IRC Logs

Channel log from 16 March 2008 (all times are UTC)

00:20	cdealer has quit IRC
00:37	Egyptian[Home] has quit IRC
04:26	joebaker has joined #ltsp
04:46	mikkel has joined #ltsp
05:08	captain_magnus has quit IRC
05:12	captain_magnus has joined #ltsp
06:18	basanta has joined #ltsp
06:21	K_O-Gnom has joined #ltsp
06:32	indradg has quit IRC
06:51	basanta has quit IRC
06:52	indradg has joined #ltsp
06:59	abadger1999 has quit IRC
07:36	alekibango has quit IRC
07:37	alekibango has joined #ltsp
07:38	daya has joined #ltsp
07:39	<daya> Hi, all
07:39	ldm in lenny is not working any idea,
07:44	nobios_ has joined #ltsp
07:50	daya has quit IRC
08:00	nobios has quit IRC
08:32	nobios_ has quit IRC
08:36	ogra_ has quit IRC
08:36	ogra_ has joined #ltsp
08:37	ogra_cmpc has quit IRC
08:37	ogra_cmpc has joined #ltsp
08:40	abadger1999 has joined #ltsp
08:45	K_O-Gnom has quit IRC
08:47	vagrantc has joined #ltsp
09:02	daduke_ has joined #ltsp
09:55	<ogra_cmpc> oooof
09:55	* ogra_cmpc just finished working through a >2500 mail backlog
09:57	<ogra_cmpc> vagrantc, http://code.google.com/p/compcache/
09:58	* laga fires of another 2500 mails to ogra_cmpc
09:58	<vagrantc> ogra_cmpc: crazy
09:58	<ogra_cmpc> it's claimed that it has nearly a 50% compression rate ... so having a swapfile in a tmpfs would actually gain us extra ram
09:58	<vagrantc> ogra_cmpc: kernel module?
09:58	<ogra_cmpc> i'll examine it in intrepid for subnotebook stuff
09:58	no idea, i havent looked inside the tarball yet
09:59	just stumbled across
09:59	if that works we could even support 24M clients out of the box without swapfiles :)
10:00	s/swapfiles/swapfiles from initramfs/
10:00	<vagrantc> getting below 32M without network swap would be key.
10:01	* ogra_cmpc cant stop being impressed by the SDHC card he bought ...
10:01	<vagrantc> SDHC ?
10:01	<ogra_cmpc> SD card with new hipeed technology
10:01	<cyberorg> great, i am not able to get nbdwap working properly, i get "Device or resource busy" on swapon /dev/nbd1
10:01	<ogra_cmpc> it getsw to transfer rates of about 20M/s
10:02	which is faster than the 18M my laptop disk does :)
10:03	or any of my usb keys
10:05	<vagrantc> ogra_cmpc: that compcache thing is slick ... it's implemented as a block device
10:05	<ogra_cmpc> cool
10:06	i'll set up a spec for next UDS :)
10:07	abadger1999 has quit IRC
10:11	<vagrantc> ogra_cmpc: you install the module and just do: swapon /dev/compcache0
10:11	<ogra_cmpc> wow
10:11	<vagrantc> it may just eat whatever ram it uses though ... not dynamically allocating it
10:11	<ogra_cmpc> i'm curious what our kernel team will say
10:22	tux_440volt has joined #ltsp
10:26	steph_ has joined #ltsp
10:31	<steph_> ogra_cmpc: would it be possible to use ltsp in a virtual machine (Xen, vbox or something else). I would like to create a production and a test environnement.
10:31	<laga> steph_: yes
10:32	<ogra_cmpc> i'm using for tests it in virtualbox
10:32	*using it for tests
10:32	* vagrantc uses virtualbox daily for ltsp development
10:33	* vagrantc used to use qemu a fair bit for ltsp development
10:33	* cyberorg too
10:33	<steph_> What about Xen (pro/cons)
10:33	<rjune__> how do I tell irc I'm not away?
10:33	/away says I'm away
10:34	<vagrantc> steph_: i haven't tried xen as a thin-client, but the main LTSP deployment i admin uses a xen ltsp server.
