IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 30 July 2010 (all times are UTC)
| 00:08 | ball has joined #ltsp | |
| 00:25 | daduke has quit IRC | |
| 00:29 | daduke has joined #ltsp | |
| 00:41 | ball is now known as ball_ | |
| 00:42 | ball_ has quit IRC | |
| 01:21 | Patina has joined #ltsp | |
| 01:47 | alkisg has joined #ltsp | |
| 02:03 | gnunux has joined #ltsp | |
| 02:13 | joulez has quit IRC | |
| 02:59 | otavio has joined #ltsp | |
| 02:59 | otavio has joined #ltsp | |
| 03:13 | chupacabra has quit IRC | |
| 03:17 | ogra_cmpc has quit IRC | |
| 03:29 | ogra_cmpc has joined #ltsp | |
| 03:47 | johnny has left #ltsp | |
| 03:48 | otavio has quit IRC | |
| 03:55 | otavio has joined #ltsp | |
| 04:01 | rjune_ has quit IRC | |
| 04:01 | johnny has joined #ltsp | |
| 04:10 | otavio has quit IRC | |
| 04:21 | komunista has joined #ltsp | |
| 04:22 | Selveste1 has joined #ltsp | |
| 04:22 | otavio has joined #ltsp | |
| 04:25 | alkisg has quit IRC | |
| 04:32 | alkisg has joined #ltsp | |
| 04:37 | sweetpi has quit IRC | |
| 04:52 | dobber_ has joined #ltsp | |
| 05:14 | otavio has quit IRC | |
| 05:18 | satellit_ has joined #ltsp | |
| 06:02 | wurzzero has joined #ltsp | |
| 06:02 | ogra_cmpc has quit IRC | |
| 06:03 | <wurzzero> Good morning!
| |
| 06:04 | ogra_cmpc has joined #ltsp | |
| 06:29 | pmatulis has joined #ltsp | |
| 06:45 | mikkel has joined #ltsp | |
| 06:49 | rjune_ has joined #ltsp | |
| 06:53 | rad4Christ has joined #ltsp | |
| 06:57 | jammcq has quit IRC | |
| 07:06 | rad4Christ has quit IRC | |
| 07:55 | rjune_ has quit IRC | |
| 07:55 | alkisg has quit IRC | |
| 08:06 | alkisg has joined #ltsp | |
| 08:09 | vagrantc has joined #ltsp | |
| 08:16 | rjune_ has joined #ltsp | |
| 08:50 | alkisg has quit IRC | |
| 09:14 | alkisg has joined #ltsp | |
| 09:41 | alkisg has quit IRC | |
| 09:54 | Gadi has joined #ltsp | |
| 09:57 | wurzzero has quit IRC | |
| 10:00 | wurzzero has joined #ltsp | |
| 10:01 | gnunux has quit IRC | |
| 10:11 | staffencasa has joined #ltsp | |
| 10:13 | <vagrantc> Gadi: aha!
| |
| 10:14 | Gadi: any comments on http://bugs.debian.org/587900 ?
| |
| 10:19 | staffencasa has quit IRC | |
| 10:23 | staffencasa has joined #ltsp | |
| 10:36 | dobber_ has quit IRC | |
| 10:36 | <vagrantc> stgraber, Gadi: plan to tag ltsp-trunk today ...
| |
| 10:37 | any pending commits, get em in soonish :)
| |
| 11:05 | <knipwim> vagrantc: i have a gentoo commit i'd like to include
| |
| 11:05 | in how many hours are you going to tag?
| |
| 11:09 | <vagrantc> knipwim: was hoping to soon ... but does gentoo even use the released version? :)
| |
| 11:13 | * vagrantc is pretty clueless about gentoo processes | |
| 11:14 | <vagrantc> knipwim: how's the gentoo port going?
| |
| 11:14 | alkisg has joined #ltsp | |
| 11:17 | <vagrantc> might be later this evening, anyways...
| |
| 11:18 | alkisg: plan to tag ltsp-trunk today ... any issues you'd like committed/
| |
| 11:18 | ?
| |
| 11:19 | alkisg: apparently, the sleep included with debian's initramfs doesn't support partial seconds ...
