IRC chat logs for #ltsp on irc.freenode.net (webchat)


Channel log from 5 July 2015   (all times are UTC)

00:00
<vagrantc>
so that's one limitation of dnsmasq
00:00
<TheProf>
vagrantc: Thank you -- I don't have any machines scheduled to turn on automatically. I imagine in the computer lab the 30 computers will be turned on by the students and there will definitely delays between them.
00:01
<vagrantc>
dnsmasq can be configured to do all sorts of complex things, and it's arguably as easy if not easier than isc-dhcp-server, for example.
00:01
<maldridge>
huh, wonder if that's what was killing one of my labs, I was waking up 150 machines with wol magic packet
00:01
I would say its definitely easier, but for sheer power and configurability, it is lacking
00:02
certainly lacking in the direction of HA
00:02
<vagrantc>
ah, never tried HA DHCP servers.
00:02
dunno what other configurability or "sheer power" you might be referring to, though.
00:03
<maldridge>
sheer power as in being able to support a few thousand boxes with nice clean configs
00:03
my preferred stack is isc-dhcpd, nsd, unbound, and pf
00:04
with that I can scale horizontally as much as I want, and its all nice and modular
00:04
<TheProf>
And here I am getting stressed out over 30 computers :)
00:05
<maldridge>
eh, I would argue the first lab is the hardest to set up
00:05
you have to make real choices there and actually configure things, once you have that done you can spin up more easily
00:07
<TheProf>
There has been definite struggles that's for sure. One semi-related one is I'm trying to get Internet from another building 250 meters away through a wireless connection as I can't get reasonable bandwidth at the school site.
00:08
<maldridge>
ah, those are always fun
00:09
<TheProf>
In addition to the link posted above, are there other resources for ltsp-pnp so I can continue to educate myself about it?
00:10
<maldridge>
I'm not sure if there are any that are fully up to date, others might know of some
00:10
the guy that wrote that page hangs out here later at night
00:10lee has left IRC (lee!~lee@loathe.ms, Ping timeout: 276 seconds)
00:10
<vagrantc>
TheProf: there's not a lot to it
00:11
which is part of what makes it so good.
00:12lee_ has joined IRC (lee_!~lee@loathe.ms)
00:12
<TheProf>
ok. I think to mitigate any issues I'm going to set up another test server and try it all out there first. Because I'm a volunteer I can't always be there to trouble-shoot when things go wrong, and teachers are rather picky about their youtube access :)
00:13
<vagrantc>
oh, the biggest problem you might encounter with fat clients is the lack of authentication ... so screen lockers will lock the users out (or locking has to be disabled)
00:14
it does some pretty crude hacks to create a local user, but doesn't have a local password by default
00:14
<maldridge>
TheProf: you should probably also use if you aren't already some kind of configuration management and a vcs so you can back out changes that didn't go as you expected
00:14
<vagrantc>
!password
00:14
<ltsp>
Error: "password" is not a valid command.
00:14
<vagrantc>
!pass
00:14
<ltsp>
I do not know about 'pass', but I do know about these similar topics: 'fat-passwd', 'bypass_acceleration'
00:14
<vagrantc>
!fat-passwd
00:14
<ltsp>
fat-passwd: to allow fat client users to change their password, install the usermode package both in the server and in the chroot, and put these in lts.conf: REMOTE_APPS=True, RCFILE_01="sed 's,^Exec=,Exec=ltsp-remoteapps ,' -i /usr/share/applications/redhat-userinfo.desktop", RCFILE_02="sed 's,^Exec=,Exec=ltsp-remoteapps ,' -i /usr/share/applications/redhat-userpasswd.desktop"
00:14
<vagrantc>
whoah! that's not what i wa looking for
00:15
<maldridge>
still a great hack though
00:16
<vagrantc>
there's a feature in recent versions of LDM where you can set LDM_HASH_PASSWORD=true in lts.conf to store a local copy
00:16
so screen lockers can work again ... but changing the password still needs to be done on the server
00:16
(it will work locally, but be reset on reboot)
00:16
so that's confusing...
