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


Channel log from 20 August 2010   (all times are UTC)

00:04vagrantc has joined #ltsp
00:06migue has joined #ltsp
00:11gentgeen__ has joined #ltsp
00:22vagrantc has quit IRC
00:36alkisg has joined #ltsp
00:38
<fluvvell>
I think I've solved my problem, two dnsmasq servers on the network.
01:05
<johnny>
hah.. that'll do it :)
01:07alkisg has quit IRC
01:09alkisg has joined #ltsp
01:22alkisg has quit IRC
01:23Guerdal82 has left #ltsp
01:28alkisg has joined #ltsp
01:31sweetpi has quit IRC
01:36irule has quit IRC
01:48alkisg has quit IRC
01:52alkisg has joined #ltsp
02:08dobber has joined #ltsp
02:28alkisg has quit IRC
02:28alkisg has joined #ltsp
02:55Da-Geek has joined #ltsp
03:02nutron has quit IRC
03:04nutron has joined #ltsp
03:05NeonLicht has quit IRC
03:13Faithful has quit IRC
03:16alkisg has quit IRC
03:33Faithful has joined #ltsp
03:38NeonLicht has joined #ltsp
03:45Da-Geek has quit IRC
03:46poff has joined #ltsp
03:50F-GT has quit IRC
03:51F-GT has joined #ltsp
04:01Faithful has quit IRC
04:08Selveste1___ has joined #ltsp
04:13Selveste1___ has quit IRC
04:48dlezcano has quit IRC
04:50poff has quit IRC
05:28alkisg has joined #ltsp
05:58alkisg has quit IRC
06:28pmatulis has joined #ltsp
06:38alkisg_ has joined #ltsp
06:40g4tsu has joined #ltsp
06:41
<g4tsu>
Hello
06:42
I've got a problem (sometimes) when I launch ltsp boot, I've this message : nfsmoutn need a path
06:43
Here is my /etc/ltsp/dhcpd.conf : http://pastebin.ca/1921344
06:43
(I'm using LTSP 5)
06:53
<alkisg_>
distro/version?
06:54
<g4tsu>
debian lenny
06:54
But I see that I have two dhcp on my network
06:54
I'll try to stop one
06:56
<sep>
g4tsu, hard to keep a conversation
06:56
<g4tsu>
Hmm even if I stop a dhcp , I've got nfsmount need a path
06:57
<alkisg_>
And the client doesn't boot? Can you post your pxelinux.cfg/default?
06:57
<g4tsu>
It boots
06:57
But I have nfsmount need a path and a "crash"
06:58
<alkisg_>
So it doesn't complete its booting, it crashes in the middle, right?
07:00
<g4tsu>
Yep
07:01
<alkisg_>
OK, can you post your pxelinux.cfg/default?
07:01
<g4tsu>
After few times, I've got an initramfs
07:01
in /var/lib/tftpbooot ?
07:01
<alkisg_>
I'm not sure where debian puts that
07:01
<sep>
look at the same place as pxelinux.0 is
07:02
<alkisg_>
/var/lib/tftpboot, /srv...
07:02
<g4tsu>
ok
07:02
I have /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default
07:02
Which contains :
07:02
DEFAULT vmlinuz ro initrd=initrd.img quiet root=/dev/nfs ip=dhcp boot=nfs
07:03
<alkisg_>
And currently you only have the dhcp server of the ltsp server active? No other dhcp servers around?
07:05
<g4tsu>
I stop the dhcp on the server. I couldn't stop the dhcp on the router
07:05
<alkisg_>
That won't work
07:05[GuS] has joined #ltsp
07:05
<g4tsu>
(Work network)
07:05
ok
07:05
<alkisg_>
You either need a 2 nic setup, or to put a proxydhcp server in your ltsp server
07:05
<g4tsu>
I'll try to isolate the network
07:05
ok
07:06
Thanks for the advice
07:06
<alkisg_>
https://help.ubuntu.com/community/UbuntuLTSP/ProxyDHCP
07:16alkisg_ has quit IRC
07:21alkisg_ has joined #ltsp
07:26
<[GuS]>
Hi guys, i have a small problem. I've implemented so far LTSP5 (in gentoo) and it works now so far (at least basic things, i didnt full tested yet). Now, since in my work we used to configure for some users to login agaist other server, this is not working. I've added the entry SERVER=IP to this user but does not login, which works on older ltsp4.2. I am missing something?