10:34	rjune__: with irssi ... "/away foo message" and "/away" says no longer away
10:35	<rjune__> vagrantc: thanks.
10:35	for some reason /away is registering as still away
10:36	<steph_> thanks
10:43	<cyberorg> vagrantc, ogra_cmpc we've sent two patches on the ML, one today for plugin and one earlier for ldm, how does it get committed in trunk?
10:44	<vagrantc> cyberorg: well, seems like you're the (open)suse developer ... should give you commit access
10:45	ogra_cmpc: regarding the ldm-arbitrary-xserver-opts ... i don't particularily care about which options are supported, but do you like the idea of moving code out of LDM itself and into the screen script?
10:46	<cyberorg> vagrantc, i'll be sorting out plugin bits in more detail, for now my main concern was to get everything working, i guess we are almost at par with ubuntu implementation feature wise :)
10:47	we get TC to boot up even on 128M ram without swap, if i could get swap working, may be it would work with 62M too
10:52	<steph_> I read there's also KVM (kernel virtual machine). I would like to give a try, but I want to choose the more stable solution (Xen, VirtualBox, KVN, VWare, etc.) Any advices?
10:53	<vagrantc> steph_: KVM pretty much requires hardware support in your CPU
10:54	steph_: virtualbox is most useful for testing- it's (exclusivevly?) GUI
10:55	steph_: but i wouldn't run virtualbox as a server ...
10:55	steph_: vmware is propreitary. blech :)
10:55	steph_: i don't know how Xen would handle as a client, but works great as a server.
10:56	<steph_> vagrantc: what do you mean about "hardware support"?
10:56	<vagrantc> steph_: qemu feels more flexible than virtualbox to me.
10:56	steph_: like, if you don't have the right CPU, it won't work
10:56	<steph_> oh, ok
10:56	<vagrantc> or rather, it'll fall back to using regular qemu
10:57	and then you may as well just use qemu
11:08	ogra_cmpc has quit IRC
11:13	<warren> xen can do either full virt or paravirt
11:14	I'm not sure if paravirt can do PXe
11:14	xen fullvirt uses qemu
11:15	<laga> vagrantc: virtualbox can also be run from the command line, FYI
11:17	<steph_> vagrantc: I read that KVM is qemu plus some great kernel level stuff. How can I know if my CPU can work with KVM?
11:22	*quote: "Ubuntu Linux will use KVM as its primary virtualization software" - Cannonical...
11:28	steph_ has left #ltsp
11:35	ogra_cmpc has joined #ltsp
11:59	joebaker has quit IRC
12:08	J45p3r has joined #ltsp
12:08	spectra has joined #ltsp
12:17	deavid has joined #ltsp
12:19	deavid has quit IRC
12:27	Egyptian[Home] has joined #ltsp
12:29	mikkel has quit IRC
12:39	deavid has joined #ltsp
12:58	deavid has quit IRC
12:58	zodman has joined #ltsp
13:30	zodman has quit IRC
13:36	K_O-Gnom has joined #ltsp
13:48	abadger1999 has joined #ltsp
13:54	daduke_ has quit IRC
13:58	mccann has joined #ltsp
14:02	cyberorg has quit IRC
14:03	cyberorg has joined #ltsp
14:10	daduke_ has joined #ltsp
14:21	mccann has quit IRC
14:22	mccann has joined #ltsp
14:25	oh207 has quit IRC
14:25	oh207 has joined #ltsp
14:36	<johnny> ugghz
14:36	genkernel has an exit 255 :(
14:37	# End of execution for now
14:37	exit 255
14:39	deavid has joined #ltsp
14:40	<vagrantc> johnny: someday you'll look back and say ... "i'm glad i got that out of the way"
14:40	<johnny> ?
14:41	<vagrantc> johnny: when you're past the initial implementation and on to maintenance and feature development :)
14:41	<johnny> i'm getting closer!
14:41	<vagrantc> sounds like it :)
14:42	<johnny> and.. i will be able to run it with i386 and virtualbox :)
14:42	i think it was worth the hour to figure that out up front..
