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


Channel log from 14 February 2010   (all times are UTC)

01:00mushroomblue has joined #ltsp
01:12abeehc has quit IRC
01:14abeehc has joined #ltsp
01:30mushroomblue has quit IRC
01:31mushroomblue has joined #ltsp
01:33mushroomblue has joined #ltsp
01:58Kicer86 has joined #ltsp
02:42alkisg has joined #ltsp
03:02Kicer86 is now known as Kicer86|AWAY
03:41highvoltage has joined #ltsp
03:59ogra has quit IRC
04:00ogra has joined #ltsp
04:02mikkel has joined #ltsp
04:27ogra_ has joined #ltsp
04:30ogra has quit IRC
04:32cyberorg has joined #ltsp
04:32cyberorg has joined #ltsp
04:35Barbosa_ has joined #ltsp
04:39Barbosa has quit IRC
05:11ogra_ is now known as ogra
05:11ogra has joined #ltsp
05:35alkisg has quit IRC
05:37leio has quit IRC
05:52leio has joined #ltsp
06:35HardDisk has quit IRC
06:35HardDisk has joined #ltsp
06:37pmatulis has joined #ltsp
08:00Gadi_eeepc has quit IRC
08:06Gadi_eeepc has joined #ltsp
08:07bobby_C has joined #ltsp
08:09Kicer86|AWAY is now known as Kicer86
08:15GGD has joined #ltsp
08:16Gadi_eeepc has quit IRC
08:21Gadi_eeepc has joined #ltsp
08:42Gadi_eeepc has quit IRC
08:49Gadi_eeepc has joined #ltsp
08:56bzh29 has joined #ltsp
08:59litlebuda has joined #ltsp
09:04bzh29 has quit IRC
09:10Gadi_eeepc has quit IRC
09:17Gadi_eeepc has joined #ltsp
09:30alkisg has joined #ltsp
09:32artista_frustrad has joined #ltsp
09:44litlebuda has quit IRC
09:50Gadi_eeepc has quit IRC
09:53Gadi_eeepc has joined #ltsp
10:02bobby_C has quit IRC
10:06Gadi_eeepc has quit IRC
10:25Gadi_eeepc has joined #ltsp
10:35Gadi_eeepc has quit IRC
10:38frederickjh has joined #ltsp
10:39Gadi_eeepc has joined #ltsp
10:56Gadi_eeepc has quit IRC
11:06vagrantc has joined #ltsp
11:13
<alkisg>
!Gadi++ :)
11:13
<ltspbot>
alkisg: Error: "Gadi++" is not a valid command.
11:22pmatulis has quit IRC
11:26GGD_ has joined #ltsp
11:29GGD has quit IRC
11:31adrianorg has joined #ltsp
11:47mikkel has quit IRC
12:01
<moldy>
hi
12:01
where is the udev rules file in karmic?
12:02
the manual says /opt/ltsp/<arch>/etc/udev/rules.d/<some-letter-and-number>-ltspfd.rule
12:02
but i don't see such a file on my system
12:03
<alkisg>
dpkg -L udev
12:04
<moldy>
alkisg: on the client?
12:04
<alkisg>
anywhere - just to see the paths...
12:05Egyptian[Home] has quit IRC
12:05
<moldy>
well, the paths differ on client and server
12:05
client just has 2 files for persistent net and cd stuff
12:06
also, the file i am looking for is not part of the udev package, i think?
12:08
it's /lib/udev/rules.d
12:09
<alkisg>
Which file are you looking for?
12:09
<moldy>
alkisg: i want to enable local hard disk access on the thin client
12:09
<vagrantc>
recent versionsm of udev moved the location ... docs should be updated.
12:09
<moldy>
LTSPManual.html says i should edit /opt/ltsp/<arch>/etc/udev/rules.d/<some-letter-and-number>-ltspfd.rule
12:11
hm, 80-ltspfsd.rules already contains an entry that looks like it should detect hard disks
12:13slidesinger has joined #ltsp
12:13alkisg has quit IRC
12:18
<vagrantc>
moldy: local hard disks are excluded, you need to set an lts.conf option ...
