| 15:58 | mikkel has quit IRC | |
| 16:03 | spectra has quit IRC | |
| 16:07 | _UsUrPeR_ has quit IRC | |
| 16:07 | etyack has quit IRC | |
| 16:08 | captain_1agnus has quit IRC | |
| 16:13 | alekibango has joined #ltsp | |
| 16:16 | <vagrantc> yay! openssh 5.1 is in debian lenny :)
| |
| 16:16 | warren has quit IRC | |
| 16:17 | <vagrantc> now if i can just cleanly distinguish between older versions, so we can do "kill -1 $PPID" on older versions, and "ssh -O exit" on newer versions ...
| |
| 16:20 | <Gadi> yay!
| |
| 16:20 | did you share that news with the rest of the gang?
| |
| 16:20 | (the ssh -O thing)
| |
| 16:21 | warren has joined #ltsp | |
| 16:22 | <Gadi> vagrantc: you are sure, btw, that it was the ssh version and not the dbus version that made the difference?
| |
| 16:23 | <vagrantc> Gadi: i downgraded to the old openssh, and it didn't work. i upgraded openssh, and it worked. multiple times. no other changes.
| |
| 16:24 | there may be other factors, but that was at least one key point.
| |
| 16:24 | <Gadi> cool
| |
| 16:24 | <vagrantc> Gadi: i didn't yet share the "ssh -O exit" stuff ... i've been meaning to post to a debian bug report about it and cc ltsp-developer
| |
| 16:24 | <Gadi> sucks to have to do a: ssh ... ssh -V
| |
| 16:24 | :P
| |
| 16:25 | <vagrantc> ?
| |
| 16:25 | <Gadi> well, I assume that's how you will be able to tell
| |
| 16:25 | <vagrantc> haven't figured out how to make it work.
| |
| 16:25 | <Gadi> serverSSHVersion=$(ssh -S ${LTSP_SOCKET} ${LTSP_VERSION} /usr/bin/ssh -V)
| |
| 16:26 | <vagrantc> that would tell you the client version
| |
| 16:26 | <Gadi> the version of the server's client
| |
| 16:26 | <vagrantc> not necessarily the server version ... though they're *likely* to be the same.
| |
| 16:27 | <Gadi> I think its a fair assumption
| |
| 16:28 | <vagrantc> the other method i thought would be to use distro-specific plugins to ldminfod
| |
| 16:28 | to pass the version
| |
| 16:28 | <Gadi> but that assumes the LTSP server = app server
| |
| 16:29 | <vagrantc> well, currently. but i think ldminfod should really belong in it's own package, along with dependencies for things that belong on app servers like a desktop and so on.
| |
| 16:30 | <Gadi> oh - ldminfod
| |
| 16:30 | i was thinking l-b-c
| |
| 16:30 | how about just an lts.conf variable - LDM_LEGACY_MODE
| |
| 16:30 | <vagrantc> that was my other approach
| |
| 16:31 | but it's really a per-server configuration, not a per client configuration.
| |
| 16:31 | <Gadi> we are working around a bug
| |
| 16:31 | true
| |
| 16:31 | <vagrantc> and our way of communicating server capabilities is currently ldminfod
| |
| 16:31 | <warren> I'm not sure what is going on with nbd and squashfs mounts
| |
| 16:31 | it seems that squashfs is using more than 131072 blocks, but nbd0 thinks 131072 is the limit.
| |
| 16:31 | <Gadi> vagrantc: I am hoping your thinking is to do ssh -O unless told otherwise by whatever mechanism, right?
| |
| 16:32 | <warren> if I use nbd-client on the server itself and mount locally it has no problem
| |
| 16:32 | <vagrantc> Gadi: yes.
| |
| 16:32 | Gadi: well, we have to append the "kill -1 $PPID" stuff onto the starting of the Xsession command, and nix the "ssh -O exit" when that happens.
