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


Channel log from 5 May 2010   (all times are UTC)

00:01johnny has left #ltsp
00:07johnny has joined #ltsp
00:12highvoltage has quit IRC
00:17highvoltage has joined #ltsp
00:18highvoltage has quit IRC
00:22highvoltage has joined #ltsp
00:27highvoltage has quit IRC
00:27vagrantc has quit IRC
00:33highvoltage has joined #ltsp
00:34highvoltage has quit IRC
00:59DeathLx has quit IRC
01:16highvoltage has joined #ltsp
01:17highvoltage has joined #ltsp
01:19try2free has joined #ltsp
01:21try2free has left #ltsp
01:40sepski is now known as sep
01:59gnunux has joined #ltsp
02:00
<gnunux>
hi
02:15highvoltage has quit IRC
02:15highvoltage has joined #ltsp
02:18ogra has quit IRC
02:22ogra has joined #ltsp
02:29yanu has quit IRC
02:31highvoltage has quit IRC
02:32highvoltage has joined #ltsp
02:32yanu_ has joined #ltsp
02:35dobber has joined #ltsp
02:38yanu_ has quit IRC
02:47highvoltage has quit IRC
02:51highvoltage has joined #ltsp
02:52highvoltage has quit IRC
02:57highvoltage has joined #ltsp
03:23highvoltage has quit IRC
03:24highvoltage has joined #ltsp
03:29yanu_ has joined #ltsp
03:31highvoltage has quit IRC
03:37highvoltage has joined #ltsp
03:47map7 has quit IRC
03:47lfaraone has quit IRC
03:47vlt has quit IRC
03:48map7 has joined #ltsp
03:48lfaraone has joined #ltsp
03:48vlt has joined #ltsp
03:52highvoltage has joined #ltsp
03:53highvoltage has joined #ltsp
04:00highvoltage has quit IRC
04:21highvoltage has joined #ltsp
04:22highvoltage has quit IRC
04:22highvoltage has joined #ltsp
04:30highvoltage has quit IRC
04:40highvoltage has joined #ltsp
05:07highvoltage has joined #ltsp
05:13alkisg has joined #ltsp
05:26mikkel has joined #ltsp
05:33highvoltage has joined #ltsp
05:36highvoltage has quit IRC
05:44highvoltage has joined #ltsp
05:45highvoltage has quit IRC
05:56highvoltage has joined #ltsp
05:56hersonls has joined #ltsp
05:59highvoltage has joined #ltsp
06:06Anon9155 has joined #ltsp
06:13Anon9155 has quit IRC
06:18pmatulis has joined #ltsp
06:21jammcq has quit IRC
06:49bobby_C has joined #ltsp
06:53frederickjh has joined #ltsp
06:57hersonls has quit IRC
06:57hersonls has joined #ltsp
07:30cliebow has joined #ltsp
07:44daduke has joined #ltsp
07:51profesor_ has joined #ltsp
07:53wsuetholz has quit IRC
07:56cliebow has quit IRC
07:57cliebow has joined #ltsp
08:00mikkel has quit IRC
08:04reynolds has joined #ltsp
08:04
<reynolds>
can anyone help me with my ubuntu upgrade? My clients wont boot
08:06
<alkisg>
Did you create a new chroot?
08:06
(I assume you mean os upgrade, right?)
08:06
<reynolds>
no do I need to?
08:06
I updated the chroot, and rebiult
08:07
<alkisg>
Yes, you usually need to build a new chroot with ltsp-build-client
08:08
<reynolds>
I upgraded from 9.10 to 10.4 on the server, updated the chroot, and rebuilt mu i385 image with the updates. I am getting error PXE-T02, and PXE-E3C
08:10
so "sudo ltsp-build-client --arch i386"
08:11
<alkisg>
Wait
08:11
The PXE error is a different thing
08:11
Running ltsp-build-client after an os update is the normal way to go, but your pxe problem is not related
08:11
What's the full error message?
08:12
<reynolds>
I also tried this fix http://ubuntuforums.org/showthread.php?t=1466911
08:12
<alkisg>
Does it get an IP and die at the tftp stage?
08:14
<reynolds>
yes it gets an ip, Then the errors say PXE-T02 : Forbidden directory, and PXE-E3C:TFTP error access violation
08:16
<alkisg>
OK, it's tftp related then
08:16
Paste your /etc/default/tftpd-hpa to pastebot
08:16
!pastebot
08:16
<ltspbot>
alkisg: "pastebot" :: The LTSP pastebot is at http://ltsp.pastebin.com. Please paste all text longer than a line or two to the pastebot, as it helps to reduce traffic in the channel. Don't forget to paste the URL of the text here.
08:19
<reynolds>
do you want me to paste that? Its kinda hard cause its the boot screen on my client
08:19
oh i see sorry
08:21
http://ltsp.pastebin.com/T7GA6B4T
08:22
<frederickjh>
Hi guys! If I add a file to /opt/ltsp/i386/etc/X11 I need to rebuilt the client for them to see the file. Is that right?
08:22
<alkisg>
It looks OK. Is tftp started? run: sudo status tftpd-hpa
08:23
frederickjh: in ubuntu (using nbd)? yeah..
08:23
ltsp-update-image
08:23
<frederickjh>
Thanks alkisg. yes ubuntu
08:25reynolds has quit IRC
08:33ogra has quit IRC
08:34ogra has joined #ltsp
08:37reynolds has joined #ltsp
08:54
<frederickjh>
Any one here ever used the XRandR parameters in lts.conf to set the size of a monitor?
08:55
I have a monitor that is coming up unknown and at a lower resolution than it can display. I want to bump it up to a higher resolution.
08:56
I have used the xrandr comands on the client in a terminal and there it works but when I add the same commands to the lts.conf in the section for that client I get no change.
08:56wsuetholz has joined #ltsp
08:57
<alkisg>
!lts.conf
08:57
<ltspbot>
alkisg: "lts.conf" :: http://manpages.ubuntu.com/lts.conf
08:57
<alkisg>
Did you try XRANDR_MODE_0 from there?
08:57
<frederickjh>
in the lts.conf? yes
08:57Fionnz has joined #ltsp
08:58yanu_ has quit IRC
08:59
<frederickjh>
My question is what is the string part? that is not listed. Is that the same string as when you run xrandr on the command line?
08:59
<alkisg>
No it's just a resolution
08:59
XRANDR_MODE_0=1024x768
08:59
<frederickjh>
XRANDR_NEWMODE_0="1024x768" 63.50 1024 1072 1176 1328 768 771 775 798 -hsync +vsync
09:00
So can I just use that line?
09:01
0 being the first display?
09:01
<alkisg>
So you need to set a new mode? It isn't there if you just use xrandr on the client?
09:01
<frederickjh>
Right now when the client comes up it is 800x600.
09:02
<Fionnz>
Hello! I have a gbit network problem in my classroom for disabled children. Using Edubuntu 10.04/ltsp-server-standalone: Server connected to a 1GBits switch port. Clients to 10/100 ports. Firstly it is really fast but the it gets stuck while loading the kernel and the max. upload rate from the Server is about 20 to 250 KBits :( anyone ahve an idea? need more infos? ethtool and mii-tool seems to be 1Gbits...
09:02
<frederickjh>
I want 1024x768.