| |
| 11:19 | alkisg: you've got some hooks in the udhcpc script that call "sleep 0.1"
| |
| 11:19 | i guess we could include /usr/bin/sleep and hard-code a path to sleep...
| |
| 11:19 | <alkisg> vagrantc: I thought I saw a debian bug about it, *closed fixed*
| |
| 11:20 | vagrantc: stgraber put those, I just left them there
| |
| 11:20 | <vagrantc> hm.
| |
| 11:20 | <alkisg> It's not in the right place anyway
| |
| 11:20 | <vagrantc> we've currently worked around it by changing it to sleep 1
| |
| 11:20 | <alkisg> It should be in local-top I think, not in init-premount
| |
| 11:21 | (man initramfs-tools describes the stages, I just haven't read that for a long time)
| |
| 11:21 | <vagrantc> local-top doesn't get called from ltsp_nbd or nfs
| |
| 11:21 | <alkisg> stgraber put it in init-premount exactly to avoid changing ltsp_nbd
| |
| 11:21 | I think we should change ltsp_nbd and call it, if udhcpc is installed
| |
| 11:21 | Now it bumps into a lot of initialization related problems
| |
| 11:22 | <vagrantc> well, it's nice for it to work for nfs scripts too, though
| |
| 11:22 | <alkisg> E.g. eth not initialized yet etc
| |
| 11:22 | OK, but still, we could put a symlink to it in nfs-top or something like that
| |
| 11:22 | Right now it's called too early
| |
| 11:23 | * alkisg reads at the initramfs-tools manpage.. | |
| 11:24 | <alkisg> local/nfs-top or local/nfs-premount, not sure which one should be more appropriate
| |
| 11:24 | vagrantc: did you commit Peter's change?
| |
| 11:25 | <vagrantc> alkisg: the sleep change i didn't commit upstream
| |
| 11:25 | but the other stuff i did
| |
| 11:26 | <alkisg> Nice. Busybox in sid still doesn't support decimals?
| |
| 11:26 | (or however they're called...)
| |
| 11:28 | <vagrantc> nope!
| |
| 11:28 | <alkisg> vagrantc: urm. It looks like it doesn't support it here either, so the 0.1 => 1 change probably needs to go upstream...
| |
| 11:28 | <vagrantc> suppose could file a wishlist bug
| |
| 11:29 | ah, so much the better.
| |
| 11:29 | <alkisg> Nah, the "sleeps" there are bad anyway
| |
| 11:29 | <vagrantc> well, simpler
| |
| 11:29 | <alkisg> I think if we move it in a later init stage, it won't be needed at all
| |
| 11:30 | <vagrantc> i don't know that you can move it much later and have it available generally
| |
| 11:30 | <alkisg> I think it's doable, as long as it's alphabetically before nfs-top etc... but I'm not sure
| |
| 11:31 | <vagrantc> it's not a sort-order issue ... some directories aren't called from the nfs scripts, for example.
| |
| 11:32 | <alkisg> vagrantc: ah, I just tried it in a client initramfs, it works fine
| |
| 11:32 | <vagrantc> where did you put it?
| |
| 11:32 | the sleep 0.1 ?
| |
| 11:32 | <alkisg> I don't know what ubuntu does to it's initramfs - `busybox sleep 0.1` doesn't work, while `sleep 0.1` in break=mount works
| |
| 11:32 | <vagrantc> maybe it's using /bin/sleep ?
| |
| 11:32 | <alkisg> *I mean busybox sleep outside of the chroot, on the server,
| |
| 11:33 | while busybox sleep 0.1 on break=mount works fine
| |
| 11:33 | <vagrantc> oh
| |
| 11:33 | that's... ummm... confusing.
| |
| 11:33 | <alkisg> I believe I saw a fixed bug in busybox about it, so not sure what's the problem - maybe the compile flags?
| |
| 11:33 | There's another version in the initramfs and another in a normal ubuntu system
| |
| 11:33 | <vagrantc> ah.
