IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 6 October 2010 (all times are UTC)
| 00:06 | alkisg has joined #ltsp | |
| 00:16 | isetann has joined #ltsp | |
| 00:26 | alkisg has quit IRC | |
| 00:36 | artista_frustrad has joined #ltsp | |
| 00:43 | <isetann> anybody i can ask concerning ltsp clients?
| |
| 00:43 | hp has joined #ltsp | |
| 00:43 | <isetann> first time user of irc
| |
| 00:43 | <hp> hello is ogra here?
| |
| 00:45 | <muppis> isetann, don't ask to ask, ask. If someone knows, he/she answers.
| |
| 00:45 | <isetann> ok thanks for the advice
| |
| 00:47 | hp has quit IRC | |
| 01:02 | <sweetpi> isetann: you need to ask the question if you want someone is to answer it :)
| |
| 01:18 | <elias_a> Is there something new concerning LTSP in the 10.10. going to be released next Sunday?
| |
| 01:19 | Ubuntu Maverick Meerkat that is...
| |
| 01:20 | alkisg has joined #ltsp | |
| 01:51 | alexqwesa__ has quit IRC | |
| 01:53 | alexqwesa__ has joined #ltsp | |
| 01:54 | dean_za has quit IRC | |
| 02:02 | gnunux has joined #ltsp | |
| 02:04 | dean_za has joined #ltsp | |
| 02:06 | jimjimovich has joined #ltsp | |
| 02:07 | <jimjimovich> Hi guys, have some questions about mounting /home on NFS. does this essentially leave the whole /home unprotected on the network?
| |
| 02:09 | <Appiah> http://tldp.org/HOWTO/NFS-HOWTO/security.html
| |
| 02:09 | sweetpi has quit IRC | |
| 02:12 | <jimjimovich> I've seen that, but I was wondering how LTSP specifically deals with this, does it somehow use usernames and passwords or does it just leave /home open for everyone?
| |
| 02:13 | I need to get Chrome working as a local app, but would hate to throw out a lot of security in order to make that happen
| |
| 02:13 | was wondering if it'd be worth mounting some other sort of network drive where people's chrome profiles could be saved and bypass the NFS_HOME thing
| |
| 02:15 | alkisg has quit IRC | |
| 02:17 | alkisg has joined #ltsp | |
| 02:41 | spectra has quit IRC | |
| 02:43 | spectra has joined #ltsp | |
| 03:23 | mikkel has joined #ltsp | |
| 04:21 | <vmlintu> jimjimovich: nfs4 gives you better security with kerberos
| |
| 04:22 | bobby_C has joined #ltsp | |
| 04:23 | <muppis> Just switched to v4 due the locking issues, but without krb5
| |
| 04:27 | Da-Geek has joined #ltsp | |
| 04:28 | <alkisg> muppis: did it solve the locking issues?
| |
| 04:28 | And, any how-tos?
| |
| 04:29 | <muppis> alkisg, yes it did. https://help.ubuntu.com/community/NFSv4Howto
| |
| 04:30 | <alkisg> ty!
| |
| 04:33 | dobber has joined #ltsp | |
| 04:47 | methril_work has quit IRC | |
| 04:48 | <jimjimovich> so, i'm assuming that the standard setup with sshfs is pretty secure, right?
| |
| 04:50 | does CIFS allow for the creation of these Process Singleton files like Chrome wants to make?
| |
| 05:01 | <alkisg> CIFS doesn't even allow ownership + permissions...
| |
| 05:02 | Maybe we could file some bug reports for chrome + googleearth + evolution to use /var/run or /var/lock for their locks, though
| |
| 05:04 | <gnunux> use /Var/run ? omg
| |
| 05:04 | that really not a good idea
| |
| 05:04 | <alkisg> gnunux: why not? Doesn't /var/lock exist for that exact reason?
| |
| 05:06 | from fhs, http://www.pathname.com/fhs/2.2/fhs-5.9.html ==> "Lock files should be stored within the /var/lock directory structure. "
| |
| 05:07 | <gnunux> not for user's application
| |
| 05:07 | otherwise you have to set rwx for all users !