09:02
<alkisg>
frederickjh: ok, if you boot the client normally, and just run xrandr on a gnome-terminal, do you see the 1024x768 resolution?
09:02
<frederickjh>
nope
09:03
I can add it and switch to it.
09:03
but when I logout it disapears.
09:03
So, I thought I could make it permanent in lts.conf
09:03
Ubuntu see the monitor as unknown.
09:03
<alkisg>
X_HORZSYNC=30.0-88.0
09:03
X_VERTREFRESH=50.0-70.0
09:04
<frederickjh>
I need those lines?
09:04
<alkisg>
frederickjh: try putting those in lts.conf, *without* specifying any XRANDR_MODE_0 or anything else
09:04
<frederickjh>
Ok, will do.
09:04
<alkisg>
If I'm right, it'll start in 1024x768 automatically after that
09:10Gadi has joined #ltsp
09:11
<Fionnz>
sorry for interrupting! nobody any hints or ideas?
09:11
if i download or upload any thing to a win share speed works fine. Only while the tftp boot the network hangs...
09:13alkisg has quit IRC
09:14
<frederickjh>
alkisg that did not work. The client is this one http://www.disklessworkstations.com/200118.html
09:14
It has a vga and a dvi output. I am using the vga.
09:15
Fionnz it could be the switch. Faulty or flaky switches are a bad thing in at LTSP environment. Booting is the hardest on them.
09:16
Fionnz, I would also check to see if QOS is turned on, if it is a managed switch, and turn it off.
09:17
<Gadi>
frederickjh: add: X_OPTION_01 = "\"ActiveDevice\" \"CRT\""
09:17Fionnz has quit IRC
09:22
<frederickjh>
Ok Gadi will try
09:23
Do I need all that space before the =?
09:23mordocai has joined #ltsp
09:24
<frederickjh>
Gadi is that in addition to the lines from alkisg above or by itself?
09:24
<mordocai>
Is there a way for an ltsp server to reboot all it's currently connected terminals at once?
09:25
<Gadi>
frederickjh: try with just that line first, and if it doesn't work, try setting the BIOS to only use the CRT outputs
09:25
*output
09:26* Gadi was just talking to an openchrome driver dev about this yesterday (assuming it is the same bug)
09:30Fionnz has joined #ltsp
09:30
<frederickjh>
ok
09:37yanu has joined #ltsp
09:37yanu has joined #ltsp
09:37
<Fionnz>
need help plz!? :)
09:39jorge_ has joined #ltsp
09:39jorge_ is now known as jcastro
09:39jcastro has joined #ltsp
09:42Egyptian[Home] has quit IRC
09:45
<frederickjh>
Well, Gadi I tried it with that line and then set the bios to crt for the boot display but none of it worked.
09:45komunista has joined #ltsp
09:46
<frederickjh>
Fionnz have you tried replacing the switch to see if that is the problem?
09:46vbundi has joined #ltsp
09:46
<frederickjh>
I had a couple of bad switches and pretty much the same problem till I trashed them.
09:47
<Gadi>
frederickjh: sounds like your monitor is not responding to ddc. add the two lines that alkisg suggested in addition to the above:
09:48
[07:02] [alkisg] X_HORZSYNC=30.0-88.0
09:48
[07:02] [alkisg] X_VERTREFRESH=50.0-70.0
09:48
<frederickjh>
ok
09:48
will try it out.
09:48yanu has quit IRC
09:50yanu has joined #ltsp
09:50yanu has joined #ltsp
09:50CAN-o-SPAM has joined #ltsp
09:54
<frederickjh>
Well I have all three lines and no change. Still comes up 800x600.
09:55
The interesting thing is Gadi that I have another client that is exactly the same machine with a different monitor and it comes up ok. So, the problem must be with the monitor.
09:55
<mordocai>
So... is there an easy way to reboot all terminal machines from the ltsp server? I'm wanting to be able to reboot them the morning after every time i change something on the client.
09:55
<frederickjh>
Maybe I need to make a xorg.conf and add a line for that cilent in lts.conf.
09:56
<Gadi>
frederickjh: can you post the Xorg log file from that client?
09:56
(the Xorg log file tells all)
09:56
<frederickjh>
you mean from the clients
09:57
<Gadi>
right
09:57
<frederickjh>
As far as I can tell there is not one.
09:57
<Gadi>
sure there is
09:57
you need to get a shell on the client
09:57
so, set: SCREEN_02=shell and SCREEN_07=ldm
09:57
<frederickjh>
I looked in /opt/ltsp/i386/etc/X11 on the server and I did not find one.
09:58
<Gadi>
that's the build directory
09:58
not the runtime directory
09:58
<frederickjh>
I have the SCREEN settings set already.
09:58
<Gadi>
so, on the client, go to a shell
09:58
<frederickjh>
Ok, so a shell on the client.
09:58yanu has quit IRC
09:58
<Gadi>
there will be a /var/log/Xorg*log
09:58
<frederickjh>
then /etc/X11
09:58
ok
09:58
<Gadi>
no no - the *log*
09:58
<frederickjh>
ok
09:59
and that is the config file?
09:59
<Gadi>
then, you can: scp /var/log/Xorg*log user@server:
09:59
and pastebin it
09:59
thats not the config - it is the *log*
09:59
:)
09:59yanu has joined #ltsp
09:59yanu has joined #ltsp
09:59
<frederickjh>
Thank Gadi I will work on it.
09:59
<Gadi>
logs roll down stairs and under chairs and into the neighbor's yard
09:59
log, logs, they're big, they're heavy, they're wood
10:00
logs, logs, they're better than bad - they're good!
10:02treyhunsucker has joined #ltsp
10:02
<treyhunsucker>
what is your guy's opinion on Lucid + LTSP
10:06
<johnny>
Gadi, lol
10:06
treyhunsucker, lots of people run it.. that's all i know
10:06
i run the equivalent of the lucid ltsp packages i think.. via stgraber's ppa
10:06
i'm ugrading to lucid sometime in the next week
10:08primeministerp has joined #ltsp
10:08
<treyhunsucker>
johnny: cool ty
10:09
johnny: do you run ubuntu server or desktop edition on your installs?
10:09
<johnny>
install* :)
10:09
not installs
10:09
desktop, because it is used as a desktop machine as well
10:09
<treyhunsucker>
actually should be intall(s) :P
10:09
cool same here
10:10
<johnny>
our point of sale software runs on the same machine
10:10
we only have 3 clients tho.. so it's not a big load on the machine
10:10
<treyhunsucker>
cool
10:10staffencasa has joined #ltsp
10:11
<johnny>
/me doesn't understand why pcsd and other stuff is started on fedora boot..
10:11
i don't even know what it does
10:12
<frederickjh>
Ok Gadi. here is the log http://ltsp.pastebin.com/iK1bt67T
10:15
<Fionnz>
frederickjh this is the tcpdump while booting
10:15
profesor@ltspserver:~$ sudo tcpdump -s 1500 -ni eth0 port 68 tcpdump: verbose output suppressed, use -v or -vv for full protocol decode listening on eth0, link-type EN10MB (Ethernet), capture size 1500 bytes 10:21:07.660903 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 08:00:27:0d:f2:8c, length 548 10:21:07.661411 IP 192.168.1.254.67 > 255.255.255.255.68: BOOTP/DHCP, Reply, length 321 10:21:09.658775 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP,
10:17
<Gadi>
frederickjh: first off - sorry, I thought it was a VIA chipset - seems it is intel - so, you can take out the X_OPTION_01 line
10:17
<frederickjh>
ok.