| |
| 16:33 | <Gadi> hey - here's an idea:
| |
| 16:33 | can you run ssh -O in the background and put a sleep after it
| |
| 16:33 | then follow up with a kill
| |
| 16:33 | that way, if ssh -O does hang, it will be killed by the kill
| |
| 16:33 | and if it doesn't everything exits nicely
| |
| 16:34 | <warren> that sounds like something Gadi would think of
| |
| 16:34 | =)
| |
| 16:34 | <Gadi> :)
| |
| 16:34 | <vagrantc> Gadi: no, because the "ssh -O exit" doesn't ever happen unless you append the kill on the initial command.
| |
| 16:34 | <Gadi> why are we still appending the kill?
| |
| 16:34 | <johnny> hmm... you guys find odd bugs
| |
| 16:35 | <Gadi> why can't we issue an ssh ... kill?
| |
| 16:35 | should be the same PPID
| |
| 16:35 | no?
| |
| 16:35 | <vagrantc> Gadi: because without the kill on older versions, it hangs on logout and never runs the K* scripts
| |
| 16:35 | <Gadi> oh...
| |
| 16:35 | I thought it was ssh -O that was causing the hang
| |
| 16:35 | <vagrantc> it's a particularly annoying bug.
| |
| 16:36 | <Gadi> it never runs ssh -O?
| |
| 16:36 | <vagrantc> correct. just hangs.
| |
| 16:36 | <Gadi> that still makes no sense to me
| |
| 16:36 | because it shouldn't run the kill either
| |
| 16:36 | :)
| |
| 16:36 | <vagrantc> well, it runs ssh -O exit after you manually kill the process, and then ssh -O hangs because it still has a socket, but the socket doesn';t go anywhere.
| |
| 16:37 | it makes perfect sense to me, and i'm at a loss how to explain it futher.
| |
| 16:37 | <warren> btw, what benefit does this ssh -O thing get us?
| |
| 16:37 | <Gadi> warren: it is the key to making everything clean
| |
| 16:37 | <vagrantc> there's basically three phases ... ssh to establish the tunnel, ssh to start the X session, and then after the X session ends.
| |
| 16:37 | warren: it properly does pam_logout
| |
| 16:38 | or pam_close_session, or whatever it's called.
| |
| 16:38 | jammcq has quit IRC | |
| 16:39 | <vagrantc> Gadi: so right now, we do kill $PPID on the tail end of starting the X session ... without it, on older versions of ssh, it hangs.
| |
| 16:39 | and $PPID is the pid of the sshd process
| |
| 16:39 | <Gadi> right
| |
| 16:39 | but "hangs" is non specific
| |
| 16:39 | it obviously continues to process more commands
| |
| 16:40 | because otherwise "kill" would never be executed
| |
| 16:40 | so, does it prevent ldm from continuing to call the rc.d scripts?
| |
| 16:40 | you say yes, I think
| |
| 16:41 | <vagrantc> when it hits the end of "... Xsession ; ltspfsmouter all cleanup" it doesn't close the ssh connection.
| |
| 16:41 | <Gadi> right
| |
| 16:41 | so, it doesn't return to the ldm controlling process
| |
| 16:41 | <vagrantc> leaving ldm waiting till that call to ssh ends ...
| |
| 16:42 | <Gadi> what if you did: ....Xsession ; ltspfsmounter ...; exit 0
| |
| 16:42 | <vagrantc> Gadi: that's actually the next thing i planned on doing ... had hoped to test it yesterday, but got distracted but a dozen other things...
| |
| 16:42 | <Gadi> ah, ok
| |
| 16:42 | its really not that "it holds open the ssh socket" or anything
| |
| 16:43 | its merely that the ssh child process does not close
| |
| 16:43 | which means the ldm parent process does not proceed to call the rc.d K scripts
| |
| 16:44 | <vagrantc> i don't know exactly what is doing it, but that's the end result, yes.