| |
| 05:07 | maybe add sticky bit but that not solve the problem
| |
| 05:08 | <alkisg> /var/lock does have a sticky bit
| |
| 05:08 | Hmmm nope, sorry :)
| |
| 05:08 | Only rwx
| |
| 05:09 | <gnunux> i'm talking about /var/run ;)
| |
| 05:23 | bobby_C has quit IRC | |
| 05:55 | methril_work has joined #ltsp | |
| 06:01 | methril_work has quit IRC | |
| 06:11 | methril_work has joined #ltsp | |
| 06:28 | <alkisg> !compiz
| |
| 06:28 | <ltspbot> alkisg: "compiz" :: if compiz is giving you problems, one way to disable it for all users is: sudo gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.mandatory --type string --set /desktop/gnome/session/required_components/windowmanager metacity
| |
| 06:30 | pmatulis has joined #ltsp | |
| 06:30 | Da-Geek has quit IRC | |
| 06:35 | toscalix has joined #ltsp | |
| 06:41 | <jimjimovich> okay, just throwing this idea out there to see what you guys think. what if you had some sort of /temp mounted with NFS and then linked the user's chrome directory to that temp directory? all the ideas i've found online revolve around using /tmp for this, but it's not very, well good at maintaining changes to a chrome profile
| |
| 06:51 | another question (from my boss). Is it possible to boot a client from a local hard drive but still have it connect to their /home folder on the server?
| |
| 06:52 | <alkisg> Sure, plain NFS without LTSP :)
| |
| 06:53 | <muppis> We use autofs to mount only those whom are logged in.
| |
| 06:53 | autofs+NFSv4
| |
| 06:54 | <jimjimovich> okay, let me rephrase his question and be a bit more specific
| |
| 06:54 | actually, i don't like his question, forget it
| |
| 06:54 | sorry
| |
| 06:54 | <muppis> Heh. :)
| |
| 06:55 | <jimjimovich> the idea of LTSP (for us anyway) is to make things simpler, not to complicate them :)
| |
| 06:58 | <muppis> I think things are more easier now even using mixed enviroment.
| |
| 06:59 | Only thing more that we made our own script to deploy changes to fats.
| |
| 07:00 | <jimjimovich> oh, for sure. i set up our first LTSP lab over 4 years ago, and at the time, it took me a week of work to get floppies (yeah, floppies!) and flash drives to work. Sound was a dream. Now, in less than a week, we have an almost operational system with everything our bosses want. Printers, scanners, skype, local Firefox, remote access, audio, video, etc
| |
| 07:01 | <muppis> Even I have set up only one production server (and several for testing purposes), tooks only 8h get it operational.
| |
| 07:02 | <jimjimovich> muppis: that's super! what distro/version you using?
| |
| 07:02 | <muppis> Ubuntu 10.04
| |
| 07:02 | <jimjimovich> me too
| |
| 07:03 | * alkisg needed about 1 years to develop some scripts for automatic ltsp deployment, but now each deployment takes about 10 clicks + 5 hours for the packages to get downloaded from the net :) | |
| 07:03 | * alkisg loves package management ;) | |
| 07:04 | <muppis> As we talk here, I'm setting yet another server for testing as virtual machine.
| |
| 07:04 | * muppis loves apt-cacher | |
| 07:04 | <muppis> Even we are in same building as local ISP, apt-cacher speed ups things to happen.
| |
| 07:13 | methril_work has quit IRC | |
| 07:14 | <jimjimovich> I've been using apt-proxy for several years with pretty good luck. haven't tried apt-cacher
| |
| 07:15 | <muppis> I tried apt-proxy in my home (3 machines..), but it hanged too often. apt-cacher has been working 4 months without any problem.
| |
| 07:15 | <alkisg> Are those really needed in LTSP labs?
| |
| 07:15 | <jimjimovich> yeah, i've had that hanging problem too
| |
| 07:16 | * alkisg tried squid as an apt cacher for non-LTSP labs, worked fine... | |
| 07:16 | <jimjimovich> well, we have a lot of ubuntu in the building, so it saves a lot of time :)
| |
| 07:16 | ogra_ac has quit IRC | |
| 07:16 | ogra_ac has joined #ltsp | |
| 07:18 | <jimjimovich> anyone using drobox or ubuntu one in an ltsp situation?
| |
| 07:18 | <muppis> alkisg, no really. Probably any proxy will do the trick.
| |
| 07:18 | <alkisg> muppis: I mean, for LTSP labs, the packages only need to be downloaded once...
| |
| 07:19 | <jimjimovich> alkisg: true, but if you got 2 labs, about 15 stand alone installs and random notebooks, a cache can make things go a lot faster :)
| |
| 07:19 | <alkisg> Sure
| |
| 07:20 | <muppis> That's my point.
| |
| 07:20 | And we are using apt-cacher for other server than just for LTSP.
| |
| 07:21 | <jimjimovich> so, I got this new setup basically ready to go, and i've learned a couple things. 1. Network bandwidth is the biggest problem. 2. I hate not having Chrome :)
| |
| 07:21 | well, I have chrome, but it's not so great as it is on a regular install
| |
| 07:22 | hersonls has joined #ltsp | |
| 07:23 | <jimjimovich> muppis: what are you doing with virtual machines?