| |
| 11:34 | <alkisg> (busybox-static, busybox-initramfs etc etc not sure about the details)
| |
| 11:34 | <vagrantc> busybox-initramfs
| |
| 11:34 | <alkisg> The problem with the current place for udhcpc is that it's broken for quite a lot of newer systems
| |
| 11:35 | <vagrantc> still use klibc/ipconfig in debian
| |
| 11:36 | though the udhcpc hooks do kick in
| |
| 11:36 | <alkisg> In many core 2 duo clients nbd can't connect because the card is not properly initialized (I assume because it's too early - if I put a break=mount, wait 2 secs, and then press ctrl+d, it boots)
| |
| 11:36 | <vagrantc> well, if we can figure where to put it that works better for both debian and ubuntu, we're set :)
| |
| 11:36 | * alkisg is using IPAPPEND 3 because stgraber put a clientid there which doesn't work for cisco routers :-/ | |
| 11:37 | <alkisg> (to prevent duplicate leases in dhcpd)
| |
| 11:37 | <vagrantc> looks like i won't manage to upload for a few hours anyways
| |
| 11:37 | <alkisg> I won't have time to work on this neither today nor in the next few days, so don't hold back on my account...
| |
| 11:38 | <vagrantc> ah, sure.
| |
| 11:38 | worse case is i leave it where it is with the sleep 0.1 patch
| |
| 11:38 | <alkisg> It will need a lot of testing anyway, that's why I haven't even started it yet
| |
| 11:38 | Sounds good to me
| |
| 11:38 | <vagrantc> could just change it to "true"
| |
| 11:38 | which will sort of act like a very brief sleep
| |
| 11:38 | <alkisg> It won't leave the cpu free for other upstart scripts
| |
| 11:39 | sleep 1 would be better, I think
| |
| 11:39 | (udev etc)
| |
| 11:40 | vagrantc: ah, could you explain 1764. ltsp-chroot: fix argument processing for --arch and --base. ?
| |
| 11:40 | wurzzero has quit IRC | |
| 11:40 | chupacabra has joined #ltsp | |
| 11:40 | <Gadi> is the busybox in the initramfs compiled against glibc? (I wonder if that is why there is always that warning on boot ...)
| |
| 11:41 | <alkisg> Gadi, no idea, but sleep 0.1 works in the ubuntu initramfs busybox...
| |
| 11:41 | <vagrantc> alkisg: it would fail with an error message somewhat like: unknown directory /opt/ltsp/'armel'
| |
| 11:41 | alkisg: so i had to trim off the '
| |
| 11:42 | hopefully people aren't crazy enough to include ' in their directory names...
| |
| 11:42 | <alkisg> vagrantc: hmmm ok but that makes it impossible to have a /opt/ltsp/vagrantc's-chroot :P
| |
| 11:42 | Heh
| |
| 11:42 | OK ok
| |
| 11:43 | Shouldn't 1765 be applied to ltsp-update-image as well?
| |
| 11:43 | <vagrantc> alkisg: maybe so
| |
| 11:43 | alkisg: i don't use it much ... soo ...
| |
| 11:44 | <alkisg> ...heh, well, since you already started committing stuff I thought I'd ask that too :D
| |
| 11:44 | OK should I do it?
| |
| 11:45 | <vagrantc> alkisg: sure, go for it
| |
| 11:45 | <alkisg> ok, in 10'.
| |
| 11:47 | <vagrantc> alkisg: so you're ltsp-chroot didn't have the ' when you used --arch ?
| |
| 11:48 | i guess i could make the sed more particular ... sed -e "s,/',/,g" -e "s,'$,,g"
| |
| 11:56 | alexqwesa_ has joined #ltsp | |
| 11:57 | komunista has quit IRC | |
| 11:57 | alexqwesa has quit IRC | |
| 12:11 | komunista has joined #ltsp | |
| 12:13 | <alkisg> vagrantc: I copied the --arch and --base code from ltsp-update-image without testing it
| |
| 12:13 | So any changes there would better go to both of those files..
| |
| 12:13 | <vagrantc> ugh... we should put them somewhere that they both can use...