10:18
It is a client from disklessworkstations.com see above for a link to the client specs.
10:18
<Gadi>
frederickjh: 2. I don't think your lts.conf params are making it - can you do: getltscfg -a|grep X_HORZSYNC at the shell
10:18lfaraone has left #ltsp
10:18
<frederickjh>
on the client?
10:18
<sbalneav>
Morning all
10:18
<Gadi>
frederickjh: yeah
10:18
<frederickjh>
ok
10:19
<Gadi>
frederickjh: the monitor is clearly not giving up its frequencies and the log doesn't say that you specified any
10:19
<Fionnz>
no this is the correct one: its on port 69
10:19
http://ltsp.pastebin.com/ChrtqLzW
10:26
<frederickjh>
Ok Gadi it returns three lines X_HORZSYNC="30.0-88.0" export X_HORZSYNC
10:26
and it tells me that there is an error on line 60 of lts.conf but it is only 59 lines long.
10:26
<Gadi>
ah!
10:26
hmm...
10:27
<frederickjh>
I think I saw that before when the file was shorter too.
10:27
<Gadi>
take a look in /var/run/ltsp-xorg.conf
10:27alkisg has joined #ltsp
10:27
<Gadi>
on the client
10:27
and see if the params made it
10:27
<frederickjh>
ok
10:28* Gadi wonders if whatever editor you are using to edit lts.conf is leaving it in a bad way
10:30
<frederickjh>
gedit
10:31
<Gadi>
are there params in /var/log/ltsp-xorg.conf?
10:32
<frederickjh>
http://ltsp.pastebin.com/jpk3QtF8
10:32
Looks like
10:33
<Gadi>
well, that's bad
10:33
hehe
10:33
hmm...
10:33
I think you may have found a scripting bug
10:33
one sec
10:34
<frederickjh>
ok, Gadi
10:35
<alkisg>
sbalneav: I'm starting a shell script from /etc/xdg/autostart and I'd like it to die when the session ends. It never receives a SIGTERM though - I suppose it's supposed to register itself to gnome-session, but being a script, it can't. Any ideas on how to (properly) terminate it a logoff?
10:35
Or, maybe putting it in the ldm scripts would help?
10:36
<sbalneav>
alkisg: Well..... (thinks).
10:36
<Fionnz>
i found the following while tcpdump: sudo tcpdump -ni eth0 -v -T tftp 'udp'
10:36
<sbalneav>
alkisg: Is this running on the server or the thin client itself?
10:37
<Fionnz>
192.168.1.27.2070 > 192.168.1.254.69: 38 RRQ "/ltsp/i386/pxelinux.0" octet tsize 0 11:32:18.114383 IP (tos 0x0, ttl 64, id 30148, offset 0, flags [none], proto UDP (17), length 42) 192.168.1.254.48043 > 192.168.1.27.2070: 14 OACK tsize 14776 11:32:18.114936 IP (tos 0x0, ttl 20, id 3, offset 0, flags [none], proto UDP (17), length 45) 192.168.1.27.2070 > 192.168.1.254.48043: 17 ERROR EUNDEF "TFTP Aborted"
10:37gnunux has quit IRC
10:37
<Fionnz>
192.168.1.27.57089 > 192.168.1.254.69: 90 RRQ "/ltsp/i386/pxelinux.cfg/ea3d1a58-e210-47ed-a0b" [|tftp] 11:32:18.147994 IP (tos 0x0, ttl 64, id 30161, offset 0, flags [none], proto UDP (17), length 47) 192.168.1.254.35436 > 192.168.1.27.57089: 19 ERROR ENOTFOUND "File not found" 11:32:18.149193 IP (tos 0x0, ttl 20, id 18, offset 0, flags [none], proto UDP (17), length 102) 192.168.1.27.57090 > 192.168.1.254.69: 74 RRQ "/ltsp/i386/pxelinux.cfg/01-08-
10:37
<alkisg>
sbalneav: on the server (or, on fat clients)
10:37
<Fionnz>
somebody know what todo? seems that the loader cant find any pxe files!?
10:38
<treyhunsucker>
Fionnz: what is your error while booting?
10:39
<sbalneav>
alkisg: wonder if you could use a (possibly slightly modified) version of my xatom program.
10:39
sit and idle, when you get an X error because the X session's ended, then terminate the script.
10:39
<alkisg>
sbalneav: as a monitor? So that when it dies (X programs die at logoff), it also kills the script?
10:40
<Gadi>
frederickjh: under the chroot dir (/opt/ltsp/i386) rename this file: /usr/share/ltsp/screen-session.d/XS20-xserver-intel-desktop-hack to XS85-xserver-intel-desktop-hack
10:40
<sbalneav>
right.
10:40
<alkisg>
Got it
10:40
<Gadi>
frederickjh: then run ltsp-update-image
10:40
<frederickjh>
Gadi I will be back in a bit (supper). I will try after that.
10:40
<alkisg>
Thanks sbalneav - I wonder why it's so hard though... maybe consolekit could help there
10:40
<sbalneav>
alkisg: Might be a useful utility for us to code up:
10:40
"xsessionterminate" or the like.
10:41
pass it a sequence of process id's to watch on the command line.
10:41
<alkisg>
yeah, like gnome-watchdog does now. I think it'd be useful...
10:41
<sbalneav>
ok
10:41
I'll code something up today.
10:41leio has quit IRC
10:41
<sbalneav>
I've got a touch of spare time this afternoon.
10:41leio has joined #ltsp
10:42
<alkisg>
Damn, need to pick up the kids asap. Thanks, bbiab.
10:42
<sbalneav>
cool
10:43icullen has joined #ltsp
10:44
<Fionnz>
treyhunsucker while starting loading vmlinuz the bandwith is very slow and need about 20 min to complete
10:44
<icullen>
Can anyone help me with LTSP in a fresh install of Ubuntu 10.04. I can't boot clients - it says "Error: failed to connect to NBD server" then dumps the client to a busybox shell. 9.10 worked OK out the box.
10:45
<johnny>
you broke it..
10:45
/me doesn't really know about a fresh install .. is going to do an upgrade
10:45
<Gadi>
icullen: you ran ltsp-build-client?
10:46
<treyhunsucker>
Fionnz: what are the specs for your server
10:47
<icullen>
i installed 10.04 using the alternate CD, selected "install an LTSP server" at the startup screen. I haven't run any commands on the server, 9.10 just worked without any further configuration
10:48
<treyhunsucker>
icullen: is it the dhcp server?
10:49
<icullen>
no, i've uninstalled dhcp3-server and we are using windows 2003 dhcp to boot the client. same configuration as we had under 9.10
10:49
<Gadi>
icullen: try running: ltsp-update-image --arch i386
10:49
<icullen>
017 Root Path /opt/ltsp/i386 066 Boot Server Host Name 192.168.0.33 067 Bootfile Name ltsp/i386/pxelinux.0
10:50
will try the update image just now
10:55
<Fionnz>
treyhunsucker edubutn 10.04 release dvd 2GB, 4*2,8 Ghz Rack-server with 2 Gbit nics
10:55
<icullen>
holy crap, that's fixed it :D
10:55
you guys are awesome, thanks a million!