| |
| 16:45 | sepone has quit IRC | |
| 16:59 | mhterres has quit IRC | |
| 17:23 | Gadi has left #ltsp | |
| 17:32 | bobby_C has quit IRC | |
| 18:15 | ltspbot has joined #ltsp | |
| 18:37 | alekibango has quit IRC | |
| 18:53 | staffencasa has quit IRC | |
| 19:00 | Lns has quit IRC | |
| 19:27 | vagrantc has quit IRC | |
| 19:33 | jammcq has joined #ltsp | |
| 19:41 | <chupacabra> hidy Jim
| |
| 19:47 | <jammcq> hey chupacabra
| |
| 19:47 | how you doin?
| |
| 19:49 | <chupacabra> good, you?
| |
| 19:49 | <jammcq> doing well. very busy, but that's sort of normal
| |
| 19:49 | <chupacabra> yup
| |
| 19:51 | * chupacabra looks to see if gentoo has a ltsp package. | |
| 20:23 | <johnny> almost
| |
| 20:23 | it needs just a few adjustments to work with recent changes in ltsp-trunk, and ldm trunk
| |
| 20:24 | i'm just waiting for donnie to merge my changes and get me commit access, then we'll have the local apps stuff too
| |
| 20:24 | well.. it does have an ltsp 4 package, but it is masked
| |
| 20:25 | chupacabra, i'd suggest that you should wait a few days before adding the ltsp overlay
| |
| 20:25 | jammcq, anything you can do to help us get some official ltsp tarballs on ltsp.org? ones based off the tagged revisions?
| |
| 20:25 | that'll be the final thing we need
| |
| 20:35 | <chupacabra> will do
| |
| 20:36 | johnny:Are you the johnny from Postnuke?
| |
| 20:37 | hey there is a supybot in here.
| |
| 20:37 | whos is that?
| |
| 21:14 | fedora_ has joined #ltsp | |
| 21:14 | fedora_ is now known as petre | |
| 21:16 | <chupacabra> [C[Cwhy wont fedora 8 install on my machine?
| |
| 21:19 | <petre> warren, ping
| |
| 21:19 | <warren> petre: pong
| |
| 21:20 | <petre> do you know why if I call nautilus from command line or script, it mounts all the ext3 partitions that are on my hard disk?
| |
| 21:20 | I mean from the live usb
| |
| 21:21 | <warren> petre: wow, does it treat it like local devices?
| |
| 21:21 | <petre> yes, they all appear on the desktop
| |
| 21:21 | weird
| |
| 21:21 | <warren> petre: that's interesting, I bet they never tested it in this weird circumstance
| |
| 21:21 | petre: although I bet the same thing happens on a plain livecd
| |
| 21:22 | petre: the thing is, almost nobody uses a livecd for more than a demo
| |
| 21:22 | petre: or to install
| |
| 21:22 | <chupacabra> but
| |
| 21:22 | why wont fedora 8 install on my machine?
| |
| 21:22 | <petre> My thought was to have a script that called system-config-network, all the user to twiddle those parts, then call nautilus with the /etc/sysconfig/network-scripts directory
| |
| 21:22 | to let them edit the newly created ethX interface
| |
| 21:23 | <warren> chupacabra: that's kind of like "why wont my toyota start?"
| |
| 21:23 | <chupacabra> hehe
| |
| 21:23 | <warren> the usual answer is "it's been 18 years, time to buy a new one"
| |
| 21:23 | <chupacabra> i was hoping there was a known issue.
| |
| 21:23 | 8 is old
| |
| 21:23 | <petre> it doesn't prevent the steps from working, but I think it might freak the user out
| |
| 21:23 | <warren> how many users will run nautilus from a command line?
| |
| 21:24 | <chupacabra> none.
| |
| 21:24 | <petre> none, the point is to put it into a script
| |
| 21:24 | * chupacabra hates nautilus | |
| 21:24 | <warren> petre: why do you want it in a script?
