00:11 | alkisg (~alkisg@ubuntu/member/alkisg) joined #ltsp. | |
00:30 | vmlintu (~vmlintu@nblzone-240-143.nblnetworks.fi) left irc: Remote host closed the connection | |
00:58 | MorningSon (~MorningSo@cpe-70-114-21-95.satx.res.rr.com) left irc: Quit: WeeChat 0.3.0 | |
01:18 | dobber (~dobber@89.190.197.77) joined #ltsp. | |
01:27 | jkinn2 (~jason@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) joined #ltsp. | |
01:31 | cyberorg (~cyberorg@opensuse/member/Cyberorg) left irc: Quit: Leaving | |
01:33 | cyberorg (~cyberorg@opensuse/member/Cyberorg) joined #ltsp. | |
01:42 | otavio (~otavio@debian/developer/otavio) left irc: Ping timeout: 250 seconds | |
01:55 | Kicer86 (~Kicer86@host-5db0eeee.sileman.net.pl) joined #ltsp. | |
01:57 | otavio (~otavio@debian/developer/otavio) joined #ltsp. | |
02:03 | gnunux (~emmanuel@194.167.18.244) joined #ltsp. | |
02:04 | <gnunux> hi
| |
02:33 | alexqwesa (~alex@109.172.15.11) left irc: Read error: Operation timed out | |
02:35 | alexqwesa (~alex@109.172.15.11) joined #ltsp. | |
03:05 | no_mind_ (7aa3f301@gateway/web/freenode/ip.122.163.243.1) joined #ltsp. | |
03:05 | <no_mind_> ping anyone around ?
| |
03:11 | ogra (~ogra@ubuntu/member/ogra) left irc: Excess Flood | |
03:11 | ogra (~ogra@p5098ed03.dip0.t-ipconnect.de) joined #ltsp. | |
03:11 | Nick change: ogra -> Guest78603 | |
03:11 | <alkisg> !question
| |
03:12 | Guest78603 (~ogra@p5098ed03.dip0.t-ipconnect.de) left irc: Changing host | |
03:12 | Guest78603 (~ogra@ubuntu/member/ogra) joined #ltsp. | |
03:12 | <alkisg> Hmmm bot's still down
| |
03:12 | Nick change: Guest78603 -> ogra_ | |
03:12 | <alkisg> "Don't ask to ask a question, just ask it"
| |
03:25 | alkisg (~alkisg@ubuntu/member/alkisg) left irc: Quit: Leaving. | |
03:50 | Nick change: zz_evil_root -> evil_root | |
03:51 | Nick change: evil_root -> zz_evil_root | |
03:51 | Nick change: zz_evil_root -> evil_root | |
03:52 | artista-frustrad (~artista_f@187.7.151.137) joined #ltsp. | |
03:53 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) joined #ltsp. | |
03:54 | artista_frustrad (~artista_f@187.7.151.137) left irc: Ping timeout: 250 seconds | |
04:00 | artista-frustrad (~artista_f@187.7.151.137) left irc: Ping timeout: 250 seconds | |
04:10 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) left irc: Quit: I Leave | |
04:11 | wima1 (~wima@91.183.60.122) joined #ltsp. | |
04:11 | <no_mind_> I have an ltsp setup on ubuntu 10.04 . What I am noticing for few days that suddenly the performance of the box degrades. All clients are slow to respond. This happens randomly any time during the day. Reboot fixes the problem. How do I find the cause of performance degrade
| |
04:14 | artista-frustrad (~artista_f@187.7.151.137) joined #ltsp. | |
04:21 | F-GT (~phantom@ppp121-44-126-79.lns20.syd6.internode.on.net) left irc: Quit: Leaving | |
04:21 | artista-frustrad (~artista_f@187.7.151.137) left irc: Ping timeout: 255 seconds | |
04:21 | F-GT (~phantom@ppp121-44-126-79.lns20.syd6.internode.on.net) joined #ltsp. | |
04:23 | <wima1> no_mind_: wat doe top tell you? is the load very high, is IO wait high? ...