| |
| 12:14 | <alkisg> Yeah a little refactoring would be nice
| |
| 12:14 | Although not all the parameters are in common
| |
| 12:16 | * vagrantc heads to lunch | |
| 12:16 | [GuS] has joined #ltsp | |
| 12:17 | satellit_ is now known as satellit_afk | |
| 12:17 | satellit_afk has left #ltsp | |
| 12:18 | <[GuS]> Hi guys, i have a question: in the company i work still uses ltsp4.2 (gentoo). Now we have customers that wants to use a usb scanner connected to a thin client. Which package from here i need to so so? http://ltsp.mirrors.tds.net/pub/ltsp/ltsp-4.2/ or from somewhere else?
| |
| 12:19 | * vagrantc sings a sad, sad song about ltsp 4.x | |
| 12:19 | * vagrantc really heads to lunch | |
| 12:19 | vagrantc has quit IRC | |
| 12:40 | otavio has joined #ltsp | |
| 12:43 | <[GuS]> lol
| |
| 12:54 | * alkisg thinks LTSP 4.2 was abandoned somewhere in the Mesozoic period... | |
| 12:58 | <[GuS]> alkisg: lol
| |
| 12:58 | damn :(
| |
| 12:58 | no help then :(
| |
| 12:58 | ltsp5 is far away to be stable in gentoo...
| |
| 12:58 | <alkisg> Yeah ltsp 4.2 help is scarce at best
| |
| 12:58 | <[GuS]> alkisg: indeed.
| |
| 12:59 | <Gadi> [GuS]: if it is the latest 4.2, I believe there was an lts.conf param like XINETD_SERVICES = "saned" or some such
| |
| 12:59 | <johnny> [GuS], the problem here is that we have no good solution for initramfs
| |
| 12:59 | actually ltsp on gentoo works fine except for that part
| |
| 12:59 | now that dracut for gentoo should be complete by end of summer
| |
| 12:59 | we should have a one click install
| |
| 13:00 | stable is almost meaningless in gentoo anyways..
| |
| 13:00 | for ltsp anyways
| |
| 13:00 | because we curate the package list
| |
| 13:00 | <Gadi> [GuS]: see: http://www.mail-archive.com/ltsp-discuss@lists.sourceforge.net/msg26734.html
| |
| 13:00 | <[GuS]> johnny: thanks for the info
| |
| 13:01 | <johnny> atm the annoying part of ltsp gentoo is having to manually set the network cards
| |
| 13:01 | because genkernel is no good
| |
| 13:01 | but as i mentioned, dracut will take care of that
| |
| 13:01 | <[GuS]> Gadi: ohh great, i was searching a lot but didnt found that
| |
| 13:01 | johnny: well, i had troubles even generating the client
| |
| 13:02 | which is posted in the bug report
| |
| 13:02 | <johnny> ah.. i'll have knipwim look at that
| |
| 13:02 | we were using funtoo stages because they were more up to date
| |
| 13:02 | but i think we're going to switch back
| |
| 13:02 | <[GuS]> which, i believe was cause of my locale
| |
| 13:02 | <johnny> as hey changed the tar compression format
| |
| 13:02 | <[GuS]> yep
| |
| 13:02 | <johnny> ah.. a multilang person to test with ;)
| |
| 13:02 | <[GuS]> i've read that on the bug report
| |
| 13:02 | johnny: haha
| |
| 13:02 | <johnny> well i'd appreciate if you could idle here
| |
| 13:02 | <[GuS]> yeah.. from Argentina :P
| |
| 13:03 | <johnny> so you can talk to knipwim
| |
| 13:03 | <[GuS]> ok, i will, since we are fighting in the company witgh the old ltsp
| |
| 13:03 | and we want to upgrade
| |
| 13:03 | <johnny> i wrote the initial code, but don't have a powerful enough laptop to actually keep using it (still!!!)
| |
| 13:03 | so i make knipwim do the heavy lifting :) hehe
| |
| 13:04 | so yes, we'll hopefully get you fixed up soon then
| |
| 13:04 | i'm wondering if we should setup a sourceforge project..
| |
| 13:04 | i wonder how big of files they let you post, if iso size, then we could just allow people to download a working stage
| |
| 13:05 | i think they let you do that..