| |
| 21:24 | petre: (btw, might want to test a F10 live cd to make sure that isn't still a bug, if it is please file)
| |
| 21:24 | <petre> click on one button on the desktop to launch the necessary pieces
| |
| 21:25 | have it automatically bring up system-config-network
| |
| 21:25 | let the user make necessary mods
| |
| 21:25 | <chupacabra> im using 8 because amahi only runs inn 8
| |
| 21:25 | <petre> after user closes that, have it open nautilus pointing at /etc/sysconfig/network-scripts
| |
| 21:26 | so user can right click on ethX, choosed gedit, make change
| |
| 21:26 | <warren> petre: might not be a bad thing for the user to know where to find the launcher for tools though.
| |
| 21:26 | especially that tool
| |
| 21:26 | and the change they make to the config file
| |
| 21:27 | <petre> it would be a good thing, but I'm not sure this would be the time to teach them paths & tools
| |
| 21:27 | <warren> because not knowing that change they can easily fuck their network later
| |
| 21:27 | <petre> yes, I suppose you're correct
| |
| 21:28 | I'm just trying to keep away from a command line if possible
| |
| 21:28 | user can open Computer icon on desktop, drill down to .../network-scripts
| |
| 21:28 | but won't do any good unless one is root
| |
| 21:29 | instructing the user to open a terminal and su - to root is not that big a deal, I suppose
| |
| 21:30 | but I keep thinking about the folks I get in my Linux Admin class who like to navigate graphically
| |
| 21:30 | CL is more powerful, IMHO, but I'm trying to make it as easy as possible
| |
| 21:33 | <warren> isn't this still easier than setting up K12LTSP?
| |
| 21:33 | fewer steps at least
| |
| 21:34 | <petre> debatable
| |
| 21:34 | k12ltsp is only about 6-8 clicks with the mouse and it installs everything
| |
| 21:34 | OTOH, there's no live demo of k12ltsp
| |
| 21:35 | aside from the doctored edubuntu CD I made
| |
| 21:35 | so in that sense, yes, this is miles ahead of that
| |
| 21:36 | The steps in the README are really quite easy
| |
| 21:36 | but I've been doing this kind of stuff for a while;
| |
| 21:37 | I fear a teacher trying it out might be inclined to run away if there's CL work
| |
| 21:37 | so, I'm trying to figure out a way to avoid it, if possible; maybe it's not possible
| |
| 21:38 | k12lstp may also get the nics backwards, so it's not foolproof
| |
| 21:38 | this approach seems less likely to do that
| |
| 21:41 | <warren> I'm really afraid of screwing your network with dhcpd
| |
| 21:41 | even I did it at work
| |
| 21:41 | twice
| |
| 21:41 | it wasn't pretty
| |
| 21:41 | they know it is me instantly now
| |
| 21:41 | =)
| |
| 21:42 | I'm thinking about adding a dhcp detector to the ltsp-dhcpd service
| |
| 21:42 | if you want to use it in a fancier way "dhcpd with failover" you have to flip a config option or something
| |
| 21:44 | <petre> I used to work for Nextel; when I started there in 2001, they wouldn't let me run linux because some engineer had previously unleashed a dhcp server on a linux box, made a mess of things
| |
| 21:44 | took me two years to get them to relent
| |
| 21:44 | <chupacabra> nont run dhcp on the ltsp server.
| |
| 21:44 | dont even
| |
| 21:44 | <petre> for a live usb demo, dhcp is necessary
| |
| 21:45 | <chupacabra> let their server point ot the ltsp server kernel
| |
| 21:45 | <petre> most teachers don't control the school's dhcp server
| |
| 21:46 | <chupacabra> ltsp over usb?
| |
| 21:46 | <petre> ltsp via usb (stick)
| |
| 21:46 | <chupacabra> ahh
| |
| 21:46 | <petre> live CD on usb stick
| |
| 21:46 | really cool, I typing this on it
| |
| 21:46 | i'm
| |
| 21:46 | I'm
| |
| 21:47 | <chupacabra> no, but you cant just throw another dhcp server on any network and not expect problems.