| |
| 07:25 | <muppis> jimjimovich, they are for testing and for customers websites.
| |
| 07:25 | <jimjimovich> cool. i'm a big fan of virtual machines myself
| |
| 07:27 | <muppis> Like now, I've setup isolated virtual network for vm's to test LTSP.
| |
| 07:28 | <jimjimovich> :)
| |
| 07:30 | bobby_C has joined #ltsp | |
| 07:34 | isetann_ has joined #ltsp | |
| 07:37 | <jimjimovich> how would you auto mount an internal hard drive on a specific thin client?
| |
| 07:37 | jammcq has quit IRC | |
| 07:38 | <Appiah> just enable localdev?
| |
| 07:39 | oh maybe that does not enable disks..
| |
| 07:40 | <jimjimovich> it doesn't seem to work
| |
| 07:41 | alkisg has quit IRC | |
| 07:42 | <Appiah> is there no LOCAL_STORAGE paramater
| |
| 07:42 | <jimjimovich> probably need to write a special script to do it or something
| |
| 07:45 | ah, there's a LOCALDEV_DENY_INTERNAL_DISKS that defaults to True
| |
| 07:47 | <Appiah> ah :)
| |
| 07:47 | docente has joined #ltsp | |
| 07:53 | <jimjimovich> LOCALDEV_DENY_INTERNAL_DISKS=False auto mounts all local hard drives via sshfs in /media
| |
| 07:58 | this could be really useful for migrating someone from a stand-alone install to ltsp
| |
| 08:10 | jimjimovich has left #ltsp | |
| 08:18 | Gadi has joined #ltsp | |
| 08:18 | <Gadi> !compiz
| |
| 08:18 | <ltspbot> Gadi: "compiz" :: if compiz is giving you problems, one way to disable it for all users is: sudo gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.mandatory --type string --set /desktop/gnome/session/required_components/windowmanager metacity
| |
| 08:21 | ogra_ac has quit IRC | |
| 08:23 | bencer_ has quit IRC | |
| 08:34 | pmatulis has quit IRC | |
| 08:34 | ogra_ac has joined #ltsp | |
| 08:45 | pmatulis has joined #ltsp | |
| 08:46 | mikkel has quit IRC | |
| 08:59 | toscalix has quit IRC | |
| 09:09 | ogra_ac has quit IRC | |
| 09:10 | ogra_ac has joined #ltsp | |
| 09:36 | jammcq has joined #ltsp | |
| 09:36 | <jammcq> hey hey
| |
| 09:43 | dean_za has quit IRC | |
| 09:48 | mikkel has joined #ltsp | |
| 10:14 | staffencasa has joined #ltsp | |
| 10:16 | alkisg has joined #ltsp | |
| 10:17 | vagrantc has joined #ltsp | |
| 10:18 | jhutchins_lt2 has joined #ltsp | |
| 10:20 | vagrantc has quit IRC | |
| 10:31 | vagrantc has joined #ltsp | |
| 10:31 | ogra has joined #ltsp | |
| 10:34 | <vagrantc> hrm.
| |
| 10:34 | last time i tested the new LDM stuff it was completely broken ...
| |
| 10:34 | would've liked to have a little more notice before merging it into trunk...
| |
| 10:36 | oh well.
| |
| 10:37 | gnunux has quit IRC | |
| 10:37 | <vagrantc> don't get me wrong, the changes sound like something good ... it just would be nice to have an opportunity to test it
| |
| 10:38 | <Kyle__> Oh, vagrantc. Awhile ago someone mentioned a bug with debian running locale-gen, even when it shouldn't, and that was keeping it from building the client image. You worked on it, and they never repsponded, right?
| |
| 10:38 | alkisg has quit IRC | |
| 10:39 | <vagrantc> don't recall this at all
| |
| 10:40 | <Kyle__> I found a mailing list post where you responded, then they never respondend and you closed the issue, that's all I know.
| |
| 10:40 | <vagrantc> that's more than i know, at this point.
| |
| 10:41 | <Kyle__> _but_, I ran into the same issue yesterday. It doesn't happen on the 64bit builds, but does on the 32. I don't know if that's the varible or not, since the 32-bit server is being run off of openVZ container.
| |
| 10:41 | <vagrantc> a reference to the post you found would be helpful...
| |
| 10:41 | * Kyle__ nods | |
| 10:41 | <Kyle__> Holdon, I'll try and dig it up (was at the other campus).
| |
| 10:41 | <vagrantc> i almost exclusively use 32 bit for testing
| |
| 10:42 | <Kyle__> Humm. Then it sounds more likely an issue with the OpenVZ container.