| |
04:23 | alkisg (~alkisg@ubuntu/member/alkisg) joined #ltsp. | |
04:24 | <no_mind_> wima1: top reports load between 0.8 to 1.2
| |
04:24 | <wima1> in general, or when it is going slow?
| |
04:25 | <no_mind_> in general or when it is going slow
| |
04:25 | it is the clients that go slow
| |
04:25 | server works fine
| |
04:25 | how do I find out IO wait ?
| |
04:25 | <wima1> in top, 3rd line. %wa
| |
04:26 | it shows the nr of processes in wait. which is usually IO
| |
04:27 | should be very low in generql
| |
04:28 | if it gets higher than ~10%, you will see the load climbing up very fast
| |
04:28 | <no_mind_> its varrying between 1.2 - 1.4
| |
04:28 | ni is 93%
| |
04:29 | oops ni is 0% but id is over 90%
| |
04:30 | msignor_work (~test@24.115.67.19.res-cmts.sesp.ptd.net) joined #ltsp. | |
04:31 | <wima1> idle, that's fine :)
| |
04:32 | and the clients are very slow at this very moment?
| |
04:34 | artista-frustrad (~artista_f@187.7.151.137) joined #ltsp. | |
04:39 | gnunux (~emmanuel@194.167.18.244) left irc: Ping timeout: 252 seconds | |
04:44 | <msignor_work> Anyone able to steer me in a direction to figure out why no one can login to LDM other than root?
| |
04:44 | <no_mind_> wima1: yes clients are very slow
| |
04:45 | wima1: I have 32 GB RAm with total of 16 cores
| |
04:54 | gnunux (~emmanuel@194.167.18.244) joined #ltsp. | |
04:57 | <alkisg> msignor_work: is that the server root account, or the client root account?
| |
04:57 | (you use different passwords on those 2? if not try to temporarily change the root password on the server)
| |
04:57 | And check the LTSP_FATCLIENT=False method I suggested yesterday
| |
04:57 | bbl
| |
04:58 | alkisg (~alkisg@ubuntu/member/alkisg) left irc: Quit: Leaving. | |
05:11 | cVsup (~cVsup@189.83.156.119) joined #ltsp. | |
05:12 | <cVsup> how can define packages to pxe clients?
| |
05:19 | how can select packages to pxe?
| |
05:30 | how can select packages to pxe??
| |
05:30 | Last message repeated 1 time(s). | |
05:30 | <cVsup> ltsp 4 have ltspadmin, in 5.1 what tool?
| |
05:46 | <wima1> no_mind_: sorry, i was away from my desk for a while... Anyway, if there is no problem with the server, and periodically all clients become very slow, I can only imagine it to be a network problem
| |
05:49 | <cVsup> hey somebody can help?
| |
05:51 | <muppis> cVsup, what you're actual looking for?
| |
05:51 | <cVsup> muppis i need select packages to boot pxe
| |
05:52 | <muppis> cVsup, so your client nic doesn't support pxe booting?
| |
05:54 | <cVsup> yes support
| |
05:54 | i need select packages to pxe
| |
05:54 | <muppis> I just don't follow you. To do what?
| |
05:56 | <cVsup> exist method to select packages in tftp image?
| |
05:57 | <muppis> Trying to install applications to be runned in client?
| |
05:57 | <cVsup> yes
| |
05:58 | <muppis> This should help to start: https://help.ubuntu.com/community/UbuntuLTSPKarmicLocalAppsFirefox
| |
06:00 | <cVsup> hm, thanks muppis
| |
06:01 | Gadi (~romm@ool-18bbe47a.static.optonline.net) joined #ltsp. | |
06:01 | ggsict (~GRAVESEND@194.80.20.10) joined #ltsp. | |
06:02 | <msignor_work> Are users on the "Server" considered to be the same users present on the "client"
| |
06:02 | i am having difficulty understanding if there are actually separate users I need to maintain
| |
06:03 | <muppis> msignor_work, you should not have any users in client, if you haven't purposly created one under chroot.
| |
06:04 | <msignor_work> muppis: ok - thats what I thought. alkisg has asked me to verify if I am loggin in as root on the client of server
| |
06:04 | ggsict (~GRAVESEND@194.80.20.10) left irc: Client Quit | |
06:04 | <muppis> Having printing problem. Locally runned Firefox switches to Letter instead A4 used here.