00:17
<TheProf>
vagrantc: I'm OK with changing passwords via the server -- the students don't change their own passwords normally.
00:17
<maldridge>
is there a reason that you can't authenticate through the pam stack with something like kerberos or ldap?
00:17
<TheProf>
but the screen locking is an issue for sure.
00:17
vagrantc: is that ltsp-bot post the solution to that?
00:17
<vagrantc>
maldridge: that's in the plans, but we tried implementing workarounds to having to set up kerberos or ldap ... and it's worked well enough, but has reached it's limits.
00:18
maldridge: our next workaround is libpam-sshauth
00:18
<maldridge>
ok, I run actual disk installs and I can confirm 100% that ldap is a pain to setup
00:18
<vagrantc>
well, it's less of a hack.
00:19
we'rehoping to get rid of LDM for LTSP6, but there hasn't been a lot of momentum on a major undertaking like that for a while.
00:19
<maldridge>
interesting
00:19
also xlock with an .xlockrc
00:20
would use a common lock screen password, but would allow for screen locking
00:20
<vagrantc>
http://wiki.ltsp.org/wiki/Dev:LTSPPamNotes
00:20
!pam
00:20
<ltsp>
I do not know about 'pam', but I do know about these similar topics: 'pad'
00:20
<vagrantc>
!libpam
00:20
<ltsp>
Error: "libpam" is not a valid command.
00:21
<vagrantc>
!ltsp6
00:21
<ltsp>
ltsp6: See: !future
00:21
<vagrantc>
!future
00:21
<ltsp>
future: Some of the highlights include working towards obsoleting LDM (using libpam-sshauth/libnss-sshsock), moving stuff out of ltsp-build-client into initramfs run-time modifications, and making it possible to maintain the chroot from a booted environment
00:21
<vagrantc>
we've already done nearly all the initramfs stuff, i guess. :)
00:21
<maldridge>
yeah, those certainly sound like large undertakings
00:23
<vagrantc>
ideally, i'd like to figure out ways to hook into any of the various remote options for pam
00:23
ldap, kerberos, even something sql based
00:23
<maldridge>
eh, its hilariously insecure, but I'd go for NIS first
00:23
<vagrantc>
though libpam-sshauth is our "in-house" attempt at hooking into pam
00:23
<maldridge>
I see
00:24
nis with kerberos would be reasonably secure, and gets around having to read X.509 docs for ldap
00:25
<vagrantc>
speaking of which, i need to fix a build failure for libpam-sshauth in debian
00:25
<maldridge>
is ltsp planning to pull any build deps that are systemd specific? and/or where could I find documentation for writing the system-specific bits of ltsp to port it to a new distro
00:26
<vagrantc>
nothing systemd specific, no
00:27
on debian, it's mostly falling back to the compatibility code
00:27
but it'd be good to improve the systemd support
00:27
maldridge: what distro?
00:29
<maldridge>
void
00:29
its runit based
00:29
<vagrantc>
this is an ancient list we used to eventually create ltsp5: http://wiki.ltsp.org/wiki/Dev:Compliance
00:30
wow, the 32MB of ram is out the window! hah.
00:30
oh, 2006
00:30
<maldridge>
wow, so network auth has been in there for a while
00:31
bbiaf, heading into work
00:31
<vagrantc>
maldridge: this might be useful: http://wiki.ltsp.org/wiki/Dev:Design
00:50
<maldridge>
I think I'm missing something
01:00vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Ping timeout: 252 seconds)
01:11gbaman has joined IRC (gbaman!~gbaman@2.29.16.44)
01:16gbaman has left IRC (gbaman!~gbaman@2.29.16.44, Ping timeout: 264 seconds)
01:31TheProf has left IRC (TheProf!~chatzilla@TOROON47-1176243489.sdsl.bell.ca, Ping timeout: 256 seconds)
02:04
<maldridge>
I'm trying to troubleshoot an nbd volume not being mounted at boot, any suggestions?
02:11vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
02:15
<vagrantc>
maldridge: the basic things to do to add support for a new distro in LTSP are appropriate hooks for ltsp-build-client to create a chroot that installs the ltsp-client and related packages (ltsp-client, ltspfs, ldm).