12:18
LOCALDEV_DENY_INTERNAL_DISKS=false
12:21Egyptian[Home] has joined #ltsp
12:29highvoltage has quit IRC
12:31highvoltage has joined #ltsp
12:56comfrey has joined #ltsp
12:56Egyptian[Home] has quit IRC
13:07
<moldy>
vagrantc: thanks
13:08
vagrantc: i can never remember where all those lts.conf options are documented
13:09
that one in particular does not seem to be documented at all
13:11Egyptian[Home] has joined #ltsp
13:12
<vagrantc>
!lts.conf
13:12
<ltspbot>
vagrantc: "lts.conf" :: http://manpages.ubuntu.com/lts.conf
13:12
<vagrantc>
there's also an upstream version
13:12
yes, ltsp-docs doesn't seem to document it ...
13:13
that's a fairly important one *sigh*
13:33arreyder has quit IRC
13:33arreyder has joined #ltsp
13:53bobby_C has joined #ltsp
13:54Egyptian[Home] has quit IRC
13:59alexqwesa_ has joined #ltsp
14:00alexqwesa has quit IRC
14:04litlebuda has joined #ltsp
14:09Egyptian[Home] has joined #ltsp
14:22bobby_C has quit IRC
14:26alexqwesa_ has quit IRC
14:26alexqwesa_ has joined #ltsp
14:28bobby_C has joined #ltsp
14:30arreyder has quit IRC
14:36arreyder has joined #ltsp
14:41leio has quit IRC
14:44arreyder has quit IRC
14:44arreyder has joined #ltsp
14:50alexqwesa_ has quit IRC
14:54alexqwesa_ has joined #ltsp
14:58frederickjh has quit IRC
14:59pmatulis has joined #ltsp
15:00Egyptian[Home] has quit IRC
15:01klausade has quit IRC
15:09etyack has joined #ltsp
15:10etyack has left #ltsp
15:15jhutchins has quit IRC
15:15jhutchins has joined #ltsp
15:18litlebuda has quit IRC
15:21jhutchins has quit IRC
15:24jhutchins has joined #ltsp
15:49
<comfrey>
vagrantc: that backport is good to go?
15:51
oh, and is it a bug that the default dnsmask.conf for debian is pointing at nbi rather than nfs?
15:56
nevermind .. i just need to build the correct arch i think...
15:57Kicer86 has quit IRC
16:01
<vagrantc>
comfrey: freegeek is using it
16:02* comfrey updates sources.list
16:02
<vagrantc>
comfrey: it points to pxelinux for clients that report pxe, and nbi for clients that report etherboot
16:03
comfrey: you'll also have to manually specify the backports when building... it's a little clunky.
16:05
comfrey: ltsp-build-client --pin-dist lenny-backports --pin-packages 'ltsp-client ltsp-client-core ltspfsd ltspfsd-core ldm' --extra-mirror 'http://www.backports.org/debian lenny-backports main' --apt-key /etc/apt/trusted.gpg # assuming you've got backports.org keys in your keyring.
16:05
think that ought to do it.
16:07
i was really hoping to have 5.1.99 on backports.org, but the broken mipsel buildd's are preventing it from migrating to testing...
16:07
i've been fighting all day with my mipsel box, but it is apparently unstable.
16:07
don't know if it's my hardware or the software...
16:08
i've got to update the backports page ... it's a little more complicated than i'd like... i wonder if it will work to just have a --backports option.
16:10Gadi_eeepc has joined #ltsp
16:15
<comfrey>
vagrantc: interesting, thanks. it looks like amd64 is not supported out of alioth.
16:15
i guess i'll head to backports.org
16:15
<vagrantc>
comfrey: you want amd64 clients? or the server packages aren't available?
16:16
<comfrey>
server packages
16:16
what do you recomend?
16:16
<vagrantc>
you could just pull the server-side packages from squeeze- there's no changes to any of those in the backports.
16:16
they were trivial rebuild backports.
16:17
oh, ltspfs may require actual backporting at this point.
16:17
well, trivial rebuild style.
16:22
<comfrey>
does the backport hav 337 more dependancies than the regular lenny package?
16:22
<vagrantc>
yes, but they should all be satisfyable in lenny.
16:23
<comfrey>
why so many more dependancies?
16:23
<vagrantc>
a lot of new features :)
16:24
if you really don't want them all, you can just use --early-packages 'ltsp-client-core $foo $bar $baz'
16:26
<comfrey>
ok, rad.