07:28mlis has joined #ltsp
07:29moose-tek has quit IRC
07:29moose-tek has joined #ltsp
07:33
<mlis>
hi i have a problem with ltsp on ubuntu 10.04 if i booting PXE client i get this massage "Error: failed to connect NBD server" anybody tell me what i should to do?
07:34
<alkisg_>
mlis: do you get a busybox prompt after that?
07:35
<mlis>
no:) i must read about this thx for now...
07:35
<alkisg_>
You don't want more help? OK...
07:40
<[GuS]>
I want! :P
07:43
<mlis>
ok i just look for this i get "BysyBox v 1.13.3 built in shell Enter help for a list of build.in commands" And that's all.
07:50wurzzero has joined #ltsp
07:54gnunux has joined #ltsp
08:03redbaritone has joined #ltsp
08:07mlis has left #ltsp
08:30Selveste1___ has joined #ltsp
08:35Selveste1___ has quit IRC
09:17wurzzero has quit IRC
09:25otavio has quit IRC
09:25otavio has joined #ltsp
09:25otavio has joined #ltsp
09:29alkisg_ has quit IRC
09:33mikkel has joined #ltsp
09:51mistik1 has quit IRC
09:57mighty-d has joined #ltsp
09:57
<mighty-d>
Hi
10:01
i was wondering if there are any *server* performance issues with a 64 bits Ubuntu 10.04 system. I have a setup where the server is running an i7 and 8GB RAM for only 12 users. The uptime measure is always above 1 and periodically increases up to 30, the memory consumption is always above 50%. The weird thing is that this server is an upgrade of a Core 2 4GB RAM running 32 bits Ubuntu 8.04 that worked flawlessly.
10:03jhutchins_lt has joined #ltsp
10:07gnunux has quit IRC
10:12dobber has quit IRC
10:14wima1 has joined #ltsp
10:15
<wima1>
does anyone else has problems with evolution-data-server processes lingering on?
10:17
<Appiah>
thats a common gnome problem
10:17
not a ltsp issue really
10:19
<wima1>
Appiah: I know, but it is on ltsp setups that these things can become a problem, so i was wondering if anyone else had problems with this
10:21
i have a server of which the load goes up to 50 about once a day for about 5 minutes. It started a few days ago
10:21
im trying to find out what is wrong, and i stumbled on all the evolution processes
10:22
<Appiah>
Yes I have a install with that problem
10:23
but its ubuntu 9.04 or so
10:24
the servers are schedulded to reboot 1 time a week and are only being used 5days a week
10:24
<wima1>
for some reason it starts swapping like crazy. 5 minutes later everything is fine
10:24
<Appiah>
so maybe some late friday there alot of proccesses but they dont have any impact on the server...
10:25
<wima1>
strange also: it has only 4 out of 6 GB memory in use, but swap is 100% full
10:28staffencasa has joined #ltsp
10:31
<wima1>
could LDM_LIMIT_ONE_SESSION be of any help here??
10:32
<Appiah>
does not for us
10:33
<wima1>
I cannot find out what it does exactly
10:33
doesn't it kill all leftover processes?
10:34
<Appiah>
it kills procceses if there's any left from the last session
10:34
when a user logs on
10:34
(if I'm correct)
10:34
<wima1>
sounds like something i want :)
10:39crmanski has joined #ltsp
10:40
<crmanski>
Hello, I have the same issue as this person: http://www.mail-archive.com/ltsp-discuss@lists.sourceforge.net/msg38555.html
10:41
Does anyone have any ideas hot to debug?
10:42wima1 has quit IRC
10:44g4tsu has quit IRC
10:47wima has joined #ltsp
10:56chupacabra has quit IRC
10:58mighty-d has quit IRC
11:01mighty-d has joined #ltsp
11:01mighty-d is now known as Guest92642
11:15wima has left #ltsp
11:20Guest92642 has quit IRC
11:21jhutchins_lt has quit IRC
11:21jhutchins_lt has joined #ltsp
11:36crmanski has quit IRC
11:46vagrantc has joined #ltsp
12:08
<klausade>
vagrantc: hello.
12:10
<vagrantc>
klausade: hi!