02:16
maldridge: https://bazaar.launchpad.net/~ltsp-upstream/ltsp/ltsp-trunk/files
02:16
the client directory contains code for ltsp-client, and the server directory for ltsp-server bits...
02:17
LDM is the display manager, and ltspfs i a fuse-based filesystem and automounter system for LTSP for mounting things like USB sticks and such...
02:19
<maldridge>
ok, so as long as I can get those things to build, its just down to writing hte bits that maintain the chroot
02:21
<vagrantc>
oh, and you'll need to make sure whatever initramfs system you use can do the read-only root with some sort of tmpfs overlay
02:22
<maldridge>
ok, that makes sense
02:22
<vagrantc>
ltsp-client and ltsp-server are largely build-your-own packages, since distros vary so widely, but ldm and ltspfs are fairly straightforward
02:22
<maldridge>
void has tried very hard to maintain more or less patch free upstream stuff
02:22
<jammcq>
hey vagrantc
02:23* vagrantc waves to jammcq
02:23
<jammcq>
how's everything going? haven't talked to anyone in a while
02:23
<vagrantc>
maldridge: well, ltsp is largely just a collection of scripts and hooks to scripts ... not really something you can just "make install"
02:23
jammcq: not much going on, unfortunately ...
02:24
jammcq: warren's stirring up some muck to get support back on track for fedora/redhat/centos and whatnot
02:24
<jammcq>
hmmm
02:24
I think he's moving to SanFran
02:26
<vagrantc>
jammcq: alkisg and i were hoping to try and make progress on LTSP6 stuff ... but it would be good to get more people with time, energy and motivation...
02:26
ltsp's kind of been in maintenance mode with infrequent bursts of activity for the last few years
02:27
<jammcq>
yeah, I miss the days when I was heavily involved
02:27
<maldridge>
vagrantc: would you say ltsp has a low barrier to entry?
02:27
<jammcq>
only 2 more years of having kids in college. hopefully I can relax a bit and get more involved when that's over
02:27
<vagrantc>
we're really close to syncig the ubuntu and debian packaging
02:28
maldridge: it's a bit unconventional, in that it's basically a glue project that ties a bunch of other software together to support network-booted computers
02:28
<jammcq>
it's my understanding that both debian and ubuntu are moving to systemd, that should make them more alike
02:29* vagrantc is pretty init system agnostic
02:29
<vagrantc>
hopefully we can keep ltsp that way
02:29
<maldridge>
jammcq: yes, they are both going to systemd
02:30
vagrantc: I ask because I am the acting president of my university's lug, and we're looking for a decent open source project to point people to
02:30
our lab is built on top of ltsp and so it seems like a nice fit
02:30
<vagrantc>
maldridge: it's a project that's sort of between software development and system administration
02:30
<jammcq>
maldridge: where are you located?
02:30
<maldridge>
jammcq: Texas
02:31
<jammcq>
cool, what part?
02:31
<maldridge>
North part
02:31* maldridge wonders why jammcq wants to know
02:31
<jammcq>
i've been to dallas, austin and houston
02:31
i'm asking cuz i'm interested to know where my project is being used
02:32
<gehidore>
maldridge: because people like to relate to other people
02:32
<maldridge>
ah, UT Dallas
02:32
<vagrantc>
maldridge: it'd be great to get more involvement, but i can't really speak to the barriers to entry... i've been a developer for almost 10 years now, and one of the few active developers for the last several
02:32
<jammcq>
ah, cool
02:32
<vagrantc>
!map
02:32
<ltsp>
I do not know about 'map', but I do know about these similar topics: 'worldmap'
02:32
<vagrantc>
!worldmap | echo maldridge
02:32
<ltsp>
maldridge worldmap: If you're using LTSP, please let the world know and share your story at http://www.ltsp.org/stories/ Your can add a nice pin to our world map at your location, plus your setup will count towards the global LTSP usage statistics.