10:56mordocai has quit IRC
10:58tstafford has quit IRC
10:58icullen has quit IRC
10:58
<Fionnz>
treyhunsucker: switch is a Trendnet with 2 GBit ports and 24 10/100 ports
10:59tstafford has joined #ltsp
11:04dobber has quit IRC
11:05highvoltage has joined #ltsp
11:07Fionnz has quit IRC
11:12Fionnz has joined #ltsp
11:12
<Fionnz>
restarted, any messages i missed?
11:14
<treyhunsucker>
Fionnz: not sure whats wrong
11:14
Fionnz: when did it start, did you change anything?
11:17
<Fionnz>
hmm i tried a new switch!
11:17
http://ltsp.pastebin.com/hGrYutqD
11:17
some logs! this could be the error: 19 ERROR ENOTFOUND "File not found"
11:17
but why?
11:22
<Gadi>
Fionnz: those errors are prefectly normal - pxelinux looks for a bunch of configuration files in pxelinux.cfg/ and uses the first one it finds that matches the client - when it does not find them, youget that error and it moves on to look for the next one
11:23
<treyhunsucker>
Fionnz: so you tried a new switch and that is when the errors started?
11:23
<Gadi>
the only weird thing in your paste is that your client aborted its first attempt to d/l pxelinux.0
11:26Fionnz has quit IRC
11:33
<highvoltage>
-/win 45
11:33treyhunsucker has left #ltsp
11:44
<sbalneav>
alkisg: https://code.edge.launchpad.net/~sbalneav/ltsp/xexit
11:45
<alkisg>
sbalneav: perfect! You rock!
11:47
<sbalneav>
It's maybe a little rough. Play with it a bit and let me know.
11:47
If it ends up being useful, we can import it intl the ltsp tree.
11:48
Xatoms to the rescue again!
11:48
"Xatoms! Is there anything they CAN'T do?"
11:48
Off for lunch
11:48
<alkisg>
And it's plain X == pretty light... yey!
11:48
Thanks sbalneav
11:48
<sbalneav>
bbiab
11:49reynolds has quit IRC
11:55shogunx has joined #ltsp
11:57hersonls has quit IRC
11:58Egyptian[Home] has joined #ltsp
12:01
<frederickjh>
Ping Gadi!
12:02
I am back and was able to get it all done.
12:02
I think we are headed in the right direction Gadi.
12:02
However now the monitor is saying out of range.
12:03
<alkisg>
frederickjh: what happened with the X_HORZSYNC/X_VERTREFRESH ?
12:03
<frederickjh>
Anyone know the new key combination to change monitor modes? It use to be Ctrl-Alt and the numpad + or - keys
12:03
I think it is still in there.
12:03
Let me check.
12:04
yah it was still in there and the X_OPTION_01="\"ActiveDevice\" \"CRT\"" line too.
12:04
I think Gadi said to take that one out.
12:05
<alkisg>
frederickjh: so, you didn't get the 1024x768 resolution with that?
12:05
Or did you get out of range?
12:05
<frederickjh>
out of range.
12:05
<alkisg>
OK, try to lower X_VERTREFRESH then
12:05
Put it up to 60.0 instead of 70.0
12:06
<frederickjh>
ok will try
12:11
ok, I put it up to 60 and got rid of the X_OPTION_01 line too, but it is still saying out of range.
12:13Kicer86 has joined #ltsp
12:14
<frederickjh>
Is there a way I can set it to 1024 by 768 pixels alksig or Gadi, as that is the highest resolution the monitor can do.
12:14
<alkisg>
Can you paste what you currently have in lts.conf?
12:14
<frederickjh>
one second
12:15
http://ltsp.pastebin.com/JmDPNx8H
12:16
# 1st Disklessworkstations.com Atom Intel 1.6 GHz - Oliver's machine at the moment is the machine we have been working on.
12:16
<alkisg>
Can you try also enabling this? XRANDR_MODE_0=1024x768
12:16
<frederickjh>
sure
12:17
be back after the client boot.
12:18mordocai has joined #ltsp
12:19
<johnny>
/me is sad that evolution keeps adding bugs that are really annoying
12:23Usuario563 has joined #ltsp
12:24
<mordocai>
So... never got an answer so i'll ask again (sorry if i'm annoying). Is there a way from the LTSP server to cause all clients to reboot?]
12:24Usuario563 has quit IRC
12:24
<frederickjh>
Ok, finally sucess. Login screen came up at 1024x768. After login the screen was 800x600 but now could be changes under System>Display.
12:25
mordocai all at the same time?
12:25
<mordocai>
frederickjh: Yeah. In case i just did a kernel update on the client or something like that.
12:25
<frederickjh>
On my system if you make a new client image and a user logs out then it goes for the reboot.
12:25
<mordocai>
Well, i'm using NFS
12:26
So i don't really make a new image
12:26
<alkisg>
frederickjh: ok, then the horizontal/vertical values need some tweaking, so that you can remove XRANDR_MODE_0. Can you paste the output of `xrandr`?
12:26
<frederickjh>
from the client?
12:26
<alkisg>
Yes
12:26
<frederickjh>
mordocai try Italc
12:28
<vbundi>
I tried iTalc and while I think it's awesome, it seemed a little slow, I would think it would be unusable in 30+ terminal environments, maybe something is up with my configuration
12:28
maybe because I'm running 1280x1024 sessions
12:29
<mordocai>
frederickjh: Hmm... looks interesting. However, we would have no use for the other features. I'm at a college and trying to configure all our kiosks with LTSP. Currently still in the "test phase" (politics and stuff to get through) but we have our kiosks spread throughout the entire campus and it'd be a pain to go reboot them manually. Though, i'm not sure at what "layer" our existing system for restarts is... maybe it works lower
12:30
<frederickjh>
http://ltsp.pastebin.com/2WcHK6nA here it is alkisg
12:30
<abeehc>
mordocai: i have root unlocked +ssh enabled; and i have a script that reboots any clients not already logged in
12:31ogra_cmpc has quit IRC
12:31
<frederickjh>
what distro are you using mordocai?
12:31
<abeehc>
works for the most part save in situations where nbd is gimped the reboot command doesn't work that well or at all
12:31ogra_cmpc has joined #ltsp
12:31
<mordocai>
frederickjh: red hat enterprise. Playing with beta 6 rightn ow
12:31
right now*
12:32
<frederickjh>
ok using ubuntu here.
12:32
<alkisg>
frederickjh: what card is this? ati? It ignores horzsync... So anyway leave XRANDR_MODE_0 as it is, and be careful not to choose a larger mode.
12:32
<mordocai>
Yeah, i prefer Debian(use it at home), but my higher ups wanted redhat
12:33
<frederickjh>
alkisg the client is this one http://www.disklessworkstations.com/200118.html look under the specs tab for the video card specs
12:34
<alkisg>
Intel? I wonder why it ignores it then... anyway
12:34
<frederickjh>
abeehc you have that reboot not working too?
12:34
<abeehc>
not sure what you mean
12:34
i'm able to reboot all my clients as neccesary
12:35
<frederickjh>
I notice that on my clients that if I power down no problems but tell them to reboot and they give a bunch of SQUASHFS errors and never do reboot.