12:12mip has joined #ltsp
12:13hersonls has joined #ltsp
12:13
<mip>
Hi Folks - I'm running 10.04.1 on asus eeepc1000he in a school lab environment. It's running great. I'd like to set it so that students can't alter the desktop picture, the panel, and also so that they can't see the credentials used to join the wireless network. Any ideas?
12:14alkisg has joined #ltsp
12:15
<mip>
For the record - I have about 120 netbooks running ubuntu, and two ltsp servers. I'd like to manage/restrict the student desktops similarly - with the addition of disabling the viewing of the wireless password on the netbooks.
12:16
Sorry - that wasn't clear. The netbooks are stand alones (not thin clients). The ltsp servers serve about 30 clients in another building.
12:17mistik1 has joined #ltsp
12:17
<johnny>
for most of it can you use sabayon
12:17
not sure how to handle the wireless thing
12:18
<mip>
Does sabayon work now?
12:18
it used to crash a lot
12:18
<johnny>
i think so..
12:18
it got a lot of work
12:18
<mip>
okay - I'll give it a shot. I remember it being fairly intuitive
12:19
<klausade>
vagrantc: i'm looking at "ltsp-build-client --extra-help" and find this strange: --no-diskless-edu-workstation install debian-edu diskless workstation environment
12:20
vagrantc: no call it --no-diskless-edu-workstation when it installs debian-edu diskless workstation environment, typo?
12:20
vagrantc: this is ltsp 5.2.4-1
12:20mip has quit IRC
12:20alkisg has joined #ltsp
12:22Mip5 has joined #ltsp
12:24poff has joined #ltsp
12:33migue has quit IRC
13:06
<vagrantc>
klausade: that must be a custom plugin added by debian-edu
13:07
klausade: look in /usr/share/ltsp/plugins/ltsp-build-client/Debian*
13:16
<klausade>
vagrantc: ok
13:25
<alkisg>
vagrantc: another idea to prevent e.g. network-manager from interfering would be to delete /etc/init.d/network-manager from the initramfs or from an early initscript. Can we do that on an nfs booted thin client? Is it a good idea?
13:25
https://bugs.launchpad.net/ubuntu/+source/ltsp/+bug/619398
13:25
(instead of the blacklisting we're currently doing...)
13:25
<vagrantc>
alkisg: can't really delete it
13:26
alkisg: might be able to mount and write over it
13:26
<alkisg>
Erm ok that would also work, but is it better than blacklisting?
13:27
(in my view, less trouble installing stuff, a little more cross distro... but still not an elegant way :-/)
13:28* alkisg wonders if upstart / init support blacklisting initscripts in some configuration file...
13:28
<vagrantc>
alkisg: alternately, is there a configuration file that /etc/init.d/network-manager sources that could disable it?
13:29
<alkisg>
There might be, but it's not the only one. E.g. we'd want to do the same for /etc/xdg/autostart/jockey-gtk.desktop
13:29
(i.e. delete it or render it ineffective)
13:30
(btw, why aren't we just mounting a tmpfs over nfs with aufs? Don't play well together?)
13:31
(instead of bind-mounts...)
13:37
<johnny>
yes
13:37
or they didn't
13:37
also. aufs isn't in the tkernel
13:37
not in debian's kernel anyways..
13:38
<alkisg>
Got it, ty
13:39
For network manager, maybe we could generate an appropriate /etc/network/interfaces file that corresponds to how we setup ethX, then network-manager wouldn't touch the interface we use...
13:49mikkel has quit IRC
13:49
<johnny>
alkisg, we're still waiting on union mounts ;(
13:49
like 5 years later
13:49
<alkisg>
Heh
13:49
nbd-server -c is the best RAM wise...
13:50
We could use _only_ that for nbd booted clients...
13:50chupacabra has joined #ltsp
13:50
<alkisg>
(==export image read/write, and use a different file to store any writes done there, and automatically discard them when done)
13:51
That saves all the tmpfs/unionfs/aufs etc ram on the client
13:53
E.g. imagine a fat client that wants to apt-get install 1 Gb of stuff temporarily - currently, it would die because of low ram, but with nbd-server -c it would work fine
13:53
(or update-db, or update-xapian-apt-index, etc etc - those won't take up any ram at all)
13:55
<vagrantc>
johnny: aufs is available in debian kernels for 4 years or so now
13:55
er, maybe 3
13:55
it's in the current stable release, anyways.