02:32
<maldridge>
vagrantc: I worked on debian for a few years, I wouldn't recommend that as a first time project to anyone, way too much paperwork to navigate
02:33* gehidore will add his story once the system is in place and used daily
02:33* vagrantc has been involved in debian for even longer
02:33
<maldridge>
I'll add an entry as soon as I can debug this nbd problem
02:34
<vagrantc>
maldridge: i find debian has a very low barrier to entry ... i got involved by sending a few emails to the bug tracking system...
02:34
<maldridge>
I enjoyed debian, but I didn't want to follow torwards systemd
02:34
vagrantc: I'm thinking for things that are roadmap level things
02:34AlexPortable has left IRC (AlexPortable!uid7568@gateway/web/irccloud.com/x-emwaqmyogwnokblj, Quit: Connection closed for inactivity)
02:34
<maldridge>
like I know some people that would be happy to help with the pamssh option, but don't like really working with the politics of things
02:35
<vagrantc>
politics?
02:35
<maldridge>
you know, people arguing over large changes
02:35
<vagrantc>
ok.
02:36* vagrantc finds the vast majority of politics in debian are people submitting patches, bug reports, committing code and sharing ideas
02:36
<vagrantc>
people get too hung up on the disagreements to see the massive-scale cooperation
02:36
<maldridge>
I tend to find that as the type of linux user in my age bracket, willing to write decent quality code, not willing to work through disputes of design choices
02:36
<vagrantc>
disputes are everywhere, just depends on how transparent those disputes are
02:37
<maldridge>
yeah, I think my first step is going to try and break the perception that linux can't do cool stuff, which is part of why I'm trying to get steam running inside ltsp
02:37
<vagrantc>
if it works on a desktop, it shouldn't be hard to do as a fat client
02:38
<maldridge>
then work through the, this is how projects work
02:38
I'm having to be sneaky so it isn't trying to download the games for every workstation
02:38
basically the games exist on a COW filesystem that the client stubbornly refuses to mount
02:40
clients complain "Cannot open NBD, no such file or directory"
02:40
<vagrantc>
what distro?
02:40
<maldridge>
ubuntu in this setup
02:41
I've checked and its definitly got the module loaded
02:41
<vagrantc>
is dhcp telling it the correct server to use, or has it passed that on the commandline?
02:42
<maldridge>
so it gets all the way up to an ldm session, this is for a second nbd volume I'm trying to mount
02:44
<vagrantc>
ah.
02:44
so LTSP's working fine, you're just trying to mount another filesystem?
02:44
<maldridge>
on the server I get a "read failed, bad file descriptor" right after it tries to create the map/diff for the second fs
02:45
yes, that's correct
02:45
<vagrantc>
so are the device nodes present in /dev/nbd* ?
02:45
and what are the nbd-client arguments?
02:45
<maldridge>
yes, there is a /dev/nbd1
02:46
<vagrantc>
/dev/nbd1 is often used for swap in LTSP
02:46
<maldridge>
nbd-client -N /opt/games 192.168.1.2 nbd1
02:46
ah, so I should pick a higher numbered one?
02:47
<vagrantc>
maybe
02:47
you may need to specify the full path ... e.g. /dev/nbd2
02:48
<maldridge>
that would seem to have done it
02:48
I forgot about the swap file
02:49
<vagrantc>
all the examples in the manpage use the full path
02:49* vagrantc goes back to reading books
02:49
<maldridge>
yeah, using /dev/nbd1 complains about writing to a closed socket, but /dev/nbd2 works
02:52
okay, so I can mount it once the machine is up, but it won't mount during boot
03:47cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 244 seconds)
03:53cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg)
03:58
<maldridge>
do fat filesystems save the log for boot time mounts anywhere?
03:58
s/filesystems/clients
04:06
so I have narrowed it down to it always failing to mount during boot, but then not once the system comes up
05:01
<work_alkisg>
are you using dns?
05:01
or the server ip?
05:03
<maldridge>
ip
05:03
should I use DNS?