12:35
<abeehc>
i have that problem too yeah
12:35
<alkisg>
Which ubuntu versions?
12:35
<abeehc>
annnoying
12:35
9.04
12:35
<alkisg>
Yeah an update is needed to use reboot -fp instead of plain reboot.
12:35
(or manual patching of the script)
12:35
<abeehc>
frederickjh: http://kb.yarmakconsulting.com/2008/01/rebooting-linux-without-reboot-command.html
12:36
i have 9.10 just testing, going to take the 9.04 server out soon enough
12:36
9.10 with the ppa
12:37
<frederickjh>
I am on ubuntu 9.10
12:37
alkisg what update?
12:39
<alkisg>
ldm-trunk/rc.d/I01-nbd-checkupdate: reboot -fp
12:39
ldm-trunk/rc.d/X99-zzz-logout-action: reboot -fp
12:39
ldm-trunk/gtkgreet/greeter.c: g_spawn_command_line_async("/sbin/reboot -fp", error);
12:39
ltsp-trunk/client/screen-session.d/XS00-halt-reboot: exec reboot -fp
12:39
In some of those lines, in previous ltsp versions, the -fp was missing
12:39
And that's what's causing the squashfs errors
12:40
Using an updated ltsp version (e.g. from stgraber's ppa) would fix the problem...
12:40
!stgraber-ppa
12:40
<ltspbot>
alkisg: "stgraber-ppa" :: https://launchpad.net/~stgraber/+archive/ppa
12:40
<frederickjh>
ok
12:40
But it should be in the new ltsp some time soon?
12:41
<abeehc>
yeah it works well with the ppa
12:41
i think i was even able to reboot my service and clients survived
12:41
<frederickjh>
It is not a big deal to me. I have a hand full of clients.
12:41
<abeehc>
maybe i was on drugs that day though not certain
12:42
<frederickjh>
Interesting on Ubuntu the system policy will not let me reboot if other users are logged in unless I give my password.
12:42
<alkisg>
That's consolekit & policykit
12:43
<johnny>
frederickjh, not just ubuntu..
12:43
<frederickjh>
Thanks alkisg and Gandi for your help!
12:43
<alkisg>
np
12:59
<komunista>
alkisg: thanks, I see it is for 9.04 too :-)
13:12wsuetholz has quit IRC
13:12wsuetholz has joined #ltsp
13:26
<sbalneav>
Back from lunch
13:26
Gandi?!
13:27* sbalneav imagines Gadi in a loin cloth
13:28komunista has quit IRC
13:31
<cliebow>
cliebow imagines sbalneav in a loincloth
13:32
and gold paint haed to toe 8~)
13:36
<sbalneav>
That'd have to be a big loin cloth.
13:36
More like a re-purposed tablecloth.
13:36
Trust me, you WOULDN'T want to eat off of it again.
13:37
cliebow: How's the new digs coming?
13:37
All done yet?
13:37
<cliebow>
ready to move in..it is very gorgeous there
13:38
still have to duild walls in cellar and all
13:38
errr build
13:39
<alkisg>
sbalneav: at xexit, why is an xprop needed? Can't the program just use an "at_exit()" hook?
13:40komunista has joined #ltsp
13:46
<alkisg>
heh, nm, /me goes on to test it...
13:47komunista has quit IRC
13:52reynolds has joined #ltsp
13:53komunista has joined #ltsp
13:55
<reynolds>
Does anyone know how I can figure out what is stressing my server CPU? Since I have updated my server to ubuntu 10.4 programs on my clients are maxing out my cpu.
13:58
<frederickjh>
top?
13:59
<highvoltage>
reynolds: I usually install htop and keep one eye on it
13:59
<frederickjh>
The other thing you can do is turn off data encryption if not needed.
13:59
add this to lts.conf
13:59
# Used to allow unencrypted traffic and cut down on CPU usage
13:59
LDM_DIRECTX=True
14:00
<reynolds>
frederickjh: how do I turn of encryption. Is that with lts.conf?
14:00
<frederickjh>
It is the configuration file for ltsp
14:00
get me get the location.
14:00
The other thing with the new ubuntu's and ltsp is the run away ica processes.
14:01
I will have no one logged on and go to 100%. I run killall ica and the cpu usage goes back to normal.
14:02
lts.conf is located in /var/lib/tftpboot/ltsp/i386
14:02
You will need super user permissions to edit it.
14:03
Add the line above under the [Default] section so that it applies to all clients.
14:03
This will be at the beginning.
14:05
<reynolds>
I already have encryption turned off it seems?
14:05
<frederickjh>
See here reynolds for more info on lts.conf https://fedorahosted.org/k12linux/wiki/LTS_PARAMETERS.TXT
14:05
Ok,
14:05
<reynolds>
thanks
14:05
<frederickjh>
If you type top in a terminal what process is taking up the cpu?
14:08
<reynolds>
got it I have a user that says it is running but it is logged out. How can I logoff a user from the command line?
14:10
<frederickjh>
I think that if you man kill or man kill all there is a way to kill all process that belong to a user.
14:10
There is a guy that has a watchdog that will do this for you.
14:11
The watch dog program sit around waiting for the user to exit and if their processes do not end shortly there after it kills them.
14:11
I can dig up a link if you are interested, reynolds.
14:12
<alkisg>
frederickjh: for the 100% cpu usage disable vino-server
14:14
<frederickjh>
reynolds I googled and found
14:14
killall -u username It will kill all the process owned by a username.
14:14
<reynolds>
I have watchdog installed By I think with the upgrade it got a little messed up
14:14
<frederickjh>
how alkisg?
14:14
<alkisg>
frederickjh: system > preferences > remote desktop or something like that > disable
14:14
<frederickjh>
thxs
14:15
<alkisg>
That's a per-user setting
14:15
<frederickjh>
It is off.
14:15
<alkisg>
ps -ef|grep vino => does that say anything?
14:16
<frederickjh>
alkisg I think what is happening is that the ica process hang around from the last session and clog up the works.
14:16
ps -ef|grep vino
14:16
1000 11959 2920 0 21:15 pts/5 00:00:00 grep vino
14:16
only the grep
14:17
<alkisg>
Do you have 100% cpu usage now? And, what does `ps -ef | grep ica` give?
14:18
ica-launcher kills any previous processes laying around, so the last session shouldn't affect it...
14:18
<reynolds>
thanks Frederickjh
14:20
<frederickjh>
No not now. Not every time but fairly often when the server boots it goes to 100% cpu. Running killall ica takes care of the problem.
14:21
<sbalneav>
alkisg: do a bzr update: found a bug.
14:21wsuetholz has quit IRC
14:22
<frederickjh>
alkisg that is a very inefficient way to search for a process. Try using the -C option like this ps -C ica
14:22
<sbalneav>
alkisg: So, if you had one script that called a run-parts on a "session-cleanup.d" directory... You'd have a nice general cleanup idea goin' on
14:22
<frederickjh>
You way turned up 18 processes when there were only 2 ica running.
14:23reynolds has quit IRC
14:23
<alkisg>
frederickjh: I do so in scripts, but I forget the switch when I just type commands at the prompt. Anyway I'm sure my hands are slower that a grep so I don't mind :)
14:23
<frederickjh>
I think that the -C option works in scripts too. I think I have used it.