| |
| 10:42 | Which is far-enough out there that I'm probably the only one who was nuts enough to try and run ltsp off of it.
| |
| 10:42 | alkisg has joined #ltsp | |
| 10:43 | <vagrantc> using some type of container doesn't sound like such a bad idea.
| |
| 10:43 | <Kyle__> That's the link I found. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=514877
| |
| 10:43 | vagrantc: It shouldn't be, and it works for 99% of the things I want to do off it it _flawlessly_.
| |
| 10:43 | vagrantc: But for some things, it's, well, funky. Like NFS.
| |
| 10:44 | <vagrantc> oh, that bug.
| |
| 10:44 | yeah, i was never able to reproduce it, and the submitter never responded...
| |
| 10:44 | * Kyle__ nods | |
| 10:45 | <Kyle__> alkisg: was around when I came in with it, he suggested I specify my locale, but it didn't help.
| |
| 10:45 | Silly tab completion.
| |
| 10:45 | <vagrantc> might be an issue with chroot and openvz?
| |
| 10:46 | though there are several other points where it chroots ...
| |
| 10:46 | * Kyle__ nods | |
| 10:48 | <Kyle__> Or it could be an issue with a slightly-broken ubuntu install. It was 8, and was updated through to 10. It didn't complain, but something could have gotten broken in that process.
| |
| 10:48 | <vagrantc> this isn't on debian?
| |
| 10:49 | * Kyle__ nods | |
| 10:50 | <Kyle__> It's a debian chroot hosted on ubuntu. Which works fine on the physical install of 64bit ubuntu 10.04
| |
| 10:50 | <vagrantc> then it's completely different code... shouldn't run locale-gen at all
| |
| 10:50 | * Kyle__ nods | |
| 10:50 | <Kyle__> vagrantc: I read the script and came across that very fact myself.
| |
| 10:50 | abeehc- has quit IRC | |
| 10:51 | <vagrantc> ah, wait ...
| |
| 10:51 | abeehc- has joined #ltsp | |
| 10:51 | <vagrantc> it does run it from the Ubuntu/025-locales plugin ...
| |
| 10:51 | Kyle__: in the debian version, i specify a full path ...
| |
| 10:52 | and test that it even exists
| |
| 10:52 | * Kyle__ blinks | |
| 10:52 | <Kyle__> Then I read the wrong script.
| |
| 10:53 | <vagrantc> check /usr/share/ltsp/plugins/ltsp-build-client/Ubuntu/025-locales
| |
| 10:53 | that's where it's failing
| |
| 10:54 | <Kyle__> vagrantc: Should it be using the debian script, because it's building a debian image, or the ubuntu script, because it's building on an ubuntu server?
| |
| 10:58 | <vagrantc> Kyle__: you're trying to build a debian image on an ubuntu server?
| |
| 10:59 | cross-distro install isn't supported.
| |
| 11:00 | dgroos has joined #ltsp | |
| 11:00 | <vagrantc> if you want to do that ... build a debian chroot, bind mount /opt/ltsp to $debian/opt/ltsp , run ltsp-build-client from within the chroot , and then run ltsp-update-kernels outside the chroot.
| |
| 11:02 | docente has quit IRC | |
| 11:09 | <alkisg> Kyle__: how are things without nbd-proxy?
| |
| 11:10 | <Kyle__> alkisg: Quite stable actually.
| |
| 11:10 | <alkisg> So it seems like it was indeed an nbd problem?
| |
| 11:10 | <Kyle__> alkisg: No flake outs so far. I'm going to make the change in my other lab and watch it.
| |
| 11:10 | <alkisg> OK, after testing do comment on that bug report...
| |
| 11:11 | <Kyle__> alkisg: We dunno. The latest kernel update from 'buntu helped a little (less frequent), but it still happened. If they stay up without the error for a few days, then I'll be compleetly willing to say that's the problem. Until then, well, we may just be lucky :)
| |
| 11:12 | <vagrantc> Kyle__: so i'm still a bit confused ... you're running an ubuntu server ... on openvz ? or the debian server is an openvz on your ubuntu server?
| |
| 11:14 | <Kyle__> vagrantc: Well, to be compleetly clear, the hardware is running a debian-based openvz server. The openvz container (vm) is an ubuntu server. I tried to build a debian client on the ubuntu server (which is a vm)
| |
| 11:14 | At this point acusations of being a sick puppy are excusable.
| |
| 11:22 | <vagrantc> Kyle__: that is a bit twisted, yes.
| |
| 11:22 | Kyle__: just build the Debian LTSP environment from a Debian LTSP server and everything will be less painful. :)
| |
| 11:22 | <Kyle__> vagrantc: I was already using that box for all the other serving purposes in the room, and the hardware I was using for the LTSP server needed to be returned to a workstation (they needed all the seats in the class).