| |
| 21:47 | <warren> chupacabra: http://wtogami.livejournal.com/27610.html
| |
| 21:47 | chupacabra: the usual way of using LTSP is to have an isolated network segment
| |
| 21:47 | <chupacabra> that is your fault and problem
| |
| 21:47 | <warren> if you want to use it in ANY other way you need manual configuration
| |
| 21:47 | <chupacabra> warren: yes.
| |
| 21:48 | I been doing ltsp since the 90s
| |
| 21:48 | <jammcq> warren: chupacabra was doing ltsp back when you were.... ok, I won't say it :)
| |
| 21:48 | <chupacabra> lol
| |
| 21:48 | thanks Jim
| |
| 21:49 | * jammcq hopes he didn't offend warren | |
| 21:50 | <warren> jammcq: I was doing LTSP back around 2002
| |
| 21:50 | <chupacabra> my advice was basic sysadmin advice.
| |
| 21:51 | <warren> chupacabra: yes, except the target user these days doesn't know anything about networking =(
| |
| 21:51 | <jammcq> yeah, and chuppy was doing it before then
| |
| 21:51 | <chupacabra> like dont run a server on your workplace network that might get boingboinged.
| |
| 21:51 | <warren> chupacabra: things are so easy these days they don't even know they are running a server.
| |
| 21:51 | these users don't know what a server is
| |
| 21:52 | chupacabra: and with my live ltsp server, you could run a server without installing anything
| |
| 21:52 | which is both cool and scary
| |
| 21:52 | <chupacabra> I thought linux was the moving target, not the sheeple that depend on it.
| |
| 21:53 | <petre> that's the irony of ltsp: it's brilliant engineering, but to the user, it looks like a 'regular' computer
| |
| 21:53 | which is good in that it doesn't frighten users, but I don't think they always appreciate how clever it is
| |
| 21:53 | <jammcq> well kidz, i'm heading to bed. it's early, but i've been staying up WAAAAAY too late lately
| |
| 21:53 | <chupacabra> ya, "Nobody Cares" as lon as it works. and two dhcp servers on one network doesn't
| |
| 21:54 | nini Jim.
| |
| 21:54 | <petre> so, testing for an existing dhcp server would be a good thing, I agree
| |
| 21:55 | would you just do a dhcp request, see if you get an answer?
| |
| 21:55 | <chupacabra> maybe one could hack dhcpd to work no matter idf their was another dhcp server on the network but it would be a bitch
| |
| 21:56 | i thought about that but how do you know its range and what if it is 253 ips
| |
| 21:56 | <warren> petre: good thing yes, but totally not fool proof
| |
| 21:56 | petre: and it is time consuming
| |
| 21:56 | <chupacabra> back to square one in my book
| |
| 21:57 | <petre> wasn't that part of the point of the bridge?
| |
| 21:57 | <warren> petre: point of the bridge was to preconfigure all the services
| |
| 21:57 | <chupacabra> hmmm. lemmee think about that. putting em on a diff subnet
| |
| 21:58 | <petre> ah, right, then just connect it all to the desired interface via the bridge
| |
| 21:58 | <chupacabra> not optimal
| |
| 21:58 | <petre> perhaps I'm overthinking this
| |
| 21:59 | since the interface is going to the eth-something, is a newbie (teacher) likely to even know that's a network interface?
| |
| 21:59 | <warren> petre: don't you need to know this for K12LTSP as well?
| |
| 22:00 | <petre> sort of: it shows you, graphically, all the interfaces it sees
| |
| 22:00 | <warren> what shows you?
| |
| 22:00 | <petre> installer
| |
| 22:00 | <warren> they wrote a custom screen for the installer?