14:24
<alkisg>
I have, it works fine
14:24
sbalneav: thanks, pulling...
14:24
<sbalneav>
just tested a logout, works peachy-keeno.
14:24
I'll maybe clean up the code a bit tonight, but it's workable now.
14:26Jinux has joined #ltsp
14:26Jinux has left #ltsp
14:26
<alkisg>
sbalneav: with that (and also by replacing `reboot -fp` to plain `reboot`) I think I have the whole "remote command execution" backend ready, with just a sh + xexit per session == minimal RAM.
14:27
I can already get client screenshots, boot, reboot them etc - I hope I'll refine it enough soon
14:27
...so that it can then be internationalized
14:27
<frederickjh>
alkisg I believe I saw somewhere that someone said it had to do with the Startup Applications Option to Automatically remember running applications when logging out. I think that the story was stuff got started twice and the next time even more.
14:27
<sbalneav>
While the xexits sitting.... (hold on...)
14:28
-rwxr-xr-x 1 sbalneav sbalneav 10488 2010-05-05 14:26 xexit
14:28
10 k binary...
14:28Joeasaurus has joined #ltsp
14:28Joeasaurus has left #ltsp
14:29Jinux has joined #ltsp
14:29
<alkisg>
htop reports 3M virtual, 700k res
14:29
Very very nice
14:29Jinux has left #ltsp
14:30
<alkisg>
frederickjh: if that's the case, then it should be easily solved with some .desktop file magic
14:30
<sbalneav>
Hmmm, I'dve thought it'd be less than that.
14:30
<alkisg>
file a bug if you can reproduce it
14:30
<sbalneav>
Definitely shouldn't be sucking up much in the way of CPU
14:30
<frederickjh>
Where would I file the bug report and what would I need for information.
14:31
<alkisg>
https://bugs.launchpad.net/ubuntu/+source/italc/+filebug
14:31
<frederickjh>
Something like log the number of running ica process during a number of reboots?
14:31
<alkisg>
Post the required steps to reproduce it
14:31
<frederickjh>
and the CPU usage.
14:32
I just might do that as it is quite annoying.
14:32
<alkisg>
Do you have the "remember session" option enabled?
14:32
<frederickjh>
Yes
14:32
<alkisg>
OK, never tried that, it might be the cause
14:32* alkisg never uses italc anymore anyway..
14:33
<sbalneav>
sbalneav@3jane:~/Source/bzr/xexit/src$ memstat | grep xexit 276k: PID 9222 (/home/sbalneav/Source/bzr/xexit/src/xexit)
14:33
memstat here for me reports 276k.
14:33
That's a keeper :)
14:33
<alkisg>
252k: PID 21252 (/home/alkisg/tmp/xexit/src/xexit)
14:34
Obviously htop reports different measures
14:35
<sbalneav>
Cripes!
14:35
116520k: PID 8414 (/usr/lib/evolution/evolution-data-server-2.28)
14:35
WHAT IS IT DOING!?!?!
14:35
<alkisg>
sbalneav: not only that... it also DOESN'T DIE AT LOGOFF :D
14:35
<sbalneav>
yeah
14:35
<alkisg>
You gnome upstream guys should take care of such stuff :P
14:35
<johnny>
does anybody here use evo? and also notice that if you get to the unread message the mark message as read after X seconds doesn't fire ?
14:36
<sbalneav>
alkisg: Hush, you.
14:36
<johnny>
in 2.30 that is..
14:36
<alkisg>
johnny: 2.30? where's that?
14:36
<johnny>
get to the unread message before the page is rendered
14:36
<alkisg>
I got 2.28 in Lucid...
14:36
<johnny>
that doesn't make sense..
14:36
you should have gotten the entire gnome 2.30 stack
14:37
<alkisg>
(10:35:59 μμ) ubottu: evolution (source: evolution): groupware suite with mail client and organizer. In component main, is optional. Version 2.28.3-0ubuntu9 (lucid), package size 2370 kB, installed size 7020 kB
14:37
<johnny>
evo 2.30 was actually branched halfway through 2.27 for 2.30
14:37
well.. blame ubuntu i guess
14:37
2.30 has imap idle finally
14:37
<alkisg>
..and a lot of broken stuff, it seems :)
14:38
Message marking works fine here
14:38
<johnny>
not alot.. just a slight timer annoyance
14:38
no.. it's if you get to the message before the whole list of messages is rendered
14:38
the the timer doesn't start to mark it as read after X seconds
14:38
so.. i have to click off it, and then back on
14:39
i'm very suprised ubuntu shipped 2.28
14:39
since 2.29 was branched halfway through 2.27
14:39
<alkisg>
Ah. I can't do it that fast - I don't use imap so it's rendered instantly
14:39
<johnny>
so they could start debonoboifying
14:41
<sbalneav>
I have to debonofy sabayon...
14:43
wait
14:43
debonobofy
14:43
I could debonofy U2
14:49
<johnny>
lol
14:50
but seriously.. why would somebody ship an un imap idle enabled evolution
14:50
i hope they plan on upgrading it later
14:50
btw.. ted ts'o really poked hard at canonical for not hiring their own engineers
14:50
<sbalneav>
Blog?
14:51
<johnny>
launchpad bug post related to ext4
14:51
i think i still have it..
14:51
http://airlied.livejournal.com/72817.html
14:51
link is in there
14:52
lol
14:52
"Since this really seems to be an upstream-relevant bug motivated by an ubuntu
14:52
bug that ubuntu seemed unable to solve, I'm going to close this as UPSTREAM and
14:52
worry about it there, rather than working the ubuntu bug through the fedora
14:52
bugzilla. "
14:53
that's on the 2nd link..
14:55pmatulis has quit IRC
14:59
<sbalneav>
heh
15:00
alkisg: more updates, and a "example" end session script that uses run-parts (stolen shamelessly from Xsession :)
15:01
<alkisg>
Nice! Unfortunately it's getting late here, I'll tes more + integrate it tomorrow though. Any plans on pushing it to ltsp after the testing?
15:02
Ah, I'd need it for fat clients as well, so it won't do me any good in either ltsp-server nor in ltsp-client-core
15:03
...maybe I'll just hack a quick separate package for it if you don't mind
15:07cliebow has quit IRC
15:07
<sbalneav>
alkisg: By all means.
15:07
<alkisg>
ty
15:08
<sbalneav>
I can even look at quickly packaging it tonight.
15:08
<alkisg>
That'd be even better ;)
15:08
<sbalneav>
I'll ping you tomorrow.