13:56
but yes, it doesn't play well with NFS
13:56
<johnny>
vagrantc, really? i'm suprised tey accepted that
13:56
since it will never ever ever be accepted into mainline
13:56
that's why fedora won't include it iirc
13:58
<vagrantc>
in debian squeeze/testing it's in the kernel package, in debian lenny/stable it's in an add-on module package.
14:00
<johnny>
yeah, cuz they removed the ability for it to be installed seperately
14:00
i assume after lenny then
14:00
which is why i'm suprised it's in the kernel
14:00
<vagrantc>
there's dkms
14:00
<johnny>
no.. i mean in the code
14:00
for aufs
14:01
<vagrantc>
i just use what i'm given. :)
14:01
<johnny>
they made it so you can't compile against the source tree externally
14:01
maybe they decided to revisit that idea tho
14:01
<vagrantc>
aufs made that?
14:01
<johnny>
i hope so.. cuz that didn't make much sense
14:01
yes
14:01
this is like 2 years ago when they did that hto
14:23alkisg has quit IRC
14:23alkisg has joined #ltsp
14:53[GuS] has quit IRC
14:59kusznir has joined #ltsp
14:59
<kusznir>
Hi all: where do I find docs on how to modify the thin client's installed software load (eg, add packages for use with ltsp-localapps)
15:00
<alkisg>
distro/version?
15:01
<kusznir>
Sorry, ltsp5 / ubuntu 9.10
15:02
<alkisg>
This is for firefox, but it's similar for others: https://help.ubuntu.com/community/UbuntuLTSPKarmicLocalAppsFirefox
15:02pmatulis has quit IRC
15:06irule has joined #ltsp
15:11
<vagrantc>
alkisg: trying to work with getting DNS configuration properly from the initramfs ...
15:13
<alkisg>
vagrantc: don't we have that already? where's the problem?
15:14
<vagrantc>
alkisg: there's a few problems ...
15:14
alkisg: it only gets the DNS_SERVER and SEARCH_DOMAIN variables from lts.conf if the hostname code was triggered...
15:15
<alkisg>
which script it that? (sorry I'm right after a format)
15:15
<vagrantc>
also, it doesn't respect the IPV4DNS0 and IPV4DNS1 variables from /tmp/net-*.conf which could be reasonable defaults for DNS_SERVER
15:16
alkisg: ltsp-trunk/client/initramfs/scripts/nfs-bottom/ltsp
15:16
<alkisg>
Ah, we're getting those from udhcp... ok, let's see that...
15:16
<vagrantc>
i think i've almost got a viable patch, but i wanted an extra pair of eyes
15:16
<alkisg>
pastebin?
15:17
<vagrantc>
hold on a second
15:18
<alkisg>
When is ` if [ "${HOSTNAME}" = "(none)" ]; then` NOT triggered?
15:19
(except for when init-premount/udhcp has alread set that, of course...)
15:19
<vagrantc>
there are some configurations where the hostname is specified via dhcp
15:19
s,specified,set,
15:19
<alkisg>
Ah those are after ipconfig, ok
15:21
Got it, yeah it needs some reordering there
15:22
<vagrantc>
alkisg: http://ltsp.pastebin.com/rPvcLxi0
15:24
<kusznir>
Hmm... my chroot broke.
15:26
<alkisg>
(noting down as I'm reading the diff, HOSTNAME can be specified in lts.conf so we should rename that variable less it's overwritten)
15:27
<kusznir>
sorry, got interrupted, didn't finish my statement.
15:29
I'm trying to install wine on my thin clients (so ltsp-localapps will run it), and when I tried to do this on ubuntu 9.10 (chroot /opt/ltsp/i386; apt-get install wine), it blows up.
15:30
http://paste.pocoo.org/show/252561/
15:30
<alkisg>
kusznir: maybe it needs some things to be mounted, like /proc etc
15:31
<kusznir>
ok, it looks like that may have worked.
15:33
<alkisg>
vagrantc: how about this? `if [ "$(cat /proc/sys/kernel/hostname)" = "(none)" ]; then` ...
15:33
(so that a HOSTNAME from lts.conf isn't overwritten)
15:34
<vagrantc>
hmmm...
15:34
<alkisg>
And then that can be simplified: if [ "$HOSTNAME" = "(none)" ] || [ -z "$HOSTNAME" ]; then ==> if [ -z "$HOSTNAME" ]; then
15:34
<vagrantc>
don't follow you there...