| |
| 11:23 | <vagrantc> weather that's a debian openvz container or not...
| |
| 11:23 | <Kyle__> vagrantc: Heh, I'll see if I can. In the meantime I'm trying to get them to (gasp) give me a dedicated box to use as a server.
| |
| 11:23 | <vagrantc> crazy talk.
| |
| 11:24 | <Kyle__> I know. Us sysadmins are such premadonnas. "You mean you need more than ONE P4 to run a computer lab?"
| |
| 11:31 | scm_ has joined #ltsp | |
| 11:31 | scm_ is now known as scottmaccal | |
| 11:32 | <scottmaccal> Mmm, good lunch. How is everyone doing?
| |
| 11:45 | dgroos has left #ltsp | |
| 11:46 | scottmaccal has quit IRC | |
| 11:56 | <Mobe> about ltsp-cluster, can i use one server to run root, cluster-control, loadbalancer and app1 roles? and the second server as app2 for this cluster?
| |
| 12:02 | vagrantc has quit IRC | |
| 12:14 | dobber has quit IRC | |
| 12:23 | jhutchins_lt2 is now known as jhutchins_lt | |
| 12:57 | jhutchins_lt has quit IRC | |
| 13:02 | elisa871 has joined #ltsp | |
| 13:02 | <elisa871> hi
| |
| 13:02 | has anyone here used "Citrix"?
| |
| 13:02 | Lns has joined #ltsp | |
| 13:02 | Lns has joined #ltsp | |
| 13:05 | <Kyle__> I have, but not voluntarily. And then only as a user.
| |
| 13:05 | elisa87 has joined #ltsp | |
| 13:06 | elisa872 has joined #ltsp | |
| 13:07 | <Kyle__> elisa872: You'll get the same answer no matter what user you /join as.
| |
| 13:07 | <elisa872> has anyone here used Citrix??
| |
| 13:07 | elisa871 has quit IRC | |
| 13:07 | <elisa872> kyle_ i got disconnected
| |
| 13:07 | could you please answer again
| |
| 13:07 | I want to find a wait to include accounting to Citrix
| |
| 13:08 | <Kyle__> elisa872: I've only used it as a user.
| |
| 13:08 | elisa872: Say what you're trying to do again? I didn't quite understand what you said.
| |
| 13:08 | <elisa872> aha,did you have credit? I mean did your usage was in a timely manner?
| |
| 13:09 | I want to give customers credit
| |
| 13:09 | <Kyle__> I really don't understand.
| |
| 13:09 | <elisa872> I want to include accounting to Citrix, do you have any idea?
| |
| 13:09 | <Kyle__> But anyway, this channel is for LTSP, a linux project. Citrix is a windows product, unless a lot has changed since I last used it.
| |
| 13:10 | elisa87 has quit IRC | |
| 13:12 | <Mobe> there is #Citrix on this network to support that
| |
| 13:17 | * Kyle__ shrugs | |
| 13:48 | ogra has quit IRC | |
| 13:52 | <Kyle__> Hu. LDM is no longer showing all my sessions.
| |
| 13:52 | * Kyle__ tries to remember how he fixed it last time. | |
| 13:54 | vagrantc has joined #ltsp | |
| 13:55 | mgariepy has joined #ltsp | |
| 14:12 | jhutchins_lt has joined #ltsp | |
| 14:14 | <abeehc-> anyone know how to get localdev cdrom/usb and whatnot working with the rdesktop script
| |
| 14:14 | i'm failing
| |
| 14:24 | loather-work has joined #ltsp | |
| 14:28 | jhutchins_lt has quit IRC | |
| 14:30 | Kyle__ has quit IRC | |
| 14:32 | Kyle__ has joined #ltsp | |
| 14:37 | vagrantc has quit IRC | |
| 14:39 | vagrantc has joined #ltsp | |
| 14:40 | <alkisg> !compiz
| |
| 14:40 | <ltspbot> alkisg: "compiz" :: if compiz is giving you problems, one way to disable it for all users is: sudo gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.mandatory --type string --set /desktop/gnome/session/required_components/windowmanager metacity
| |
| 14:40 | <Kyle__> alkisg: Is that the answer to everything? ;)
| |
| 14:40 | pmatulis has quit IRC | |
| 14:44 | <alkisg> :)
| |
| 14:44 | It does work though :)
| |
| 14:44 | Now if only nfs could stay working.... :-/
| |
| 14:44 | <Kyle__> Heh.
| |
| 14:44 | alkisg: What's happening with your NFS?