| |
06:04 | <msignor_work> (root can login to x, no one else can) - not sure how to tell what they logged into
| |
06:05 | sigh, nevermind - root@ltspXXX# is pretty obvious :)
| |
06:05 | <muppis> :)
| |
06:10 | <msignor_work> muppis: you know where I can enable more logging for ltsp? specifically so I can look at what kind of errors may be happening to prevent a login?
| |
06:12 | <wima1> i'm trying to set up a secondary dhcp server on my network, but I see very strange results...
| |
06:12 | eqch dhcp server has it's own pool of ip addresses
| |
06:13 | and each dhcp server uses itself as an application server
| |
06:14 | now it can happen that the pxe rom gets an address from a different server then the ltsp image
| |
06:14 | if this happens, the mouse doesn't work!
| |
06:14 | all the rest works fine...
| |
06:15 | ideas anyone?
| |
06:16 | Trixboxer (~Trixboxer@office.supportdepartment.net) joined #ltsp. | |
06:16 | <msignor_work> do I *need* xorg and a window manager on my Server if I am using fatclients? no right?
| |
06:17 | <muppis> msignor_work, you can think fat clients as local installation, only loaded from over network.
| |
06:18 | msaul_ (~administr@CPE001cf0f564ef-CM0011ae00c1f8.cpe.net.cable.rogers.com) joined #ltsp. | |
06:19 | <msaul_> Hi, was wondering if I could get help ssh ing from server to client.
| |
06:20 | I can ssh from client to server, but other way get connection resfused on port 22
| |
06:20 | I'm using Edubuntu 10.10 LTSP
| |
06:21 | <Appiah> did you enable ssh server in the chroot?
| |
06:22 | and remeber to set a root passwd
| |
06:24 | <msaul_> That would probably be the problem. Let me check that out...
| |
06:30 | <msignor_work> Now I have taken a step backwards - I can not login as anyone now.. just says "no response from server: restarting" any suggestions?
| |
06:31 | <Appiah> msignor_work: what did you change?
| |
06:31 | like IP...
| |
06:32 | <msignor_work> honestly - i edited lts.con to set my fatclient to false for further troubleshooting, per another user here
| |
06:32 | when it broke, i undid it and now its the same - lol. i see the request initiated in the SSH log on the server, just nothing else past that
| |
06:33 | <Appiah> well check the logs
| |
06:33 | msaul_ (~administr@CPE001cf0f564ef-CM0011ae00c1f8.cpe.net.cable.rogers.com) left irc: Quit: Ex-Chat | |
06:33 | <Appiah> ltsp-update-sshkeys might help too
| |
06:35 | <msignor_work> appiah: do i need to be concerned with this - mountd[2034]: could not open /var/lib/nfs/.rmtab.lock for locking: errno 30 (Read-only file system)
| |
06:35 | this is following the successful mount statement for my NFS home folder
| |
06:44 | msignor_work (~test@24.115.67.19.res-cmts.sesp.ptd.net) left irc: Quit: Leaving | |
06:44 | msignor_work (~test@24.115.67.19.res-cmts.sesp.ptd.net) joined #ltsp. | |
07:02 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) joined #ltsp. | |
07:09 | <wima1> if i run ltsp-update-sshkeys, should i then also do a ltsp-update-image?
| |
07:09 | MorningSon (~MorningSo@cpe-70-114-21-95.satx.res.rr.com) joined #ltsp. | |
07:10 | matrix3000 (~matrix@rrcs-70-61-255-226.central.biz.rr.com) joined #ltsp. | |
07:13 | alkisg (~alkisg@ubuntu/member/alkisg) joined #ltsp. | |
07:25 | <msignor_work> Does LTSP automatically assume the SSH server its connecting to is the host given during the DHCP initialization for boot? or is it in a config file somewhere?
| |
07:27 | <matrix3000> i thought it was a tftp config file?