| |
| 13:05 | <[GuS]> :)
| |
| 13:05 | well, i will try these days to back on the test and see if i can make some progress installing it
| |
| 13:05 | <johnny> we'll help you get it together
| |
| 13:05 | <[GuS]> thanks :)
| |
| 13:06 | <johnny> i think sourceforge might be the way to go for the source snaps and the stage
| |
| 13:06 | <[GuS]> this was my start http://bugs.gentoo.org/show_bug.cgi?id=177580#c95
| |
| 13:06 | well, in fact http://bugs.gentoo.org/show_bug.cgi?id=177580#c91
| |
| 13:07 | so, the other guy from the same company found that supposed is a locale thing: http://bugs.gentoo.org/show_bug.cgi?id=177580#c98
| |
| 13:08 | so we are stuck in there
| |
| 13:08 | <knipwim> i'm looking at the bug now
| |
| 13:08 | <[GuS]> =)
| |
| 13:08 | wurzzero has joined #ltsp | |
| 13:09 | <alkisg> johnny: http://sourceforge.net/projects/knoppel/files/ ==> 725 MB
| |
| 13:10 | (greek knoppix)
| |
| 13:11 | <knipwim> johnny: what problem is solved by maintaining other stage files?
| |
| 13:12 | ah
| |
| 13:12 | you mean working client installs
| |
| 13:25 | <wurzzero> Hi..
| |
| 13:25 | I'm having problems to configure my ltsp server and i did some confgiurations to work, so decided to explain the hole situation to see if anyone has another to do this confs...
| |
| 13:25 | komunista has quit IRC | |
| 13:26 | <wurzzero> i paste the problems in http://pastebin.com/Fdeau51Q
| |
| 13:26 | i thought it's better to read...
| |
| 13:27 | but i can paste here to...
| |
| 13:30 | alkisg has quit IRC | |
| 13:48 | vagrantc has joined #ltsp | |
| 13:55 | <wurzzero> Hi, vagrantc... i'm having problems configuring ssh with another port and yesterday you told me to use SSH_OVERRIDE_PORT, but in my tests it only works with 'SCREEN_XX=ldm'
| |
| 13:56 | i paste a resume of the problem here http://pastebin.com/Fdeau51Q
| |
| 13:58 | <vagrantc> i'm leaving in just a second
| |
| 13:59 | <wurzzero> all right...
| |
| 14:01 | <[GuS]> ok, finally i've talk to my partnert at work and after many modifications we reach to the ldm login at the client, where the characters are being displayed as squares. This could be a locale problem?
| |
| 14:03 | <otavio> vagrantc:
| |
| 14:06 | vagrantc has quit IRC | |
| 14:11 | <johnny> [GuS], try with plain..
| |
| 14:11 | then you'll know
| |
| 14:11 | <[GuS]> johnny: what do you mean by plain?
| |
| 14:11 | <johnny> a plain locale
| |
| 14:11 | like C
| |
| 14:11 | <[GuS]> ahh ok
| |
| 14:11 | <johnny> just make it work first
| |
| 14:12 | then try making it work with your locale
| |
| 14:16 | <wurzzero> i'm using Ubuntu 10.04 i386 and set the autologin for ldm, but when i want to exit the client get in a loop and don't turn off.
| |
| 14:17 | anyone know how to fix this?
| |
| 14:27 | <knipwim> [GuS]: i found that when you get squares, you're missing fonts in the client environment
| |
| 14:28 | <[GuS]> knipwim: ahh ok
| |
| 14:28 | nice tip
| |
| 14:29 | <knipwim> i installed dejavu on the client
| |
| 14:29 | <[GuS]> knipwim: thanks, i will prepare my laptop to make tests as client
| |
| 14:29 | <knipwim> i found a solution to your language problem as well
| |
| 14:31 | already mailed the fix to the quickstart maintainer
| |
| 14:31 | i'll also post it in the gentoo bug
| |
| 14:32 | Anon9930 has joined #ltsp | |
| 14:33 | Anon6378 has joined #ltsp | |
| 14:39 | Anon6378 has quit IRC | |
| 14:49 | irule has joined #ltsp | |
| 14:50 | <knipwim> johnny: since fonts are not installed by default in the client
| |
| 14:50 | pmatulis has quit IRC | |
| 14:50 | <knipwim> i'm proposing to install them via the PACKAGES var
| |
| 14:50 | in ltsp-build-client.conf
| |
| 14:51 | giving the user a choice for which font to use
| |
| 14:51 | <[GuS]> <knipwim> already mailed the fix to the quickstart maintaine ---> i saw the bug report, great.