| |
| 14:45 | <alkisg> It working. Then it stops working (looooong timeouts). After 1 minute, it's working again...
| |
| 14:45 | The weird thing is that I have similar problems with sshfs...
| |
| 14:46 | <Kyle__> alkisg: ubuntu for the NFS server?
| |
| 14:46 | <alkisg> Yes - it's working fine with vbox clients though
| |
| 14:46 | * Kyle__ had problems getting statd to stay running with ubuntu. | |
| 14:46 | <Kyle__> Oh. Humm.
| |
| 14:46 | <alkisg> It doesn't seem like a server problem
| |
| 14:47 | <Kyle__> Well, just double check that....
| |
| 14:47 | How many clients?
| |
| 14:53 | komunista has joined #ltsp | |
| 14:55 | <komunista> hi
| |
| 14:55 | vagrantc: today I updated chroot in debian testing with NBD image
| |
| 14:56 | this update invokes update-iniramfs, which generate new initrd with 600 permissions
| |
| 14:56 | <vagrantc> insisting on using NBD still, eh?
| |
| 14:57 | <komunista> and ltsp-update-image copy this initrd fith these premissions and finnaly clients can't boot - by initrd.img not found
| |
| 14:57 | <vagrantc> i've never seen that happen before, even when testing ltsp-update-image
| |
| 14:57 | <komunista> vagrantc: :-)
| |
| 14:58 | vagrantc: I dont't know where is problem (if update-initramfs or ltsp-upadate-image), but this is what I find today
| |
| 15:03 | vagrantc: I was using the NBD with Ubuntu more than year, when I swith to Debian now, I also use it...
| |
| 15:06 | vagrantc has quit IRC | |
| 15:07 | vagrantc has joined #ltsp | |
| 15:11 | Gadi has left #ltsp | |
| 15:17 | ogra_ac has quit IRC | |
| 15:20 | elisa872 has left #ltsp | |
| 15:21 | ogra_ac has joined #ltsp | |
| 15:34 | mikkel has quit IRC | |
| 15:35 | mgariepy has quit IRC | |
| 15:42 | <Kyle__> Anyone have an idea about the sessions list missing in fat clients?
| |
| 15:44 | <alkisg> Kyle__: didn't we put a call to a local ldminfod for that purpose?
| |
| 15:44 | Or that wasn't you? Anyway I committed the change upstream, just download the newer version of that file + install ldm-server in the chroot
| |
| 15:46 | http://bazaar.launchpad.net/~ltsp-upstream/ltsp/ldm-trunk/revision/1279
| |
| 15:46 | ogra_ac has quit IRC | |
| 15:47 | ogra_ac has joined #ltsp | |
| 15:48 | <Kyle__> alkisg: Yes, but I updated the systems, and now it's not working.
| |
| 15:49 | <alkisg> Upload your ldm script to pastebin to check if the change is indeed there
| |
| 15:54 | <Kyle__> http://pastebin.com/s3STFBK0
| |
| 15:54 | alkisg: Nothing changed since last time.
| |
| 15:54 | <alkisg> Kyle__: that script doesn't have the new code, add it
| |
| 15:54 | See the link I pasted above
| |
| 15:55 | <Kyle__> alkisg: Cool.
| |
| 16:01 | hersonls has quit IRC | |
| 16:01 | <Kyle__> alkisg: Errr, stupid question, how can I pull down that whole file? I'm so dense sometimes....
| |
| 16:02 | nevermind, I'm blind.
| |
| 16:02 | <alkisg> Click on " browse files at revision 1279"
| |
| 16:05 | * Kyle__ nods | |
| 16:05 | <Kyle__> I'm dense.,
| |
| 16:07 | alkisg: Other than a missing newline, and a missing comment, those look the same....
| |
| 16:08 | dobber_ has joined #ltsp | |
| 16:09 | <stgraber> ok, rdp support in ldm is fixed, now to improve the menus a bit, then I'll try to fix the session/language dialog tomorrow ;)
| |
| 16:09 | <vagrantc> stgraber: so little to no packaging changes with the new version?
| |
| 16:10 | stgraber: and this definitely sounds like ldm 3.0 to me ...
| |
| 16:10 | <alkisg> Kyle__: check the link I pasted, you're missing the lines that start with...
| |
| 16:10 | <stgraber> vagrantc: for Ubuntu, no packaging change at all. Only thing I'll need to fix (but that's an autotool issue) is to get wwm in /usr/bin/wwm
| |
| 16:11 | <alkisg> Kyle__: 94 # For this to work, ldm-server needs to be installed in the chroot
| |
| 16:11 | 95 if boolean_is_true "$LTSP_FATCLIENT" && [ -f /usr/sbin/ldminfod ]; then
| |
| 16:11 | 96 ldminfod > /var/run/ldm/$SRV
| |
| 16:11 | ...