| |
07:28 | <Gadi> msignor_work: by default, it will assume the same as the *nbd* server
| |
07:28 | msignor_work: you can set it in lts.conf with LDM_SERVER
| |
07:29 | <msignor_work> gadi: thanks, something is goofed up - i jumped the gun though and updated the ssh keys again and image - so it will be another hour before I can test again -- thanks again
| |
07:32 | <Gadi> msignor_work: not sure how much of the theory of operation you know already, but jic: when you log into ldm on the client, it sets up an ssh tunnel to the server using the creds you provide. Then, for local apps/fat client, it uses the tunnel to grab the auth info, so you never need to make users in the chroot
| |
07:33 | msignor_work: if you need to debug, you can set a "shell" on one screen and "ldm" on another. The shell will drop you to a sh prompt as root on the client
| |
07:33 | <cVsup> when start dbus on ltsp client return error
| |
07:33 | Could not create file /var/lib/dbus/machine-id:
| |
07:34 | Could not create file /var/lib/dbus/machine-id: Read-only file system
| |
07:34 | <Gadi> cVsup: are you on debian?
| |
07:34 | <cVsup> yes
| |
07:34 | <Gadi> debian uses NFS for the chroot and bind-mounts for rw dirs
| |
07:35 | <alkisg> [06:14] *** [wima1] if this happens, the mouse doesn't work! ==> that's usually when you send a different kernel with tftp, than the one on nfs/nbd
| |
07:35 | <Gadi> it therefore specifies all rw dirs
| |
07:35 | <cVsup> server:~/rdesktop-1.6.0#
| |
07:35 | ops
| |
07:35 | <Gadi> cVsup: if you have additional dirs that you need to be read-write, you need to edit a config file in the chroot
| |
07:35 | <wima1> ah, yes. that makes sense. Thanks!!
| |
07:36 | <cVsup> Gadi in lts.conf?
| |
07:36 | <Gadi> cVsup: no, it will be in the chroot
| |
07:36 | <alkisg> [06:05] *** [msignor_work] sigh, nevermind - root@ltspXXX# is pretty obvious :) ==> no it isn't, that's your session, I was asking about authentication, to see if ldm actually contacts the server or not
| |
07:36 | <Gadi> look for /opt/ltsp/i386/etc/..... I forget offhand the filename (I dont use debian)
| |
07:37 | !debian
| |
07:37 | damn
| |
07:37 | <alkisg> !no_bot_available :)
| |
07:37 | <Gadi> and there's no bot
| |
07:37 | <alkisg> Scotty said he'd resurrect it 1 week ago
| |
07:40 | [06:35] *** [msignor_work] appiah: do i need to be concerned with this - mountd[2034]: could not open /var/lib/nfs/.rmtab.lock for locking: errno 30 (Read-only file system) ==> yes, that means you don't use the normal ubuntu/nbd chroot anymore, something from your debian/nfs got mixed to it
| |
07:42 | <cVsup> Gadi
| |
07:42 | wima1 (wima@91.183.60.122) left #ltsp. | |
07:43 | <cVsup> i look for mtab and fstab but nothing found
| |
07:43 | chupacabra (~chupacabr@cpe-70-112-10-77.austin.res.rr.com) joined #ltsp. | |
07:46 | Nick change: ogra_ -> ogra | |
07:52 | <Gadi> cVsup: it looks like these are the defaults: rw_dirs="/var/lib/xkb /var/log /var/spool /var/tmp /tmp /etc/console-setup /var/
| |
07:52 | lib/pulse /var/lib/dbus /var/cache/hald /var/cache/ltsp /var/lib/urandom"
| |
07:52 | and you can override in lts.conf with: LTSP_RW_DIRS
| |
07:53 | there's also:
| |
07:53 | copy_dirs="/root /home /var/cache/ltsp-localapps /etc/rsyslog.d /etc/cups /media
| |
07:53 | /etc/cron.d /etc/udev/rules.d"
| |
07:53 | and bindfiles="/etc/network/interfaces /etc/hostname /etc/hosts /etc/syslog.conf /et
| |
07:53 | c/fstab /etc/resolv.conf /etc/X11/xorg.conf /etc/passwd /etc/group /etc/localtim
| |
07:53 | e"
| |
07:53 | (see: /opt/ltsp/i386/usr/share/ltsp/ltsp-bindmounts)
| |
07:56 | dobber (~dobber@89.190.197.77) left irc: Ping timeout: 276 seconds | |
08:06 | <msignor_work> alkisg: this happened last night after I got home and changed the LTSP_FATCLIENT to false, and commented out the NFS info
| |
08:07 | alkisg: I don't have X installed on the server -- so I assume that is why it didn't work
| |
08:07 | when i un-did it however.. then it really broke
| |
08:08 | i still have my original ubuntu image from when you and I talked a few days ago (i made a backup)
| |
08:08 | <alkisg> msignor_work: can you test some things right now?
