|08:49||ltspbot has joined #ltsp|
|08:49||sbalneav has joined #ltsp|
|08:50||Gadi has joined #ltsp|
sbalneav: Scotty !!!!!!!!!!!!!!!!!!!!!!!
sbalneav: how cold is it today?
|09:05||F-GT has quit IRC|
what's that in American ?
it's 9F here
only slightly colder there
(-22) degrees Celsius = -7.6 degrees Fahrenheit
According to gooooooooooooooooooooooooooooooooooooooooooogle
I missed the '-'
so, it's a bit colder where you are
Mite nipply, yes. :)
Marlon's in Manaus, it's 30C there right now
|09:09||F-GT has joined #ltsp|
|09:09||fbzdvbylp has joined #ltsp|
Betcha they have Brahma there :)
|09:15||alkisg has quit IRC|
|09:15||alkisg has joined #ltsp|
jammcq: I am doing LTSP over AC power.
careful you don't fry yourself
it's pretty decent, 200mbits-ish
|09:20||Pod3 has left #ltsp|
|09:22||Pod21 has joined #ltsp|
|09:22||Pod21 has left #ltsp|
|09:23||BF-Pod has joined #ltsp|
|09:23||BF-Pod has left #ltsp|
|09:24||BF-Pod has joined #ltsp|
ah, just checking I can speak now
unfortunately, due to the CTCP spam that freenode is having
I added the "unidentified users cannot speak" flag to this channel
btw... hi jammcq, etyack, sbalneav
|09:26||wimiko has joined #ltsp|
|09:27||ogblkflatnh has joined #ltsp|
case in point?
Lumiere: seems rather unfriendly to first timers coming here for help
they're not smart bots
jammcq: with any luck
in 2 days we won't need it with the move to the new ircd
but for now
k, I guess
they're spamming trojan horses to people
which turns them into spammers
jammcq: every channels' been getting pretty hammered. :-/
I dont' recall #ltsp getting spams
but, I haven't been around much lately
it got em overnight
about 10 times
in 4 hours
it's been a problem all week
i've seen it even longer, almost the entire month
not always in #ltsp
etyack: yea, but #ltsp got targeted yesterday
I passed on the suggestion to add +r
which blocks them from joining
is is possible to mount the root path on a client in writable mode?
Using what, NFS root?
With nbd root, it'd be impossible
with nfs root, it'd be possible, but if you've got more than one client, it'd be a disaster.
|09:49||guaka has joined #ltsp|
since multiple clients would be tromping all over each other.
|09:50||hersonls has quit IRC|
|09:50||uvtiqq has joined #ltsp|
hi, some client machines aren't booting anymore, they get to "Loading initrd.img.........ready" (with some more dots of course)
|09:51||Faithful has quit IRC|
guaka: either a networking, dhcp, or tftp problem
|09:51||GGD_ has joined #ltsp|
I'm trying to compile the nvidia driver, which I thought may be necessary to run on the client
|09:51||Faithful has joined #ltsp|
sbalneav, dnsmasq seems to be running fine, I do find "tftp: client does not accept options" in daemon.log but that doesn't seem to be a problem (according to some posts I found through google)
the network is working pretty well still, but if it's some specific weird network problem, how to look for problems?
Don't know. I don't know anything about dnsmasq
I always use stock tftp-hpa and dhcp3-server
|09:54||Appiah has quit IRC|
sbalneav, dnsmasq is a DHCP server that also does DNS caching, which is convenient if you want to have names for your machines
I have names for my machines too. I just use bind.
ah, dnsmasq usually requires less attention than bind..
FYI -- anyone who can't talk will need to register. We're not ignoring you. See http://bit.ly/1MtZU for more information
but DHCP seems to be allright
guaka: Well, like I say, I'm not familiar with dnsmasq.
guaka: try to download the newest pxelinux.0 from debian
I've seen a lot of newer PCs having problem with the version that Ubuntu ships...