| |
| 16:12 | <stgraber> vagrantc: yeah, I was hesitating between bumping to 2.2 or to 3.0. I'd like to wait for the new SSH backend and freerdp to be implemented before calling that 3.0 though.
| |
| 16:12 | cfkane has left #ltsp | |
| 16:12 | <vagrantc> ah
| |
| 16:12 | <Kyle__> alkisg: It has those in the one I was using & pastebinned.
| |
| 16:13 | <vagrantc> stgraber: i tried just s,rdesktop,freerdp,g earlier this year, and that worked.
| |
| 16:13 | stgraber: so it seems like a drop-in replacement, at least with the old screen script.
| |
| 16:13 | <alkisg> Kyle__: ah, I was searching for the comment, you don't have that but you have the rest...
| |
| 16:13 | * Kyle__ nods | |
| 16:14 | <Kyle__> I did put the new one in place, I just think it looks functionally the same.
| |
| 16:14 | <alkisg> Right. And you do have ldm-server installed in the chroot?
| |
| 16:14 | <stgraber> vagrantc: yep, though ideally I'd use the library (libfreerdp) instead of the command line (xfreerdp) so we can have a better integration with ltspfs and whatever sound backend we use (as in, have new network shares appear on the Windows side as we plug them)
| |
| 16:15 | vagrantc: that should also give us better error reporting. So far I've been playing with a tool call remmina which is some kind of universal remote desktop client (rdp/nx/vnc)
| |
| 16:15 | <vagrantc> stgraber: sounds very cool.
| |
| 16:15 | <stgraber> vagrantc: it has a nice plugin infrastructure and already uses libfreerdp
| |
| 16:15 | so we might be able to share some code there or at least look at how it's done
| |
| 16:16 | <Kyle__> alkisg: It was working, so I thought I must..but apparently I didn't...which is confusing.
| |
| 16:16 | <stgraber> vagrantc: ideally, I'd like to get rid of most code that's just doing some expect-like actions and replace that with clean libraries that we can share with other projects
| |
| 16:18 | <vagrantc> stgraber: sounds much better, yes.
| |
| 16:19 | might we hear from libpam-ssh2 this weekend?
| |
| 16:19 | ogra_ac has quit IRC | |
| 16:20 | <stgraber> vagrantc: I was discussing with Scotty on jabber earlier. He didn't have much time to work on libpam-ssh2 for a while but might have some time to look at it/discuss at the hackfest
| |
| 16:20 | <vagrantc> i hope so.
| |
| 16:21 | at some point, password expiry was working with ldm, so i enabled it at freegeek ... and then it stopped working with some version, but i hate to disable password expiry.
| |
| 16:21 | so we've just been dealing with the consequences.
| |
| 16:22 | <stgraber> yeah, having a clean library should help us deal with that properly. Also, the new plugin infrastructure in ldm should be easy to extend to have per-plugin hooks to deal with password expiry.
| |
| 16:22 | ogra_ac has joined #ltsp | |
| 16:22 | <stgraber> a few school districs I'm working with would like to just start a local firefox when a password expires so the user can change it online and retry
| |
| 16:23 | dgroos_ has joined #ltsp | |
| 16:23 | <stgraber> same thing for customers who need a "I forgot my password" option. They usually already have an online form to deal with that, we just need to provide an option that starts a web browser.
| |
| 16:26 | dgroos_ is now known as dgroos | |
| 16:31 | komunista has quit IRC | |
| 16:32 | <vagrantc> stgraber: does it have hooks at login time, too? i.e. i want to set up something that dynamically determines which server to log in to by default, rather than whenever ldm happened to boot.
| |
| 16:32 | or have a list of fallback servers ...
| |
| 16:34 | <Mobe> hi can i build ltsp-cluster on 2 servers, so that first has root, clustercontrol, loadbalancer and application1 roles and second is just application2?
| |
| 16:34 | <stgraber> vagrantc: I know that the RDP plugin doesn't, I still need to check for the SSH plugin though. My guess is that it's there for SSH but not yet for RDP.
| |
| 16:35 | * jammcq is looking forward to seeing this at bts. the brazilians that are coming have done alot with load balancing. | |
| 16:35 | <stgraber> vagrantc: Having that for SSH is critical for us and it'll very likely be for RDP so I'm probably not going to deploy that in large scale production until it's there.
| |
| 16:35 | Mobe: sure
| |
| 16:36 | vagrantc: currently we have an ltsp-cluster hook in the ssh scripts that re-update the LDM_SERVER if necessary. That's used so that we can safely remove servers from the cluster without having some clients to fail to login.