| |
08:10 | toscalix (~toscalix@213.99.235.34) joined #ltsp. | |
08:12 | Roasted (~jason@206.82.22.190) joined #ltsp. | |
08:12 | Roasted (~jason@206.82.22.190) left irc: Changing host | |
08:12 | Roasted (~jason@unaffiliated/roasted) joined #ltsp. | |
08:12 | test__ (~test@24.115.67.19.res-cmts.sesp.ptd.net) joined #ltsp. | |
08:12 | Nick change: test__ -> msignor_work_1 | |
08:12 | <msignor_work_1> alkisg: sry - killed my net connection when i turned on VPN
| |
08:13 | <alkisg> (05:08:49 PM) alkisg: msignor_work: can you test some things right now?
| |
08:13 | msignor_work (~test@24.115.67.19.res-cmts.sesp.ptd.net) left irc: Ping timeout: 260 seconds | |
08:13 | Nick change: msignor_work_1 -> msignor_work | |
08:13 | <msignor_work> alkisg: image is at 55% building. Also, I forgot to mention that I added the root_squash_fs, and I don't have the error anymore with the NFS lock
| |
08:14 | alkisg: If I can use the current image while the new one is building then I do whatever I need to now
| |
08:14 | <alkisg> msignor_work: login as root since you can't do anything else
| |
08:14 | And pastebin a `ps aux` from your server and one from your client
| |
08:14 | If you don't want to show them at public, PM me with the link
| |
08:15 | <msignor_work> root will not let me login now..
| |
08:15 | let me make sure I didn't typo anything in lts.conf... thats the only file I edited other than my exports
| |
08:18 | no_mind_ (7aa3f301@gateway/web/freenode/ip.122.163.243.1) left irc: Ping timeout: 252 seconds | |
08:21 | dobber (~dobber@213.169.45.222) joined #ltsp. | |
08:22 | <msignor_work> alkisg: http://pastebin.com/5JYEUz11 <-- server
| |
08:23 | alkisg: client wont let me do anything now.. just says "no response from server, restarting"
| |
08:23 | Roasted (~jason@unaffiliated/roasted) left irc: Ping timeout: 255 seconds | |
08:24 | <alkisg> msignor_work: put SCREEN_02=shell and SCREEN_07=ldm in lts.conf so that you have a shell in vt2
| |
08:26 | msignor_work: your LDM_SSHOPTIONS are wrong
| |
08:26 | They break your login
| |
08:27 | <msignor_work> aaahhhhh - ok.. makes sense.. what did I have wrong?
| |
08:27 | <alkisg> An unfinished -o at the end
| |
08:27 | LDM_SSHOPTIONS="-o StrictHostKeyChecking=no -o CheckHostIP=no -o LogLevel=silent"
| |
08:27 | <msignor_work> ahhhh crap. i must of hit enter or something and clipped the end of it let me try that
| |
08:30 | alkisg: I owe you a beer or 3.