14:42	since i leanred something i can use elsewhere
14:42	<vagrantc> with my last virtualbox upgrade, i started having problems with older versions of pxelinux :(
14:43	<johnny> actually, it's cuz of reading warren's code
14:43	i saw him using setarch, so i went and looked it up
14:43	<vagrantc> although i'm not entirely sure it was the last upgrade, as it was working even with the new version at least a few times.
14:43	<johnny> i've gotten two fixes added to our chroot creator script in the process :)
14:44	this exit255 thing i was actually stuck on for an hour or two before
14:44	i kept getting this after building the kernel
14:44	error: LTSP client installation ended abnormally
14:45	with not much info :(
14:45	looked all around the ltsp code, which is a benefit somewhat
14:45	saw the trap on_exit
14:46	looked in gekernel.. and bam
14:46	exit 255
14:47	funny comment in the script tho
14:47	## unionfs is deprecated, use aufs
14:47	but there is no --aufs flag :(
14:49	stgraber has quit IRC
14:52	stgraber has joined #ltsp
14:53	<johnny> and this exit 255 only exists in the forked genkernel we're using, the one that supports unionfs
14:53	the other one doesn't have it..
14:54	time to find the author..
14:54	and of course.. he is out today :(
14:56	tux_440volt has quit IRC
15:02	stgraber has quit IRC
15:05	bobby_C has joined #ltsp
15:25	abadger1999 has quit IRC
15:41	daduke_ has quit IRC
15:51	abadger1999 has joined #ltsp
15:53	GodFather has joined #ltsp
16:12	nobios has joined #ltsp
16:25	K_O-Gnom has quit IRC
16:33	<vagrantc> ogra_cmpc: i just realized ... if it's only a small number of lts.conf options ... you can pass them on the boot prompt
16:37	cyberorg has quit IRC
16:39	GodFather has quit IRC
16:58	cyberorg has joined #ltsp
17:00	deavid has quit IRC
17:16	nobios has quit IRC
17:17	abadger1999 has quit IRC
17:19	abadger1999 has joined #ltsp
17:48	bobby_C has quit IRC
17:50	mistik1_ has joined #ltsp
17:51	mistik1 has quit IRC
17:51	mistik1_ is now known as mistik1
17:57	abadger1999 has quit IRC
18:05	Faithful has joined #ltsp
18:21	J45p3r has left #ltsp
19:03	rcy` has joined #ltsp
19:08	RyanRyan62 has joined #ltsp
19:21	RyanRyan52 has quit IRC
19:34	RyanRyan52 has joined #ltsp
19:36	RyanRyan62 has quit IRC
19:42	MacIver has quit IRC
19:47	MacIver has joined #ltsp
19:52	RyanRyan52 has quit IRC
19:52	stgraber has joined #ltsp
19:54	Egyptian[Home] has quit IRC
19:55	Egyptian[Home] has joined #ltsp
20:02	mcfloppy is now known as kaffee
20:02	kaffee is now known as mcfloppy
20:22	RyanRyan52 has joined #ltsp
20:26	Ryan52 has quit IRC
20:27	Ryan52 has joined #ltsp
20:34	lucas__ has joined #ltsp
20:37	rcy` has quit IRC
20:44	gonzaloaf has quit IRC
21:13	abadger1999 has joined #ltsp
21:51	lucas__ has quit IRC
21:59	subsume has joined #ltsp
22:00	subsume has left #ltsp
22:00	subsume has joined #ltsp
22:00	<subsume> ARP timeout.... what could that be?
22:02	<MacIver> ever see alf?
22:03	<subsume> You talkin' to me?
22:03	=)
22:03	Yes.
22:05	<warren> vagrantc, +1 (moving X server config out of ldm and into the screen script)
22:05	vagrantc, makes it easier to unify it with xdmcp and other screen scripts as well.
22:08	<subsume> I've got a thin client that is clearly seeing the server on boot (changes to DHCP settings effect the startup client IP, Mask, DHCP, gateway).... yet I'm getting ARP timeouts
22:08	I've got lots of other non-thin-clients using the server as a DHCP and they can access network / internet fine
22:10	<MacIver> is the pxe booting?