| |
| 14:52 | dobber_ has joined #ltsp | |
| 14:52 | <knipwim> johnny: already tested, either dejavu, font-bitstream-type1 or freefonts work
| |
| 14:54 | wurzzero has quit IRC | |
| 15:08 | GodFather has joined #ltsp | |
| 15:15 | alkisg has joined #ltsp | |
| 15:30 | <johnny> ah,so we should install a font :)
| |
| 15:32 | <knipwim> are you okay with the ltsp-build-client.conf solution?
| |
| 15:32 | Gadi has left #ltsp | |
| 15:36 | <johnny> hmm.. no, there needs to be a default
| |
| 15:36 | ltsp-build-client.conf shouldn't be required
| |
| 15:36 | unless we're installing those conf files..
| |
| 15:37 | but ubuntu doesn't
| |
| 15:37 | i don't think debian does either
| |
| 15:37 | knipwim, also.. hopefully in the next few weeks we can use dracut
| |
| 15:37 | since gsoc is almost finished
| |
| 15:38 | perhaps it's already been completed, i'm not sure
| |
| 15:38 | <knipwim> johnny: not according the developer's blog
| |
| 15:38 | <johnny> knipwim, really tho, we should be putting it in a profile file probably
| |
| 15:38 | <knipwim> which font?
| |
| 15:38 | <johnny> it doesn't seem like it should be that hard to accomplish, i'm not sure why it is taking them so long
| |
| 15:39 | <knipwim> i opt for dejavu
| |
| 15:39 | <johnny> dejavu seems reasonable
| |
| 15:39 | yes
| |
| 15:40 | really tho.. we should probably switch away from quickstart, if agaffney isn't going to maintain it
| |
| 15:40 | perhaps catalyst can do the job
| |
| 15:40 | <knipwim> i don't know catalyst
| |
| 15:43 | hmm, looks interesting
| |
| 15:52 | <[GuS]> is a replacement for quickstart?
| |
| 15:52 | GodFather has quit IRC | |
| 15:56 | <johnny> not quite
| |
| 15:56 | it's the gentoo iso build system
| |
| 15:56 | quickstart is simpler
| |
| 15:58 | <[GuS]> oks
| |
| 15:59 | my experience with Gentoo is almost new (since i work here). I am more debian based.
| |
| 16:00 | dobber_ has quit IRC | |
| 16:00 | mikkel has quit IRC | |
| 16:04 | dobber_ has joined #ltsp | |
| 16:04 | dobber_ has quit IRC | |
| 16:10 | dobber_ has joined #ltsp | |
| 16:16 | [GuS] has quit IRC | |
| 16:23 | staffencasa has quit IRC | |
| 16:23 | otavio has quit IRC | |
| 16:26 | shamino has quit IRC | |
| 16:30 | shawnp0wers has joined #ltsp | |
| 16:33 | shamino has joined #ltsp | |
| 16:45 | pmatulis has joined #ltsp | |
| 17:18 | Kyle__ has joined #ltsp | |
| 17:19 | <Kyle__> Hello. I've got an LDM question. Can it be unbundled from LTSP and used on it's own? Say in an ubuntu lab?
| |
| 17:19 | Or is it too tightly coupled into LTSP?
| |
| 17:21 | <alkisg> It's quite difficult, it's tightly coupled. Why would you want ldm in a non-ltsp client?
| |
| 17:21 | <johnny> why would you want to?
| |
| 17:21 | <Kyle__> If I understand it right, LDM can be used for authentication.
| |
| 17:22 | That's why. Simple authentication, without the need for ldap.