15:08
sbalneav@3jane:~$ cat /etc/Xexit.d/11kill-evo-dataserv
15:08
pgrep -f evolution-data-server | xargs kill -9
15:08
true
15:08
sbalneav@3jane:~$
15:08
Works like a charm :)
15:09
all we need's an Xsession.d file for it and we're in business :)
15:09
Ah, I feel like I did something useful today :)
15:10* alkisg just noticed that sbalneav's hostname doesn't conform to RFC 952 :P :D
15:10
<alkisg>
You surely did
15:10
I'll put it to work tomorrow morning
15:12
Thanks a lot sbalneav... good night everyone
15:12alkisg has quit IRC
15:22lucascoala has quit IRC
15:25reynolds has joined #ltsp
15:30Kicer86 has quit IRC
15:31komunista has quit IRC
15:32CAN-o-SPAM has quit IRC
15:32yanu has quit IRC
15:32rmainard has joined #ltsp
15:35yanu has joined #ltsp
15:35yanu has joined #ltsp
15:35ogra_cmpc has quit IRC
15:37ralth has joined #ltsp
15:37yanu has quit IRC
15:38
<ralth>
Hello! I'm trying to get google chrome working on an Ubuntu Lucid fat client. I set SSH_FOLLOW_SYMLINKS to false, but it's still not working. Any ideas?
15:38
<rmainard>
Hello everyone, I am running an ltsp cluster with ubuntu 8.04 as my load balancer and control center and edubuntu 9.1 for my root and app servers. I have added XSERVER=vesa, X_ORG_0=1024x768, LDMDIRECTX=True to the lts.conf all of which made it possible to log into my thin clients. the problem is when I log out or have a failed login the screen goes blank and it never recovers. Do you have any ideas as to what this could be? Any information would be hel
15:38
pful and appreciated, thanks in advance
15:40yanu has joined #ltsp
15:40yanu has joined #ltsp
15:41ogra_cmpc has joined #ltsp
15:42treyhunsucker has joined #ltsp
15:42
<treyhunsucker>
is there a special way to update the chroot in lucid?
15:44
<vbundi>
you mean to update your software?
15:44
like apt-get update apt-get dist-upgarde?
15:45
I've just been doing 'sudo chroot /opt/ltsp/ARCH apt-get update && apt-get dist-upgrade -y
15:46
<treyhunsucker>
vbundi: yup
15:46
<vbundi>
yeah I dunno if there is a special command to do that is shorter
15:47
<treyhunsucker>
vbundi: you haven't been using the process that was used for 9.10 and 9.04
15:47
<vbundi>
yeah the ltsp system I am replacing is 4.2
15:47
so I figured I'd skip right to lucid ;)
15:48
<treyhunsucker>
lol
15:48ralth has quit IRC
15:48
<treyhunsucker>
vbundi: i think there is a more complicated method to update it
15:48
<vbundi>
treyhunsucker: https://help.ubuntu.com/community/UbuntuLTSP/UpdatingChroot
15:48
<treyhunsucker>
more of the properway
15:48
yes i have that link
15:48
but does it still apply
15:48
read the bottom paragraph
15:48
<vbundi>
check out the script at the bottom
15:48
ya
15:48
<treyhunsucker>
i mean the top line
15:49
Built in to Ubuntu 10.04 Ubuntu 10.04 already has the ability to provide recommended updates by default. The rest of the instructions on this page are not necessary.
15:49
<vbundi>
oh well
15:49
I suppose I need to read a little better... it sort of DOES imply there is a special command for lucid
15:50
something like ltsp-upgrade-image --arch
15:50
that would make sense to me
15:50ralth has joined #ltsp
15:51
<rmainard>
Hello everyone, I am running an ltsp cluster with ubuntu 8.04 as my load balancer and control center and edubuntu 9.1 for my root and app servers. I have added XSERVER=vesa, X_ORG_0=1024x768, LDMDIRECTX=True to the lts.conf all of which made it possible to log into my thin clients. the problem is when I log out or have a failed login the screen goes blank and it never recovers. Do you have any ideas as to what this could be?
15:51
<ralth>
Is there any way to check from the client and see if SSH_FOLLOW_SYMLINKS=false is being processed correctly?
15:52
<rmainard>
ralth are you writing to me?
15:52
<ralth>
Nope, just asking a question. :-)
15:52
<rmainard>
k
15:52Selveste1 has joined #ltsp
15:54
<johnny>
ralth, you should my understanding is that you should be using nfs instead of sshfs
15:54
there's an option to do that now
15:54
alkisg knows about it
15:54
but if you wanna test follow symlinks, read the docs for sshfs
15:54
man sshfs
15:55
and try mounting a directory on your own server that contains a symlink
15:55
that's how you'll know
15:55Egyptian[Home] has quit IRC
15:55
<ralth>
ok, thanks
15:58
<treyhunsucker>
johnny: is there any documenation on how to update the chroot of a lucid install
15:59
<Gadi>
\o/ starting sound from udev works nicely
16:00* Gadi makes mental note to change that in our code
16:00
<treyhunsucker>
lol
16:00* Gadi sidesteps another race condition
16:02ralth has quit IRC
16:03
<Gadi>
out of curiosity - any of you folks see a difference playing flash content on AMD processors versus Intel?
16:04
<vbundi>
nope
16:06
<treyhunsucker>
rmainard: have you tried using a different video settings?
16:06
<vbundi>
user perceivable difference that is..
16:06
<treyhunsucker>
Gadi: the biggest problem with that question is that most people don't have both that are the same speed
16:07
gadi: so it ends up being what machine i'm running it on......but most of my text boxes run xeon so my answer will be intel :)
16:07
<vbundi>
treyhunsucker: yeah, the question might as well be... Do you notice a difference driving to work between a Sunfire or a Cavalier
16:07alexqwesa has quit IRC
16:09
<rmainard>
treyhunsucker: like what? I had to put the lts.conf settings in to get it to log in. I tried the default and that cause a lockup on login.
16:09alexqwesa has joined #ltsp
16:09Egyptian[Home] has joined #ltsp
16:10
<treyhunsucker>
rmainard: how long have you had this system?
16:10
<Gadi>
vbundi: I only ask bec I was talking with an openchrome dev yesterday - comparing a via-cpu-based system with an intel-atom-based system vis-a-vis running local flash content and it is noght and day
16:10
he showed me a link that described how adobe uses an intel optimizer to optimize flash for intel cpus
16:10
at the detrement to other cpus
16:10
<treyhunsucker>
Gadi: let me guess intel killed via correct?
16:10
Gadi: yup yup
16:11
Gadi: you see the notice that IE9 is also trying to push adobe flash out the door
16:11
<Gadi>
this would suggest that it is not the graphics card but the cpu
16:11
vis-a-vis flash
16:11
<rmainard>
treyhunsucker: I've been testing and re-building this cluster for quite a while now and this is the first day on this installation.