Just copy pxelinux.0 from that package to /var/lib/tftpboot/ltsp/i386/pxelinux.0
A bug has been filed for Ubuntu to update to a newer version, but it hasn't been solved yet.... :-/
ah, ok, I'll try that... in fact, the problem only appeared recently, I don't think anything has changed, and only on one specific type of machine, it seems now (I had to wait till someone left)
|10:00||grantk has joined #ltsp|
I've seen e.g. PCs booting once out of 10 with syslinux 2:3.63+dfsg-2ubuntu3, and 10 out of 10 with the debian version.
|10:08||GGD has quit IRC|
hmm, with the debian pxelinux.0 it gets stuck right after "Trying to load: pxelinux.cfg/default", that's slightly worse than getting stuck after "Loading initrd.....ready" :/
or do I need to kick or push something before it likes another pxelinux.0?
|10:16||zky has joined #ltsp|
|10:23||scottmaccal__ has joined #ltsp|
|10:25||scottmaccal_ has quit IRC|
|10:25||shawnp0wers has quit IRC|
|10:28||gnunux has quit IRC|
|10:30||bschfuezy has joined #ltsp|
|10:33||hersonls has joined #ltsp|
|10:34||highvolt1ge has left #ltsp|
|10:34||highvoltage has joined #ltsp|
|10:35||hersonls has quit IRC|
|10:36||hersonls has joined #ltsp|
|10:39||hersonls has joined #ltsp|
|10:40||hersonls has quit IRC|
|10:41||hersonls has joined #ltsp|
|10:42||hersonls has quit IRC|
|10:44||hersonls has joined #ltsp|
|10:49||goqxrbd has joined #ltsp|
|10:52||shawnp0wers has joined #ltsp|
If I enable Xinerama, should there be a mention of it in /var/log/Xorg.7.log ?
|10:54||alkisg has quit IRC|
|11:02||scottmaccal_ has joined #ltsp|
|11:03||scottmaccal__ has quit IRC|
I hope I do not repeat myself here, but I believe registration blocked me before. I was looking for documentation on ldm load balancing. I noticed in the LTSPManual.pdf that section 21.5 is missing a reference to somewhere, it just shows "See for details.". I think it should say see Chapter 23 but I am not sure.
Anyway I wanted to ask if anyone has seen any other documentation on ldm load balancing?
|11:15||spectra has joined #ltsp|
|11:19||bobby_C has joined #ltsp|
|11:20||etyack has quit IRC|
|11:30||guaka has quit IRC|
|11:45||bobby_C has quit IRC|
|11:59||BF-Pod has left #ltsp|
|12:16||appiah has joined #ltsp|
|12:17||etyack has joined #ltsp|
|12:20||Kicer86 has joined #ltsp|
|12:33||alkisg has joined #ltsp|
|12:34||alexqwesa has joined #ltsp|
|12:40||rjune has quit IRC|
|12:42||MaRX-Mod1 has quit IRC|
|12:42||MaRX-Mode has joined #ltsp|
|12:42||MaRX-Mode has joined #ltsp|
|12:44||ffl has joined #ltsp|
|12:45||artista_frustrad has joined #ltsp|
|12:55||rjune has joined #ltsp|
Gadi, You know who resara is, yes?
rjune: they are a company not a person, and sure do
|12:58||* Gadi expects them to be in the channel, too|
|12:58||* Gadi looks around the room|
|12:58||vagrantc has joined #ltsp|
are you a scotticicle today?
It's -7.9 F here today
Or -22 in proper units :) So yeah, it's a mite nipply here.
|13:04||staffencasa has quit IRC|
|13:04||staffencasa has joined #ltsp|
|13:04||* Gadi shudders when scotty uses words like nipply|
|13:05||* jammcq just giggles|
u and ur giggling
heh heh, he said nipply, heh heh
now i have pictures in my head for my portland flight tomorrow
thanks Gadi :P
oh sure, blame me
ogra: it's not exactly warm here, but it's not really cold :)
yeah, i saw something around 10°C
after a week of -10 to -15°C here that will be like spring for me :)
sbalneav: I told you that r124 of ltspfs-trunk works well, right?