16:26
off for tea while 530 MB come down
16:36alkisg has joined #ltsp
16:42
<Gadi_eeepc>
vagrantc, alkisg: I modified ltsp-update-image today making it a bit more intelligent (hopefully) in establishing ports and writing configs
16:43
Tests well on my end, but looking forward to some additional testing
16:43
<alkisg>
Hey Gadi. I'm looking at it right now :)
16:43
<vagrantc>
Gadi_eeepc: yeah, saw that.
16:43
<alkisg>
Gadi_eeepc: would you mind if I excluded /proc in mksquashfs by default?
16:43
Because right now, it'll break with fat clients...
16:43
<vagrantc>
comfrey: just committed a plugin to make backports support much simpler.
16:44
alkisg: you'll have to figure out some way to add it back in
16:44
<comfrey>
vagrantc: rad ... you need any work these days?
16:44
<vagrantc>
alkisg: i.e. the emty directory needs to be there.
16:44
<Gadi_eeepc>
alkisg: oh, I see - there's code to unmount /proc but if it does not unmount, it still get included, eh?
16:44
<alkisg>
vagrantc: ah. I saw it as an example in the mksquashfs man page, and I thought it would work out of the box...
16:45
<Gadi_eeepc>
perhaps we can use umount -l?
16:45
or are we anti that?
16:45
<alkisg>
Gadi_eeepc: I had added an unmount -l for the fat clients case, in case everything goes wrong - I don't have any clean solutions right now, so a workaround should suffice...
16:45
<Gadi_eeepc>
ok
16:45
<alkisg>
umount -l works for me. I know it's not clean, but it's not ltsp-update-image's fault..
16:45
OK, I'll add it tomorrow
16:46
Another thing, about /etc/default/ltsp-update-image
16:46
<Gadi_eeepc>
I didnt muck with the generation code
16:46
mostly just the port stuff
16:46
and the cfg file writing
16:46
<alkisg>
1) It should be included *before* the parameters handling, and
16:46
2) shouldn't that go to /etc/ltsp/ltsp-update-image.conf, like all other files?
16:47
(you didn't change that, I was this way before...)
16:47
<stgraber>
btw, I'd like to tag 5.2 on Wednesday if possible. That's assuming we get nbd-proxy fixed by then ...
16:47
so I can upload 5.2 just before FF on Thursday
16:47
<Gadi_eeepc>
I definitely agree with #2
16:47
<alkisg>
Gadi_eeepc: about 1), it just means that command line parameters override the defaults in the file.
16:47
<Gadi_eeepc>
as for #1, I dont think it matters
16:48
oh, I see
16:48
ur right
16:48
command line should override
16:48
Il move it to the top and recommit
16:49
<alkisg>
So generally it is: 1) conffiles 2) parameters 3) defaults...
16:51
<Gadi_eeepc>
pushed
16:52
alkisg: I realized why there was some weirdness as to what ltsp-update-image does versus ltsp-update-kernels. Its simply because not all distros call l-u-i
16:52
but aall call l-u-k
16:52
so, all the nbd port stuff for the image needs to be part of l-u-i
16:53
but, I tried to be careful in the calculations to find ports that are not used by other chroots or other running services on the system (with netstat)
16:54
<alkisg>
Thanks Gadi, that was a nice work.
16:55
Could we somehow not hardcode "quiet splash" ?
16:55
<Gadi_eeepc>
oh yeah!
16:55
I was going to add a -o option
16:55
to pass BOOTOPTARGS
16:58
I mean BOOTPROMPT_OPTS
16:59bobby_C has quit IRC
16:59
<vagrantc>
maybe we could set up some common values that both ltsp-update-image and ltsp-update-kernels use?
16:59
<alkisg>
echo "BOOTPROMPT_OPTS=\"quiet splash nbdport=${NBD_PORT}\"" >${THIS_CHROOT}/etc/ltsp/update-kernels.conf
16:59
<vagrantc>
a common configuration
16:59
<alkisg>
Maybe nbdport shouldn't be part of BOOTPROMPT_OPTS
17:00* alkisg votes for /etc/ltsp/update-kernels.conf, both on the chroots and on the server...
17:04
<Gadi_eeepc>
thats already how it should work
17:05
if you pass -o OR if you set BOOTPROMPT_OPTIONS in a server side config file, l-u-i will write those options correctly in the chroot and run update-kernels in the chroot
17:05
<alkisg>
Gadi: append ro initrd=initrd.img quiet splash nbdport=${NBD_PORT}
17:06
BOOTPROMPT_OPTS aren't used there...