05:27
it looks like they get half way through mounting because if I try to use the device later, it complains its already attached
05:30
found it, I'm pretty sure that the FSTAB entry is being read before the INIT steps complete
07:12
does the lts.conf file support include statemends
07:13
*statements
07:14gbaman has joined IRC (gbaman!~gbaman@2.29.16.44)
07:18gbaman has left IRC (gbaman!~gbaman@2.29.16.44, Ping timeout: 244 seconds)
07:32gbaman has joined IRC (gbaman!~gbaman@2.29.16.44)
07:38gbaman has left IRC (gbaman!~gbaman@2.29.16.44, Ping timeout: 248 seconds)
07:53gbaman has joined IRC (gbaman!~gbaman@2.29.16.44)
07:59gbaman has left IRC (gbaman!~gbaman@2.29.16.44, Ping timeout: 255 seconds)
08:04gbaman has joined IRC (gbaman!~gbaman@2.29.16.44)
08:21gbaman has left IRC (gbaman!~gbaman@2.29.16.44, Remote host closed the connection)
08:30vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving)
08:35gbaman has joined IRC (gbaman!~gbaman@2.29.16.44)
08:56gbaman has left IRC (gbaman!~gbaman@2.29.16.44, Ping timeout: 276 seconds)
09:19ricotz has joined IRC (ricotz!~rico@ubuntu/member/ricotz)
09:49telex has left IRC (telex!~telex@freeshell.de, Remote host closed the connection)
09:50telex has joined IRC (telex!teletype@freeshell.de)
09:53gbaman has joined IRC (gbaman!~gbaman@2.29.16.44)
09:58gbaman has left IRC (gbaman!~gbaman@2.29.16.44, Ping timeout: 264 seconds)
10:33ogra_ has left IRC (ogra_!~ogra_@p5098ed03.dip0.t-ipconnect.de, Ping timeout: 252 seconds)
10:34ogra_ has joined IRC (ogra_!~ogra_@p5098ed03.dip0.t-ipconnect.de)
10:42gbaman has joined IRC (gbaman!~gbaman@2.29.16.44)
10:59cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 246 seconds)
11:01cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg)
11:37cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 250 seconds)
11:37cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg)
11:53cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Remote host closed the connection)
12:01gbaman has left IRC (gbaman!~gbaman@2.29.16.44, Remote host closed the connection)
13:39gbaman has joined IRC (gbaman!~gbaman@dab-crx1-h-21-1.dab.02.net)
13:51adrianorg has left IRC (adrianorg!~adrianorg@179.187.29.137.dynamic.adsl.gvt.net.br, Ping timeout: 264 seconds)
13:53adrianorg has joined IRC (adrianorg!~adrianorg@179.183.71.11)
14:12gbaman has left IRC (gbaman!~gbaman@dab-crx1-h-21-1.dab.02.net, Ping timeout: 252 seconds)
14:33gdi2k has joined IRC (gdi2k!~gdi2k@203.177.248.238)
15:24vervelak has left IRC (vervelak!~vervelak@139.91.248.3, Quit: leaving)
15:25vervelak has joined IRC (vervelak!~vervelak@139.91.248.3)
15:27vervelak has left IRC (vervelak!~vervelak@139.91.248.3, Client Quit)
15:28vervelak has joined IRC (vervelak!~vervelak@139.91.248.3)
16:03khildin has joined IRC (khildin!~khildin@ip-62-235-40-152.dial.scarlet.be)
17:07khildin has left IRC (khildin!~khildin@ip-62-235-40-152.dial.scarlet.be, Ping timeout: 248 seconds)
17:08andygraybeal has left IRC (andygraybeal!~andy@h104.60.88.75.dynamic.ip.windstream.net, Ping timeout: 256 seconds)
17:53nahoskins has left IRC (nahoskins!~nahoskins@64.147.84.76, Ping timeout: 246 seconds)
18:04vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
19:25lee_ is now known as lee
19:51vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Ping timeout: 256 seconds)
20:07vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
20:07vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Client Quit)
20:11ricotz has left IRC (ricotz!~rico@ubuntu/member/ricotz, Quit: Ex-Chat)
20:26telex has left IRC (telex!teletype@freeshell.de, Remote host closed the connection)
20:28telex has joined IRC (telex!teletype@freeshell.de)
21:36vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
22:09gbaman has joined IRC (gbaman!~gbaman@host86-184-75-156.range86-184.btcentralplus.com)