No, but good to hear. So that's two big bugs off the list.
vagrantc: did you fix your localdev stuff?
1) "group ownership issue" 2) "No Auth Localhost"
ah - the group ownership thing is upstream and tested?
|13:11||* Gadi hasn't gotten there yet|
|13:11||artista_frustrad has quit IRC|
vagrantc: :( it's almost *got* to be that something is not writable somewhere
Gadi: can't imagine where... the same exact code works on debian lenny
Gadi: so it's got to be behavior changed in something outside of debian
Gadi: er, outside of ltspfs
Gadi: my main guess is that the udev rules fire too early
Gadi: before it's made writeable
i guess i could write some hooks in the initramfs to mount stuff earlier
that would make sense - but only if it works with, say, usb sticks once booted
in fact, maybe mount everything in the initramfs
Gadi: well, the usb sticks work once booted
Gadi: cdpinger is still borked
i still haven't had a chance to test with real hardware, either
well, thats a good theory then
vagrantc, did debian switch from hal to devkit already ?
ogra: heck if i know
i think there changed quite a lot with that switch
95% of the stuff is dirctly handled in udev or via libudev now
vagrantc: got a sec?
|13:34||otavio has quit IRC|
|13:37||etyack has quit IRC|
|13:43||cliebow has quit IRC|
|13:49||etyack has joined #ltsp|
|13:58||shawnp0wers has quit IRC|
|14:01||mfdutra has joined #ltsp|
|14:02||shogunx has quit IRC|
|14:04||bobby_C has joined #ltsp|
Gadi: sort of
|14:08||otavio has joined #ltsp|
|14:11||waranha has quit IRC|
|14:11||waranha has joined #ltsp|
|14:14||waranha has quit IRC|
|14:14||waranha has joined #ltsp|
|14:14||waranha has left #ltsp|
|14:15||wagner has joined #ltsp|
|14:16||ogra has quit IRC|
|14:16||ogra_ has joined #ltsp|
|14:17||ogra_ has left #ltsp|
|14:17||ogra has joined #ltsp|
|14:23||Kicer86 has quit IRC|
vagrantc: I think the "pgrep" line in the call_cdpinger function in ltspfs_entry needs a "-x", so that if $1=cdrom, it does not match "cdrom1", etc
similar to what is done in ltsp-init-common
there's probably a significant amount of refactoring that could be done there anyways...
|14:30||alexqwesa has quit IRC|
well, short of replacing some calls to "basename" and "awk", I think ltspfs_entry is actually not too terrible
now, that we got a lot of my older ugliness out of there
|14:31||alexqwesa has joined #ltsp|
Gadi: yeah, i'll add your commit with cdpinger.
vagrantc: Need any help from a c programmer on cdpinger?
sbalneav: though i want to get another upload of ltsp in to debian first.
What's failing on it for you?
|14:40||* Gadi is going to commit a cleanup of basename...|
sbalneav: it just plain don't work.
sbalneav: which isn't much to go on, i know... :)
sbalneav: probably need to figure out some debugging hooks...
heh, does it segfault, or just quietly sit there doing nothing?
sbalneav: it runs fine, it's not not catching an insertion/removal events
or doesn't seem to be
revert to the python version ;)
|14:41||* sbalneav slaps ogra with a wet noodle.|
or rewrite in perl
back in the good ol days, when we had unlimited ram to burn
|14:42||* ogra pulls up some ketchup|
now with the economic downturn, we have to resort to C
well, C means you have to listen to the right kernel ioctl events
sbalneav: it's something different between debian lenny/stable and debian squeeze/testing|sid/unstable
sbalneav: lenny works fine
my suspicion is changes in udev handling or some such
or the kernel
udev and kernel changes go hand in hand
sbalneav: i think it's been broken for a long time though... since july at least
If you "start it manually", does it work?
sbalneav: so figuring out some improved debugging hooks/options/etc would probably help...