| |
08:30 | <alkisg> :)
| |
08:31 | <msignor_work> issue was 2 fold . I was not able to login as my non root user becaue I already had an X config in the folder
| |
08:31 | then in my attempt to troubleshoot i deleted half the line in lts
| |
08:31 | but i logged in as a non root user just now
| |
08:37 | <alkisg> (and I was wondering why you needed to update keys + image ... :))
| |
08:37 | alkisg (~alkisg@ubuntu/member/alkisg) left irc: Quit: Leaving. | |
08:39 | dobber (~dobber@213.169.45.222) left irc: Ping timeout: 276 seconds | |
08:44 | komunista (~slavko@adsl-195-098-015-248.dynamic.nextra.sk) joined #ltsp. | |
08:45 | cliebow (~cliebow@WatchGuard.ellsworth-hs.ellsworth.k12.me.us) joined #ltsp. | |
08:45 | dobber (~dobber@78.90.16.77) joined #ltsp. | |
08:51 | primeministerp (~ppouliot@64.119.153.82) joined #ltsp. | |
08:55 | msignor_work (~test@24.115.67.19.res-cmts.sesp.ptd.net) left irc: Quit: be back later | |
08:55 | cVsup (~cVsup@189.83.156.119) left irc: Ping timeout: 252 seconds | |
08:57 | cVsup (~cVsup@189.83.175.189) joined #ltsp. | |
09:03 | <cVsup> Gadi these files opt/ltsp/i386/usr/share/ltsp/ltsp-bindmounts none exist
| |
09:03 | patrickmw (~pwright@ip68-231-4-153.ph.ph.cox.net) left irc: Quit: Ex-Chat | |
09:05 | <cVsup> hmm found
| |
09:05 | msaul_ (~administr@CPE001cf0f564ef-CM0011ae00c1f8.cpe.net.cable.rogers.com) joined #ltsp. | |
09:05 | <cVsup> . /opt/ltsp/i386/etc/default/ltsp-client-setup
| |
09:06 | <msaul_> I needed to install ssh in chroot, then I use apt-get ssh
| |
09:06 | I get error: Err file:/media/cdrom/ maverick/main openssh-server amd64 1:5.5p1-4ubuntu4
| |
09:06 | File not found
| |
09:07 | I know path is not correct as per my CDROM on my server, was wondering if any solution...
| |
09:16 | I can use apt-get to install other apps, so I was curious for a "work-around" to install it...
| |
09:16 | jkinn2 (~jason@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) left irc: Ping timeout: 252 seconds | |
09:17 | jkinn2 (~jason@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) joined #ltsp. | |
09:19 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) left irc: Read error: Connection reset by peer | |
09:21 | cVsup (~cVsup@189.83.175.189) left irc: Ping timeout: 255 seconds | |
09:21 | cVsup (cVsup@189-91-13-13.cpe.dvl.mastercabo.com.br) joined #ltsp. | |
09:22 | drdata (~drdata@tw2.udallas.edu) joined #ltsp. | |
09:22 | toscalix (~toscalix@213.99.235.34) left irc: Remote host closed the connection | |
09:23 | patrickmw (~pwright@ip68-231-4-153.ph.ph.cox.net) joined #ltsp. | |
09:24 | gnunux (~emmanuel@194.167.18.244) left irc: Quit: Ex-Chat | |
09:25 | cVsup (cVsup@189-91-13-13.cpe.dvl.mastercabo.com.br) left irc: Ping timeout: 246 seconds | |
09:26 | cVsup (~cVsup@189.12.14.211) joined #ltsp. | |
09:30 | rjune (~rjune@oh-67-77-122-143.sta.embarqhsd.net) joined #ltsp. | |
09:40 | mistik1 (mistik1@unaffiliated/mistik1) left irc: Read error: Connection reset by peer | |
09:40 | mistik1 (mistik1@unaffiliated/mistik1) joined #ltsp. | |
09:41 | <cVsup> sombody can use rdesktop with smartcard?
| |
09:42 | msaul_ (~administr@CPE001cf0f564ef-CM0011ae00c1f8.cpe.net.cable.rogers.com) left irc: Quit: Ex-Chat | |
09:54 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) joined #ltsp. | |
09:58 | dobber (~dobber@78.90.16.77) left irc: Remote host closed the connection | |
10:04 | drdata (~drdata@tw2.udallas.edu) left irc: Ping timeout: 252 seconds | |
10:08 | jsass (~jsass@fw.acurrus.com) joined #ltsp. | |
10:08 | <jsass> hello everyone!