| |
| 17:22 | <alkisg> It's much easier to use ldap for that. But you can also use fat clients instead
| |
| 17:22 | dobber_ has quit IRC | |
| 17:22 | <johnny> Kyle__, it's just a frontend to ssh
| |
| 17:22 | so you'd still have to use ldap
| |
| 17:22 | <Kyle__> Oh. Damn.
| |
| 17:22 | <alkisg> johnny: not realy
| |
| 17:22 | *really
| |
| 17:22 | * Kyle__ is listening. | |
| 17:23 | <johnny> i mean.. if you need distributed authentication..
| |
| 17:23 | <alkisg> ldm copies the authenticated user to the client passwd, so it would work without ldap
| |
| 17:23 | <johnny> i guess i'm still missing what authentication you need
| |
| 17:23 | <alkisg> But, it's much much easier to maintain a fat client lab than a plain ubuntu lab
| |
| 17:23 | <johnny> in which gdm doesn't provide
| |
| 17:24 | <alkisg> Kyle__: sbalneav is planning on implementing a libpam-ssh authentication library, which some next version of ldm would use
| |
| 17:24 | <Kyle__> johnny: GDM works with whatever the underlying OS works with, but only so far. If you use it with LDAP, it can be a bit of pain.
| |
| 17:25 | <alkisg> That library could then be used from gdm or other managers, so that users would authenticate with ssh on your non-ltsp server.
| |
| 17:25 | <Kyle__> alkisg: Neat.
| |
| 17:25 | <johnny> your question was too vague
| |
| 17:25 | <alkisg> But really, fat client labs are much easier :D
| |
| 17:25 | <johnny> about what problem you're trying to solve
| |
| 17:26 | <Kyle__> alkisg: With the ubuntu flavored fat-client, how different is it than stock 'buntu?
| |
| 17:26 | <alkisg> Not at all. NFS is used for /home, ssh authentication though ldm, but everything runs locally
| |
| 17:26 | So you just maintain a chroot instead of e.g. 12 separate ubuntu installations
| |
| 17:27 | On a very new ubuntu lab the clients started in 50 secs with 9.04
| |
| 17:27 | I've installed 10.04 fat clients on that lab server, and they now start in 13 secs
| |
| 17:27 | <Kyle__> johnny: unified login and groups, so I don't have to create all the maps and hacks to get every group pemission right through LDAP.
| |
| 17:28 | johnny: Dozens of packages need special groups, all of which would have to be mapped, or hacked a bit, to get working with ldap.
| |
| 17:28 | <johnny> you can implement whatever pam method you want ... ldap or not
| |
| 17:28 | <Kyle__> alkisg: It can use any network FS for /home though, correct?
| |
| 17:28 | <alkisg> Kyle__: sure, you may even use local /home if you want to
| |
| 17:29 | No need for a network fs
| |
| 17:29 | <Kyle__> alkisg: I actually have need for gluster in this case, that's why I'm checking.
| |
| 17:29 | <alkisg> You'd need a small script to mount /home on boot
| |
| 17:29 | The nfs one is a couple of lines, so it should be pretty easy
| |
| 17:29 | <Kyle__> Not a problem.
| |
| 17:30 | <alkisg> What's your experience with gluster?
| |
| 17:30 | <Kyle__> alkisg: Limited, but I put it through some decent testing for my needs, and it seems stable.
| |
| 17:31 | * alkisg reads the web site... | |
| 17:32 | <Kyle__> My problem (one of them) is that I'll need a few terabytes of space for students to use, but haven't got the budget to get enough fast-enough drives for the server.
| |
| 17:32 | <alkisg> If you have normal hard drives, you could use those. Backup will be a problem then though - you'd need to solve that somehow
| |
| 17:33 | (I mean, mount /home locally)
| |
| 17:34 | Maybe rsync'ing the local homes to the server at nights or something like that...
| |
| 17:35 | <Kyle__> alkisg: I'm going to be creating a gluster-cluster where group A creates an appended volume, group B mirrors it, and one huge honkin drive on the server mirrors it.
| |
| 17:35 | <alkisg> Sounds neat, please post your experiences somewhere when you're done setting it up...