| |
| 22:01 | <petre> but it might get them backwards
| |
| 22:01 | <warren> (I haven't used K12LTSP since maybe 2003)
| |
| 22:01 | <petre> Eric might have
| |
| 22:01 | <johnny> chupacabra, yes
| |
| 22:01 | * johnny wonders if he remembers your first name.. | |
| 22:01 | <johnny> the name mike comes to my head
| |
| 22:01 | <warren> getting it backwards is really bad!
| |
| 22:01 | <petre> if it finds two nics, it automatically chooses one for clients, other for public
| |
| 22:02 | yes, it is, and sometimes what you chose during installation is not what you'd get after reboot
| |
| 22:02 | after installation finished
| |
| 22:02 | so, it wasn't bullet-proof
| |
| 22:02 | it was a regular question on the list
| |
| 22:03 | "I installed k12ltsp with no problems, but now my clients won't boot" kinda stuff
| |
| 22:05 | in my installs, it usually picks the gbit nic for eth1, not what I want, but I figure that's just my karma
| |
| 22:05 | and if the box has, say, two gbit nics, the user will have to choose, and may still get it backwards
| |
| 22:06 | which, as you say, is really bad
| |
| 22:06 | I don't know that there's a foolproof way to prevent it
| |
| 22:07 | I gotta go to bed, been up since 3:30 this morning
| |
| 22:08 | I've got a bunch of screen shots for assembling an html README, but as I said, I'd like to avoid a commandline if I can
| |
| 22:08 | because I think that may turn some people off
| |
| 22:08 | I'll sleep on it
| |
| 22:09 | petre has quit IRC | |
| 22:21 | <chupacabra> johnny: yes
| |
| 22:21 | was outside
| |
| 22:21 | cool
| |
| 22:22 | i remember that nick switch problwem
| |
| 22:22 | was fun for a few weeks
| |
| 22:22 | floppyfw did that too.
| |
| 22:23 | put the command line in a gtk box or something
| |
| 22:23 | oops
| |
| 22:23 | mccann has quit IRC | |
| 22:25 | shogunx has quit IRC | |
| 22:31 | shogunx has joined #ltsp | |
| 22:41 | twinprism has quit IRC | |
| 22:41 | mistik1 has quit IRC | |
| 22:41 | F-GT has quit IRC | |
| 22:41 | topslakr__ has quit IRC | |
| 22:41 | popey has quit IRC | |
| 22:41 | psycodad has quit IRC | |
| 22:41 | Faithful has quit IRC | |
| 22:41 | ace_suares has quit IRC | |
| 22:41 | johnny has quit IRC | |
| 22:41 | yanu_ has quit IRC | |
| 22:41 | ltsppbot has quit IRC | |
| 22:41 | score has quit IRC | |
| 22:41 | loather has quit IRC | |
| 22:41 | stgraber has quit IRC | |
| 22:41 | psycodad has joined #ltsp | |
| 22:42 | twinprism has joined #ltsp | |
| 22:42 | Faithful has joined #ltsp | |
| 22:42 | score has joined #ltsp | |
| 22:42 | mistik1 has joined #ltsp | |
| 22:42 | F-GT has joined #ltsp | |
| 22:42 | ace_suares has joined #ltsp | |
| 22:42 | johnny has joined #ltsp | |
| 22:42 | yanu_ has joined #ltsp | |
| 22:42 | popey has joined #ltsp | |
| 22:42 | topslakr__ has joined #ltsp | |
| 22:42 | ltsppbot has joined #ltsp | |
| 22:42 | loather has joined #ltsp | |
| 22:42 | stgraber has joined #ltsp | |
| 22:43 | Faithful has quit IRC | |
| 22:44 | Faithful has joined #ltsp | |
| 22:49 | X0d_of_N0d has quit IRC | |
| 22:59 | rjune has joined #ltsp | |
| 23:03 | vagrantc has joined #ltsp | |
| 23:11 | dchristiaan has joined #ltsp | |
| 23:11 | <dchristiaan> hi everyone]
| |
| 23:11 | i need some help ..