I can add some syslogging to it.
Sweet, sweet sysloggy goodness.
Syslog and lobsters.
That'd make a good name for a blog!
|14:55||wagner has quit IRC|
|14:55||scottmaccal_ has quit IRC|
there's no CD in the drive...
so starting it manually probably won't make much difference...
sbalneav: well... after starting cdpinger manually... it noticed the insertion event... but didn't notice the removal event
|15:05||hersonls has quit IRC|
maybe i should run the code from ltspfs-trunk ... just in case those code cleanups helped anything
|15:11||Wastrel_ has quit IRC|
|15:11||elias_a has quit IRC|
|15:20||artista_frustrad has joined #ltsp|
vagrantc: just commited a few more cleanups
|15:20||* Gadi stops mucking now...|
|15:21||Gadi has left #ltsp|
heh. i just noticed this patch in my debian dir... http://ltsp.pastebin.com/m2529e366
that should probably go upstream
been there since ltsp 5.1.90, i'm guessing
|15:25||GGD_ has quit IRC|
|15:29||grantk has left #ltsp|
|15:32||otavio_ has joined #ltsp|
|15:33||mikkel has quit IRC|
|15:35||lucascoala has joined #ltsp|
|15:36||etyack has quit IRC|
FYI -- anyone who can't talk will need to register. We're not ignoring you. See http://bit.ly/1MtZU for more information
|15:38||vvinet has quit IRC|
|15:40||artista-frustrad has joined #ltsp|
|15:41||artista_frustrad has quit IRC|
|15:44||bobby_C has quit IRC|
|15:44||otavio has quit IRC|
|15:56||mgariepy has quit IRC|
|15:59||artista_frustrad has joined #ltsp|
|16:01||artista-frustrad has quit IRC|
|16:18||artista_frustrad has quit IRC|
|16:18||shogunx has joined #ltsp|
|16:23||mfdutra has quit IRC|
|16:29||jhutchins_lt has quit IRC|
|16:42||ogra has quit IRC|
|16:47||jhutchins_lt has joined #ltsp|
|16:47||alkisg has quit IRC|
|16:56||jammcq has quit IRC|
|17:03||mfdutra has joined #ltsp|
hey guys. where is the hostname set in the client boot process
I'm grepping the scripts and can't find it
possibly from the initramfs
depends on if dhcp sends a hostname...
I have "ltsp1657" here
is there a logics?
yes, that's in the initramfs for sure
dhcp doesn't send it, but I wanna the hostname to set from the IP PTR record in dns
I'll write a script for that
yeah, that's the default for when dhcp doesn't send a client-hostname ...
ltspX where X is some number based on the ip address
alternately can base it on mac address
yep. we have our DNS all set here
it's the IP addr in hexa notation
ex: 192.168.17.77 -> c0a8114d
i don't quite know how to get dhcp to send the hostname...
|17:08||jhutchins_lt has quit IRC|
it's important for me to have a FQDN as hostname, because we have about 70 VLANs
and I have subdomains for each one
|17:10||* vagrantc runs a network with about 15-20 terminals|
a little more here :)
|17:11||bengoa has joined #ltsp|
probably wouldn't be hard to write code that just looks in DNS
sets it to that, if anything is found... of course, DNS timeouts would stall boot
5-10 lines in python
|17:14||jhutchins_lt has joined #ltsp|
|17:17||bengoa has quit IRC|
|17:22||vvinet has joined #ltsp|
|17:27||jhutchins_lt has quit IRC|
what are you guys babbling about?