15:35
<alkisg>
Suppose I put in lts.conf this:
15:35
[mac:address]
15:35
HOSTNAME=myclient
15:35
Currently, that will name the client == myclient
15:36
read HOSTNAME < /proc/sys/kernel/hostname ==> it breaks that
15:36
<vagrantc>
ah.
15:36
because lts.conf is before
15:36
<alkisg>
Right, and the same var name is used
15:37
<vagrantc>
could just use a different variable name ...
15:37
<alkisg>
But, reading /proc/sys/kernel/hostname serves a second purpose: if it's not (none), then that means that udhcp took care of everything
15:37
<vagrantc>
ugh...
15:37
<alkisg>
yeah we should move udhcp out of init-premount :)
15:38
Not only it's messy now, it's also too soon, and fast clients some times won't even boot
15:38
<vagrantc>
i'm also in freeze now, so if i want to fix this, it needs to be relatively simple fix...
15:38
<alkisg>
I think that single line would fix it:
15:38
`if [ "$(cat /proc/sys/kernel/hostname)" = "(none)" ]; then`
15:38
(instead of using HOSTNAME there)
15:39
<vagrantc>
that would case: echo "$HOSTNAME" > /proc/sys/kernel/hostname
15:39
to fail when HOSTNAME wasn't specified in lts.conf
15:40
<alkisg>
Why? won't that be executed in that case? HOSTNAME="$HOSTNAME_BASE$HOSTNAME_EXTRA"
15:40
<vagrantc>
there appear to be two lines ... one of which seems to have an extraneous entry...
15:40
ah, sure enough
15:41
<alkisg>
The reordering also makes checking for (none) redudant, i.e.: [ "$HOSTNAME" = "(none)" ] || [ -z "$HOSTNAME" ]; then ==> if [ -z "$HOSTNAME" ]; then
15:41
<vagrantc>
right...
15:41
<alkisg>
So after your pastebin, I think only 3 lines need to be changed ...
15:42
<vagrantc>
in fact, it's even redundant before, no?
15:42
<alkisg>
delete 21, and change 22 and 33
15:42
<vagrantc>
i think i'll just change the variable name on 21 and 22
15:43
<alkisg>
No, it was needed in the case where udhcp was used but no dns name was provided at all
15:43
In that case, net-eth0 woudn't have a "HOSTNAME=" (empty) entry, so HOSTNAME would still be (none)
15:43
<vagrantc>
and that will break how?
15:44
<alkisg>
Now with the reordering it won't break - but checking for (none) was needed before..
15:45
<vagrantc>
i don't see how checking for it as a variable vs. checking for it directly makes a difference with that...
15:45* alkisg curses his english :)
15:46
<alkisg>
Ok what I mean now is this: If you change the variable at 21/22, you may also delete `if [ "$HOSTNAME" = "(none)" ]` below, as it's not needed anymore.
15:46
It won't make sense to leave it, there's no code path that will make HOSTNAME=(none) after the rename
15:47
<vagrantc>
alkisg: http://ltsp.pastebin.com/e8xfG2z7
15:48
<alkisg>
Seems fine to me
15:49
<vagrantc>
makes me a little nervous...
15:50* alkisg can't test, he's still installing stuff :-/
15:51
<alkisg>
vagrantc: doesn't ipconfig put a DNS_SERVER entry at net-eth0.conf, except for IPV4DNS0/IPV4DNS1?
15:52
(I don't have an example handy)
15:52
<vagrantc>
alkisg: i don't think so...
15:52
but i don't really know
15:52
<alkisg>
Ah - better check if DNS_SERVER already has a value
15:53
If udhcp runs, then it puts DNS_SERVER in net-eth0.conf, which all the values it got from dhcp (e.g. more than 2 dns servers)
15:53
<vagrantc>
source code for klibc doesn't return DNS_SERVER
15:53
you and your silly udhcpc :P
15:53
<alkisg>
So udhcpc relies on nfs-bottom/ltsp to properly set those
15:53
Heh
15:53
I'm not even using it!
15:54
<vagrantc>
but wasn't it you who wrote all that?
15:54
<alkisg>
stgraber wrote the first version, and then i rewrote it mainly because I wanted it to properly handle IPAPPEND :)
15:55
hmmm sorry let me check something else...