17:06
<Gadi_eeepc>
where is that/
17:06
<alkisg>
192
17:06
*191
17:07
<Gadi_eeepc>
doh
17:07
will fix it
17:07
of course that part is only called as a last resort
17:07
<alkisg>
Gadi_eeepc: but if you put BOOTPROMPT_OPTS there, you'll have 2 nbdports
17:07
<Gadi_eeepc>
normally, it grabs the file from the chroot
17:07
no, it'll be ok
17:08
<alkisg>
That's why I think that nbdport shouldn't be part of BOOTPROMPT_OPTS, in line 167
17:08
<Gadi_eeepc>
oh, and I prolly dont need nbdport on the kernel line
17:08
just the append
17:09
<alkisg>
Yup
17:10
<Gadi_eeepc>
ok
17:10
pushed
17:10
gotta run for a while
17:10
let me know if I screwed anything else up! :)
17:11
<alkisg>
Nah, you did a fine job. It's the architecture that's screwed...
17:15alexqwesa_ is now known as alexqwesa
17:16
<alkisg>
I still think that this line will result in infinite nbdports:
17:16
echo "BOOTPROMPT_OPTS=\"${BOOTPROMPT_OPTIONS} nbdport=${NBD_PORT}\"" >${THIS_CHROOT}/etc/ltsp/update-kernels.conf
17:21
Ah, they're not sourced from update-kernels, they're regenerated on each run...
17:21
ok
17:22
So I'd need to set BOOTPROMPT_OPTS="splash" on the server, and that would get me BOOTPROMPT_OPTS="splash nbdport=2001" on the chroot. Got it.
17:27
<Gadi_eeepc>
you would set BOOTPROMPT_OPTIONS="splash"
17:27
BOOTPROMPT_OPTS is an internal variable
17:28
btw, despite how the code might look, it is still the chroot's update-kernels right ow that creates that pxelinux.cfg file
17:28
should we *only* have the server create that file?
17:29
I would need to grab those pieces for TIMEOUT and IPAPPEND
17:31
tho, all the code is already there
17:32
maybe I should remove it from the server side and add flags to write TIMEOUT and IPAPPEND variables to the chroot's conf file
17:33
hmm... and we should handle nbdroot in a sane way
18:56alkisg has quit IRC
19:27vagrantc has quit IRC
19:52adrianorg has quit IRC
20:01Gadi_eeepc1 has joined #ltsp
20:01Gadi_eeepc has quit IRC
20:03Gadi_eeepc1 has quit IRC
20:04Gadi_eeepc has joined #ltsp
20:13GGD_ has quit IRC
20:24shogunx has quit IRC
20:48GGD has joined #ltsp
21:00
<Gadi_eeepc>
stgraber: can you upload latest upstream to your ppa, say tomorrow, so we can test more easily before tagging?
21:00
<stgraber>
Gadi_eeepc: sure
21:00
Gadi_eeepc: will do so later tonight
21:01
<Gadi_eeepc>
excellent - if you could do ltsp, ldm and ltspfs that would be great
21:01
<stgraber>
I didn't see any commit to ldm or ltspfs today and I did an upload last night
21:01
but I'll have a quick look see if something changed
21:02Gadi_eeepc has quit IRC
21:03
<stgraber>
ltsp done (rev 1665)
21:04
ldm is already up to date (rev 1242)
21:04
and so is ltspfs-trunk (rev 138)
21:12pmatulis has quit IRC
21:20Gadi_eeepc has joined #ltsp
21:33
<stgraber>
Gadi_eeepc: they are all in my PPA (though maybe ltsp is still building)
21:34
just checked, they are all built.
21:42
<Gadi_eeepc>
stgraber: excellent - I am looking forward to hearing from folks how they test on lucid
21:45
stgraber: are they built for karmic, as well?
22:01F-GT has quit IRC
22:02
<stgraber>
Gadi_eeepc: nope, that's Lucid only. I have no clue if a backport would work on karmic
22:02
I'll be testing that when backporting 5.2
22:17GGD has quit IRC
22:17F-GT has joined #ltsp
22:30F-GT has quit IRC
22:48F-GT has joined #ltsp
23:27Egyptian[Home] has joined #ltsp
23:43johnny has left #ltsp
23:53johnny has joined #ltsp