| |
| 16:37 | <alkisg> jammcq: what's the policy for ltsp* names? Could I name a tool of mine "ltsp-classmanager" or something similar without worrying about trademarks? :)
| |
| 16:38 | <jammcq> alkisg: I don't hold the trademark on ltsp. that's DisklessWorkstations that holds it. But, if you are writing something to be used for ltsp, i'd say go ahead and include ltsp in the name
| |
| 16:39 | <Mobe> ok, then i have enough hardware for one weekend setup
| |
| 16:39 | <alkisg> jammcq: ah, sorry, ty though :)
| |
| 16:39 | <jammcq> no prob
| |
| 16:39 | <stgraber> Mobe: are you installing that in OpenVZ or bare-metal ?
| |
| 16:39 | jammcq has quit IRC | |
| 16:40 | vagrantc has quit IRC | |
| 16:41 | <Mobe> bare metal, (or should I go VZ?)
| |
| 16:42 | <stgraber> if you have any experience with vz, I'd suggest it as you'd be able to have your different services isolated and both application servers to be absolutely identical (you could then just copy the VZ from one box to the other)
| |
| 16:44 | <Mobe> sounds good, i have 2 32bit intel desktops with 4G ram each and 1 older with 2G
| |
| 16:45 | i need this setup just for 1 weekend for 20+ terminals to run 24 hours
| |
| 16:46 | the older one is going to be routing ltsp and windows networks no adsl
| |
| 16:47 | i have no experience with vz but am familiar with concept of virtualservers
| |
| 16:49 | vagrantc has joined #ltsp | |
| 17:00 | [GuS] has joined #ltsp | |
| 17:00 | [GuS] has joined #ltsp | |
| 17:00 | isetann has quit IRC | |
| 17:00 | isetann_ has quit IRC | |
| 17:01 | artista_frustrad has quit IRC | |
| 17:03 | * alkisg never got NFS working with that problematic lab (statd running and everything else seemingly ok), but at least it's now working fine with sshfs... | |
| 17:10 | alkisg has quit IRC | |
| 17:12 | <Kyle__> Ack. OK, with ldm-server on my images, they don't even finish booting.
| |
| 17:12 | * Kyle__ sighs | |
| 17:12 | <Kyle__> Sometimes, fat clients are a PITA.
| |
| 17:33 | vagrantc has quit IRC | |
| 17:41 | <abeehc-> https://bugs.launchpad.net/ubuntu/+source/gnome-screensaver/+bug/530044
| |
| 17:43 | dgroos has quit IRC | |
| 17:45 | bobby_C has quit IRC | |
| 17:49 | johnny has left #ltsp | |
| 17:51 | ogra_ac has quit IRC | |
| 17:56 | ogra_ac has joined #ltsp | |
| 17:56 | <Kyle__> anyone awake who uses fat clients?
| |
| 18:02 | ogra_ac has quit IRC | |
| 18:03 | <Kyle__> Humm... What, on the server side, needs to be running for LDM?
| |
| 18:06 | * Kyle__ thinks that could be the whole problem | |
| 18:13 | johnny has joined #ltsp | |
| 18:14 | vagrantc has joined #ltsp | |
| 18:16 | [GuS] has quit IRC | |
| 18:41 | dobber_ has quit IRC | |
| 18:49 | staffencasa has quit IRC | |
| 18:54 | Lns has quit IRC | |
| 19:16 | pmatulis has joined #ltsp | |
| 19:18 | vagrantc has quit IRC | |
| 20:06 | pmatulis has quit IRC | |
| 21:18 | mistik1 has quit IRC | |
| 21:19 | mistik1 has joined #ltsp | |
| 22:00 | johnny has left #ltsp | |
| 22:04 | alexqwesa__ has quit IRC | |
| 22:04 | sutula has quit IRC | |
| 22:04 | sutula has joined #ltsp | |
| 22:14 | alex789 has joined #ltsp | |
| 22:35 | johnny has joined #ltsp | |
| 22:39 | Faithful has joined #ltsp | |
| 22:44 | Faithful has joined #ltsp | |
| 22:50 | ogra_ac has joined #ltsp | |
| 23:07 | MorningSon has quit IRC | |
| 23:20 | jhutchins_lt has joined #ltsp | |
| 23:23 | jhutchins_lt2 has joined #ltsp | |
| 23:23 | jhutchins_lt has quit IRC | |
| 23:30 | Faithful has quit IRC | |
| 23:36 | jhutchins_lt2 has quit IRC | |
| 23:44 | Faithful has joined #ltsp | |
| 23:57 | vagrantc has joined #ltsp | |