| |
| 23:13 | <vagrantc> !question
| |
| 23:13 | <ltspbot> vagrantc: "question" is if you have a question about ltsp, please go ahead and ask it, and people will respond if they can. please also mention the linux distro and release you're using. :)
| |
| 23:14 | Egyptian[Home] has joined #ltsp | |
| 23:14 | <dchristiaan> ok
| |
| 23:15 | i'm using OpenSuse 11 with KIWI-LTSP
| |
| 23:15 | i've install everythin
| |
| 23:15 | <warren> is anyone else annoyed at kiwi users coming here?
| |
| 23:15 | we can't help them
| |
| 23:15 | <dchristiaan> and successfully boot the client
| |
| 23:15 | <warren> kiwi is too different
| |
| 23:16 | <dchristiaan> ouch :(
| |
| 23:16 | when i tried to login on the client
| |
| 23:16 | <warren> dchristiaan: KIWI is a fork of LTSP upstream. LTSP behaves almost the same on all other distros. You need to get help from the KIWI community.
| |
| 23:17 | <dchristiaan> ok then .. thanks a lot
| |
| 23:17 | * Ryan52 thinks there needs to be a #kiwi | |
| 23:17 | <vagrantc> the basic concept of kiwi isn't soo different
| |
| 23:17 | the only difference really is fewer plugins, really.
| |
| 23:18 | for ltsp-build-client
| |
| 23:18 | <warren> the last few people cmoing here for kiwi help
| |
| 23:18 | they mentioned tools and config I've never heard
| |
| 23:18 | of
| |
| 23:19 | their networking config is completely different
| |
| 23:19 | I can't troubleshoot that
| |
| 23:19 | forget it, waste of time bitching about kiwi
| |
| 23:19 | * warren sleep | |
| 23:19 | <dchristiaan> i got error mesasge said --> ssh: Could not resolve hostname REPLACE_SERVER_IP: Temporary failure in name resolution
| |
| 23:20 | after that my client goes to logon windows again
| |
| 23:31 | Ok,, NOT talking abount KIWI
| |
| 23:31 | i need suggestion
| |
| 23:32 | <vagrantc> after defending the similarities with KIWI, my hunch here is that it's something totally different with kiwi-ltsp, and i have no idea how to fix it. :)
| |
| 23:32 | <dchristiaan> which one is better .... OpenSuse 11 + KIWI-LTSP or OpenSuse 11 + LTSP-5 ?
| |
| 23:32 | forget about my question .. i already fixed it :)
| |
| 23:32 | <vagrantc> sounds like REPLACE_SERVER_IP is something that should have gotten substituted with a real ip address, but didn't.
| |
| 23:33 | <dchristiaan> i forgot to change IP_SERVER on lts.conf :(
| |
| 23:33 | stupid things :)
| |
| 23:33 | <vagrantc> hmmm... IP_SERVER ... that also sounds like a kiwi-ltsp-ism
| |
| 23:33 | <dchristiaan> Bro, what do u think better is ? OS 11 + LTSP-5 or OS11 + KIWI-LTSP ?
| |
| 23:34 | <vagrantc> i don't think there is an ltsp5 implementation for opensuse, just kiwi-ltsp, which is similar to ltsp5 in basic concept.
| |
| 23:34 | but it seems like they haven't maintained a lot of the conventions followed by other ltsp5 distros.
| |
| 23:35 | but maybe that's because opensuse doesn't maintain a lot of conventions used by other distros ... i don't really know.
| |
| 23:36 | <dchristiaan> what about OpenSuse 11 + LTSP-4.2 ?
| |
| 23:36 | is it good ?
| |
| 23:36 | <vagrantc> ltsp 4.2 is basically unmaintained at this point.
| |
| 23:37 | <dchristiaan> so ? which one should i use to implemen LTSP on OpenSuse 11 ..
| |
| 23:38 | previously i success implenet OpenSuse 10.3+LTSP-4.2 at my office ..
| |
| 23:46 | dchristiaan has quit IRC | |