22:11	<subsume> I did some playing around and I got the ARP message to go away but now it says "TFTP open timeout"
22:12	http://dpaste.com/39755/ <---- my dhcp conf
22:12	<MacIver> is the ip it's trying to get to running the tftp server?
22:12	<subsume> Now its simply hanging on TFTP "TFTP....."
22:12	yes, the DHCP IP is correct
22:12	192.168.0.10 is the server
22:12	and all other machines that use this as DHCP work.
22:13	<MacIver> you could try tftp from another machine on the network
22:14	see if it can connect to the server
22:15	<vagrantc> warren: glad you like the idea.
22:17	<subsume> MacIver: I apt-got tftp but its pretty ambiguous as to whether its actually connecting
22:18	ping works but tftp <ip> does not much
22:18	<MacIver> subsume: you can try to actually download the ltsp/i386/pxelinux.0
22:19	<subsume> ok
22:21	MacIver: even that command just hangs
22:21	transfer timed out, eventually
22:21	Do I need to be resetting TFTP as well as DHCP when making changes?
22:22	<MacIver> subsume: iirc tftp server should be started by inetd on the server machine
22:22	<subsume> MacIver: its started. I wonder if it should be restarted when I change dhcp.
22:23	<MacIver> subsume: i don't see why it would
22:23	<subsume> so timeout. but I get the timeout message no matter what IP I try as server
22:23	spectra has quit IRC
22:23	<MacIver> probably because none of those servers are running tftp?
22:24	<warren> vagrantc, or rather, it might be acceptable to be in ldm only if it shares the config with other screen scripts
22:24	vagrantc, but that seems more difficult
22:24	<MacIver> how do you know tftp server is going?
22:24	<warren> If in doubt, run wireshark on the server and watch for TFTP traffic
22:25	<subsume> MacIver: 192.168.0.10 is running TFTP.... I know this because.... I started it myself?
22:25	<warren> run wireshark and see what the client is asking for exactly
22:25	I figured out my clients were asking for the wrong arch this way
22:26	<MacIver> subsume: started it yourself...hmmm, like i said, normally on distros i've seen it automagically started by inetd
22:26	<subsume> MacIver: I restarted it myself is what I mean, sorry.
22:26	warren: ok...wireshark installing
22:27	<MacIver> inetd 'starts' it whenever someone connects :-p
22:28	<warren> subsume, setup a filter before you start logging or you'll have too muc
22:28	too much
22:31	<subsume> warren: is wireshark graphical?
22:31	<warren> subsume, optionally graphical
22:31	<subsume> ok
22:32	warren: could you recommend a filter?
22:32	<warren> sorry, nee dto sleep
22:33	losing consciousness
22:33	<subsume> k
22:33	thanks
22:50	mccann has quit IRC
22:58	<subsume> Nothing.
22:58	<MacIver> so no response from the server?
22:59	<subsume> I'm not getting anything with wireshark
23:00	I scoured logs. Tres difficile.
23:00	<MacIver> using the gui?
23:00	<subsume> MacIver: Nah... headless server
23:00	tshark
23:00	<MacIver> subsume: you could run it ssh -Y remotely :-D
23:01	<subsume> wow.
23:02	I'm on macosx but let me try that from another client
23:02	<MacIver> why not from macosx?
23:02	you can do it from there
23:02	<subsume> got errors.
23:03	getting errors from client too
23:03	X window system error
23:07	A less ambiguous tftp client would be ideal =)
23:07	this tftp thing sucks bad.
23:13	http://dpaste.com/39762/
23:13	MacIver: in case the actual thin client output sparks something....
23:15	When I comment out the next server option in DHCP, I get TFTP open timeout.
23:15	I am not sure which error is preferable.
23:23	* johnny builds an initramfs
23:34	<subsume> MacIver: I'm able to see other client requests for ltsp/i386/pxelinux.0 using wireshark
23:34	MacIver: but they time out
23:35	vagrantc has quit IRC
23:36	daya has joined #ltsp
23:44	<daya> Hi, all is lenny uses ldm or sdm,
23:45	I am not prompted to GUI

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

Channel log from 16 March 2008 (all times are UTC)