(oh I should've scrolled down first :p)
hmmm... i should try testing with ext2 again ... squashfs appears to hang my thin client
hmmm... getltscfg:264 blocked for more than 120 seconds....
i'm going to say this is borked.
|17:44||GGD has joined #ltsp|
|17:48||jhutchins_lt has joined #ltsp|
|17:53||GGD has quit IRC|
|17:58||wimiko has quit IRC|
|18:00||jhutchins_lt has joined #ltsp|
|18:06||jhutchins_lt has quit IRC|
|18:32||pmatulis has quit IRC|
|18:32||Faithful has quit IRC|
|18:33||Faithful has joined #ltsp|
|18:42||kaos01 has joined #ltsp|
|18:46||otavio_ has quit IRC|
|18:57||pmatulis has joined #ltsp|
|19:02||mfdutra has quit IRC|
|19:05||kaos01 has quit IRC|
|19:23||Faithful has quit IRC|
|19:24||Faithful has joined #ltsp|
|19:27||GGD has joined #ltsp|
|20:24||try2free has joined #ltsp|
|20:31||artista_frustrad has joined #ltsp|
stgraber: seems like it would be a good idea to use /var/run/sendsigs.omit, /lib/init/rw/sendsigs.omit and/or /lib/init/rw/sensigs.omit.d/* for making sure nbd doesn't hang on reboot
finally got an upload in to debian with some patches since 5.1.99 ...
and now i'm doing a bunch of NBD testing
i still am stumped with why X won't start with NBD+squashfs+aufs
(the method most similar to ubuntu)
it works fine with NBD+tmpfs bind mounts
i also confirmed that X won't start with NBD+ext2+aufs
my favorite combination, in theory, would be NBD+ext2+aufs
no need for ltsp-update-image.
but any time i add aufs into the mix, it hangs.
well, X simply doesn't start...
the machine actually continues to work... i can switch vt's to one with a shell...
Somehing whacky with aufs?
X simply doesn't display the greeter.
that's my guess...
|20:50||Selveste1___ has quit IRC|
|20:50||evilx has quit IRC|
|20:50||jhutchins has quit IRC|
|20:50||moquist has quit IRC|
|20:50||moquist has joined #ltsp|
|20:50||jhutchins has joined #ltsp|
|20:50||Selveste1___ has joined #ltsp|
|20:50||evilx has joined #LTSP|
|20:58||shogunx has quit IRC|
|20:59||shogunx has joined #ltsp|
|21:06||Faithful has quit IRC|
|21:07||Faithful has joined #ltsp|
|21:14||ball has joined #ltsp|
vagrantc: I'm gonna head to bed early tonight. I'll spend some time adding syslog support to cdpinger tomorrow.
|21:25||squarepeg has joined #ltsp|
|21:26||squarepeg has quit IRC|
sbalneav: thanks :)
|21:27||cliebow has joined #ltsp|
|21:35||cliebow has quit IRC|
|21:35||Faithful has quit IRC|
|21:35||artista_frustrad has quit IRC|
|21:36||Faithful1 has joined #ltsp|
|21:52||pmatulis has quit IRC|
|21:56||vvinet has quit IRC|
what on earth is this: http://bugs.debian.org/563958
Is it a trout in a dustbin lid?
|22:06||artista-frustrad has joined #ltsp|
i don't really get what a "Modular TFTP/Fuse supplicant
but they actually have a package regarding ltsp...
tftp via fuse?
something crazy, yes.
|22:24||artista-frustrad has quit IRC|
|22:29||Egyptian[Home] has quit IRC|
|22:30||Egyptian[Home] has joined #ltsp|
|22:33||Faithful1 has quit IRC|
|22:40||shogunx has quit IRC|
|22:44||GGD has quit IRC|
stgraber: nbd-proxy after ltsp-5.1.99 is totally broken for me...
stgraber: 5.1.99 works fine, though
|23:38||alkisg has joined #ltsp|
|23:43||vmlintu has quit IRC|