16:11
<treyhunsucker>
i'm all for html5
16:11
<Gadi>
evidently flash is almost 100% software rendered
16:12
<treyhunsucker>
rmainard: ok just curious why you used 8.04 lts instead of 10.04 lts or 9.10
16:12
<vbundi>
Gadi: yeah I don't know if you remember, but I had VIA systems that took 5 minutes just to boot ubuntu 9.10 terminal sessions
16:12
<Gadi>
vbundi: well, thats a different issue
16:13
<vbundi>
Gadi: pretty much but the point is the same... the software is not designed to utilize certain features (if they exist) in that processor
16:13
<Gadi>
vbundi: I am just isolating the performance of flash - the theory holds up because rendering other 2d seems to not be as bad
16:13
vbundi: ah I see what u mean
16:13
<vbundi>
Gadi: now in AMD and Intel's case... they have SSE2, SSE3, SSE4.x, AMD has 3DNOW
16:14
<abeehc>
i think hardware accel off the video driver is the most pertinent
16:14
<vbundi>
Nvidia chipsets use hardware decoders now for HD video, VDPAU under linux utilizes this.... without it.. your CPU usage is insane while trying to decode x264 1080p
16:15
<abeehc>
to be fair it's not exclusively nvidia making available hardware accell
16:15
but vpdau is the best method out there for sure
16:15
<vbundi>
no, ATI does it too
16:15
<abeehc>
ati's vpdau equivalent hasn't matured yet i don't think
16:16
<Gadi>
right, but flash does not use hardware accel
16:16
<abeehc>
with the new intel stack my clients do pretty well
16:16
it does if it's available
16:16
<rmainard>
treyhunsuker: I was using 9.10 but the writeup and all of the examples on the ltsp-cluster website uses 8.04 for those servers and 9.04 for the root and app servers. I had that config on my last test but decided to try updating to 9.1 for the root and app to see if it worked better with some of the apps I want to use. I did have the same problem with that config as well
16:16
<Gadi>
so, none of that is relevant if you are playing flash content
16:16alexqwesa has quit IRC
16:16
<abeehc>
otherwise my laptop would melt at youtube
16:16
flash uses hardware accell
16:16
or you are junk
16:16
short of xeons
16:17
<vbundi>
Gadi: hardware accell for graphics? maybe not, but for lots of other stuff it may use 'cpu hardware accel' if you know what I mean
16:17
<treyhunsucker>
rmainard: you try might the edubuntu room here
16:17
have you tried #edubuntu
16:17
<abeehc>
http://www.kaourantin.net/2008/05/what-does-gpu-acceleration-mean.html
16:18
<rmainard>
No, but I had the same problem with my non-edubuntu test, I'll give it a try. Thanks
16:19alexqwesa__ has joined #ltsp
16:20
<abeehc>
flash is not cool in any case
16:20
http://www.youtube.com/testtube
16:21treyhunsucker has left #ltsp
16:23alexqwesa__ has quit IRC
16:24alexqwesa__ has joined #ltsp
16:25mordocai has quit IRC
16:25
<Mava>
remember seeing in xkcd some comic about new kernel.. guy told that new kernel supports upto 4096 cpu's. another asked does the flash still work, answer was " who needs flash ? "
16:26
<abeehc>
i cannot find that damned comic for the life of me
16:26
if you find what number it was let me know
16:27
<vbundi>
abeehc: interesting read on the flash gpu stuff
16:28alexqwesa__ has quit IRC
16:28
<vbundi>
abeehc: you can see 3D models on Blizzards website for Warcraft characters
16:28
<Mava>
promise, cant remember it straightly, but also looking for it -D
16:29
<abeehc>
vbundi: where? the first page pegs my poor laptop hehe
16:30
oh i see
16:30
a goblin
16:30
<vbundi>
um here
16:31
<abeehc>
tried to right click the 3d goblin and my chrome tab crashed
16:31
<vbundi>
abeehc: try this http://www.wowarmory.com/character-sheet.xml?r=Sen%27jin&cn=Segolene&gn=Premonition
16:32
<abeehc>
hehe i dunno
16:32
chrome tab crashes tryin to load that page
16:32
my laptop fails
16:33
<vbundi>
well, my desktop which is a pretty decent quad core hits about 20% cpu usage while spinning a 3d model on there heh
16:33
<abeehc>
hehe
16:33
well even that number is pretty freakin high for my tastes
16:33
<vbundi>
but it's neat to see how it works.
16:33
<abeehc>
there we go ff opened it with no problem
16:34
so this is non-flash?
16:34
<vbundi>
no, it's flash
16:34
<abeehc>
that rocks
16:34
oh
16:35
<vbundi>
rocks? oh yeah I can't wait till they start coming out with 3D adds with flying saucers telling me that I'm a lucky winner.. haha
16:36
<abeehc>
i play quakelive outta the browser ;)
16:36
<vbundi>
wha?
16:36
ommg
16:36
<abeehc>
it's pretty much q3a but even my old p4 1.8 laptop couldplay it via firefox quite well
16:37
i'm sure it's not flash though
16:37
<vbundi>
man id software is so cool
16:38
<abeehc>
i'm also a fan
16:38
<vbundi>
John Carmack deserves a medal or two
16:38alexqwesa has joined #ltsp
16:39
<abeehc>
http://www.addyosmani.com/resources/googlebox/ non-flash
16:39
and trippy
16:40
<vbundi>
wow
16:40
still uses a bit of cpu but less than that warcraft thing
16:40
<abeehc>
agreed
16:41
http://addyosmani.com/blog/13-incredibly-fresh-new-canvas-demos/
16:41
that testube linke i posted earlier allows you to use youtube w/o flash for a lot of videos
16:44
<vbundi>
that is so cool
16:44
the 13 things that is
16:44
youtube w/o flash is neat too
16:45alexqwesa_ has joined #ltsp
16:46alexqwesa has quit IRC
16:50alexqwesa__ has joined #ltsp
16:51alexqwesa_ has quit IRC
16:52
<vbundi>
abeehc: wow that quake live is amazing lol, takes me back
16:55alexqwesa__ has quit IRC
16:56alexqwesa__ has joined #ltsp
17:05alexqwesa__ has quit IRC
17:06alex789 has joined #ltsp
17:11alexqwesa has joined #ltsp
17:13
<rmainard>
What could be causing an ltsp client to go to a blank screen after logout instead of returning to the login prompt?
17:14
<johnny>
bad video driver?
17:14
it's not unknown
17:14
some drivers don't perform vt switching very well
17:14alex789 has quit IRC
17:14
<rmainard>
I am using the vesa setting in lts.conf
17:14
<johnny>
my position still stands..
17:14
but really.. i was thinking about the kernel framebuffer driver
17:15
<rmainard>
is it possible to change that
17:15
?
17:17alexqwesa_ has joined #ltsp
17:18alexqwesa has quit IRC
17:22alexqwesa_ has quit IRC
17:24rmainard has quit IRC
17:25alexqwesa_ has joined #ltsp
17:28frederickjh has left #ltsp
17:30alexqwesa_ has quit IRC
17:31Gadi has left #ltsp
17:32alexqwesa_ has joined #ltsp
17:37alexqwesa_ has quit IRC
17:41alexqwesa__ has joined #ltsp
17:46alexqwesa__ has quit IRC
17:48alexqwesa__ has joined #ltsp
18:01alex789 has joined #ltsp
18:02johnny has quit IRC
18:03alexqwesa__ has quit IRC
18:30vagrantc has joined #ltsp
18:52bobby_C has quit IRC
19:07johnny has joined #ltsp
19:19johnny has left #ltsp
19:20johnny has joined #ltsp
19:44vagrantc has quit IRC
20:03
<sbalneav>
Evening all
20:20rjune has quit IRC
20:41waldo323 has joined #ltsp
20:52waldo323 has quit IRC
21:11lucascoala has joined #ltsp
21:16rjune has joined #ltsp
21:20rjune has quit IRC
22:15ogra_cmpc has quit IRC
22:31ogra_cmpc has joined #ltsp
22:47gentgeen__ has quit IRC
23:00gentgeen__ has joined #ltsp
23:01alkisg has joined #ltsp
23:21gentgeen__ has quit IRC
23:26gentgeen__ has joined #ltsp
23:30alkisg has quit IRC