15:57
All seem ok EXCEPT when udhcp gets more than 2 dns servers
15:57
<vagrantc>
oh, i'll also add DNSDOMAIN
15:58
what's it do then?
15:58
<alkisg>
In that case, it puts a DNS_SERVER entry at net-eth0.conf with all the dns servers
15:58
So, if possible, better check if DNS_SERVER is empty before assigning it the IPV4DNS0/IPV4DNS1 values
16:00
<vagrantc>
yeah, that's in my latest and greatest...
16:00hersonls has quit IRC
16:01
<vagrantc>
alkisg: http://ltsp.pastebin.com/AdgQZ3dQ
16:01
<alkisg>
SEARCH_DOMAIN="${SEARCH_DOMAIN:-$DNSDOMAIN}" would be simpler
16:02
Looks fine
16:02
<vagrantc>
i like the explicitly obviousness
16:03
ok, committing
16:07moose-tek has quit IRC
16:08moose-tek has joined #ltsp
16:12rpm_ has joined #ltsp
16:14
<rpm_>
Is there a way to secure the admin web address of ltsp cluster-control-center?
16:14redbaritone has quit IRC
16:26
<rpm_>
hello? Is anybody home?
16:28
<alkisg>
rpm_: stick around, maybe someone which uses ltsp-cluster will see your question...
16:29
<rpm_>
thanks alkisg
16:29
<alkisg>
Or you can also try the mailing list
16:30
<rpm_>
thanks I'll give that a try while I'
16:30
I'm waiting
16:42vagrantc has quit IRC
16:47vagrantc has joined #ltsp
16:48gentgeen__ has quit IRC
17:15poff has quit IRC
17:17alkisg has quit IRC
17:19alkisg has joined #ltsp
17:24alkisg has quit IRC
17:25irule has quit IRC
17:46vmlintu has quit IRC
17:47jhutchins_lt has quit IRC
17:48shawnp0wers has joined #ltsp
17:53rjune has quit IRC
17:57hiltunen has quit IRC
18:02jhutchins_lt has joined #ltsp
18:08rjune has joined #ltsp
18:08Mip5 has quit IRC
18:20moose-tek has quit IRC
18:21moose-tek has joined #ltsp
18:25jhutchins_lt has quit IRC
18:35efra has quit IRC
18:37ogra has quit IRC
18:37jhutchins_lt has joined #ltsp
18:38ogra has joined #ltsp
18:40staffencasa has quit IRC
18:44shawnp0wers has quit IRC
18:48jhutchins_lt has quit IRC
19:37
<rpm_>
I am getting "error: failed to connect to NBD server" in ubuntu 10.04 is there a solution?
19:38Selveste1___ has joined #ltsp
19:46ogra_cmpc has quit IRC
19:58ogra_cmpc has joined #ltsp
20:04vagrantc has quit IRC
20:05
<johnny>
rpm_, come back when more people are here
20:06
try in about 9 hours
20:06
or later
20:06
i don't have time to help right now. i just want to let you know that this time for this channel is not very populated
20:10gentgeen__ has joined #ltsp
20:10jhutchins_lt has joined #ltsp
20:15jhutchins_lt has quit IRC
20:16sweetpi has joined #ltsp
20:25gentgeen__ has quit IRC
21:02MornignSon has quit IRC
21:04MornignSon has joined #ltsp
21:17BWMerlin has joined #ltsp
21:30Ghidorah has joined #ltsp
21:31
<Ghidorah>
Hello does any one have any experience with WDS and LTSP(PXE) co-existing on the same subnet?
21:47
<johnny>
how about you explain the problem you're having with it
21:47
altho this is not a good time for support for this channel, almost none of the resident experts are about at this time of day
22:09Guerdal82 has joined #ltsp
23:37cellofellow has joined #ltsp
23:38
<cellofellow>
I've got a client, a nice new ASUS machine, that kernel panics not long after boot.
23:44
this was in /var/log/daemon.log
23:44
in.tftpd[6988]: tftp: client does not accept options
23:46
<johnny>
that's not that relevant probably
23:46
i've seen it on my machines before
23:49
<cellofellow>
ok
23:50
i'm thinking it may be a graphics card or NIC issue. It has a GeForce 310M, not sure if that's relavant or not. How do I disable kernel mode setting? It panics the moment it switches from the plain-text 80-column output to the KMS stuff.
23:50
Which may also mean it can't mount the root device.