| |
| 17:35 | <Kyle__> alkisg: I will :)
| |
| 17:35 | OK, I'm digging through the LTSP wiki... am I looking for the term fat client?
| |
| 17:35 | Cuase I dont' see it...
| |
| 17:36 | <alkisg> https://help.ubuntu.com/community/UbuntuLTSP/FatClients
| |
| 17:36 | !fatclients
| |
| 17:36 | <ltspbot`> alkisg: Error: "fatclients" is not a valid command.
| |
| 17:36 | <Kyle__> Ahh, I was on the completely wrong wiki.
| |
| 17:36 | <alkisg> !learn fatclients as You may find some info about the Ubuntu/LTSP implementation of fat clients at https://help.ubuntu.com/community/UbuntuLTSP/FatClients
| |
| 17:36 | <ltspbot`> alkisg: The operation succeeded.
| |
| 17:38 | <alkisg> Kyle__: I've no idea how gluster works, but if you think it could also be used to export the chroot instead of nbd/nfs, please provide some feedback on that too.
| |
| 17:39 | <Kyle__> I will.
| |
| 17:39 | * alkisg would love the chroot being automatically mirrored locally on the client hard disks... | |
| 17:39 | <alkisg> It would make it possible to run a very large number of fat clients even on 100mbps networks
| |
| 17:39 | <Kyle__> This will be a very odd setup :) But I'm looking forward to it.
| |
| 17:40 | K, printed that out. I'm gonna go read that on the train.
| |
| 17:41 | <alkisg> k, good luck
| |
| 17:41 | Kyle__ has left #ltsp | |
| 17:59 | irule has quit IRC | |
| 18:28 | shawnp0wers has quit IRC | |
| 18:30 | Selveste1 has quit IRC | |
| 18:46 | shawnp0wers has joined #ltsp | |
| 19:05 | shawnp0wers has quit IRC | |
| 19:19 | alkisg has quit IRC | |
| 19:22 | highvoltage has quit IRC | |
| 20:01 | highvoltage has joined #ltsp | |
| 20:03 | highvoltage has quit IRC | |
| 20:06 | highvoltage has joined #ltsp | |
| 20:26 | pmatulis has quit IRC | |
| 20:32 | otavio has joined #ltsp | |
| 20:56 | wurzzero has joined #ltsp | |
| 21:12 | intelliant has joined #ltsp | |
| 21:13 | <intelliant> hello Appiah
| |
| 21:13 | sorry could not get back on the ldap client config issue
| |
| 21:13 | have made some progress since then
| |
| 21:13 | but it seems pam auth fails
| |
| 21:14 | you had asked for the auth.log output
| |
| 21:14 | the symptom now is that it allows a ldap user to log in and then throws him out within a few seconds
| |
| 21:21 | vagrantc has joined #ltsp | |
| 21:37 | RiXtEr has quit IRC | |
| 21:37 | RiXtEr has joined #ltsp | |
| 21:40 | <intelliant> now the problem is that without a root user that a default ubuntu install disallows, I am not even able to ssh back into the vm
| |
| 21:40 | so can't provide you with the logs :(
| |
| 21:41 | will have to restart the work on a cloned vm and will then post the logs before I close the session
| |
| 21:56 | intelliant has quit IRC | |
| 22:09 | otavio has quit IRC | |
| 22:16 | <vagrantc> bah. i'm too tired to do an ltsp upload ... even though it's probably a pretty trivial upload...
| |
| 22:52 | vagrantc has quit IRC | |
| 23:09 | try2free has joined #ltsp | |
| 23:11 | sweetpi has joined #ltsp | |
| 23:20 | try2free has quit IRC | |
| 23:21 | wurzzero has quit IRC | |
| 23:42 | sutula has quit IRC | |
| 23:42 | jhutchins has quit IRC | |
| 23:42 | hiltunen has quit IRC | |
| 23:42 | bix0r has quit IRC | |
| 23:43 | jhutchins has joined #ltsp | |
| 23:44 | hiltunen has joined #ltsp | |
| 23:47 | sutula has joined #ltsp | |
| 23:47 | bix0r has joined #ltsp | |