| |
10:09 | <evil_root> hi jsass
| |
10:09 | <jsass> whoops :)
| |
10:09 | Nick change: jsass -> _UsUrPeR__ | |
10:09 | <evil_root> lol
| |
10:09 | <_UsUrPeR__> on another computer >_>
| |
10:10 | <evil_root> use a bouncer ;)
| |
10:10 | <_UsUrPeR__> I am afraid to admit this, but I am too lazy to ghost my other account
| |
10:10 | <evil_root> thats funny as hell
| |
10:11 | <_UsUrPeR__> ok, so I am trying to get a DVI output working with a VIA VX800 chipset.
| |
10:11 | I swear that I had it working two days ago with DVI, but that seems to be a fading memory now
| |
10:11 | what I am seeing now is a horribly unusable and distorted screen when the client gets to ldm
| |
10:12 | <evil_root> thats weird as hell, you using dvi to dvi or are you converting dvi to vga or hdmi?
| |
10:13 | <_UsUrPeR__> HOWEVER, if I put the "X_OPTION_01 = "\"ForcePanel\" \"true\"" option in, the client works properly at a 640x480 resolution
| |
10:14 | evil_root: naw, VIA chipsets and chrome 9 hardware always have given issues
| |
10:14 | err "been issues" I should say
| |
10:14 | though typically there's some weird workaround, I am having problems finding one on my own this time.
| |
10:15 | <evil_root> yea sorry but i have never herd of it
| |
10:19 | shawnp0wers (~Adium@linuxjournal/staff/shawnp0wers) joined #ltsp. | |
10:20 | <Gadi> _UsUrPeR__: take out that option and try setting resolution with: XRANDR_MODE_0
| |
10:23 | <_UsUrPeR__> Gadi: one moment
| |
10:23 | shawnp0wers (~Adium@linuxjournal/staff/shawnp0wers) left irc: Client Quit | |
10:25 | <_UsUrPeR__> gadi: without force_panel enabled, the screen appears to have an improper refresh rate
| |
10:25 | everything is illegible
| |
10:27 | <Gadi> _UsUrPeR__: do you use a panel?
| |
10:27 | if not, dont force a panel
| |
10:27 | <_UsUrPeR__> gadi: yes. Plugged in with DVI
| |
10:28 | <Gadi> thats not a panel
| |
10:28 | <_UsUrPeR__> ?
| |
10:28 | <Gadi> panel == lvds panel
| |
10:28 | <_UsUrPeR__> oh. OH
| |
10:28 | well no, certainly not then :
| |
10:28 | )
| |
10:28 | <Gadi> the only reason you get 640x480 is that you now have an additional xrandr channel
| |
10:28 | and it is using its modes
| |
10:29 | can you get to a legible shell?
| |
10:29 | or ssh in?
| |
10:29 | <_UsUrPeR__> Gadi: no. I had to set up SSH
| |
10:29 | but I can SSH in
| |
10:29 | alkisg (~alkisg@ubuntu/member/alkisg) joined #ltsp. | |
10:29 | <Gadi> if you ssh in, take a look at:
| |
10:29 | 1. xorg log file
| |
10:29 | cVsup (~cVsup@189.12.14.211) left irc: Ping timeout: 260 seconds | |
10:30 | <Gadi> 2. export DISPLAY and XAUTHORITY to the proper values and run xrandr
| |
10:30 | cVsup (~cVsup@189.12.14.211) joined #ltsp. | |
10:30 | <Gadi> 1. will tell you what modes/frequencies it is using
| |
10:30 | 2. will tell you what outputs it sees
| |
10:30 | <_UsUrPeR__> interesting results I just had: http://pastebin.com/QNrRvama
| |
10:30 | this particular lts.conf works
| |
10:30 | just got it working.
| |
10:31 | <JamesB192> (silly)I don't suppose it could be a flat panel attached to the dvi plug at the back of his box but connected via the analog signal lines(/silly)
| |
10:32 | <_UsUrPeR__> like a DVI > VGA adapter?
| |
10:32 | that's typically something that has not worked well... at least for me.
| |
10:33 | <alkisg> Hi all. _UsUrPeR__: is the new ltsp logo released under some specific license, so that it can be included in some distro themes etc?
| |
10:34 | * _UsUrPeR__ runs away to ask his co-workers while waving his hands like Grover | |
10:34 | <alkisg> :)
| |
10:36 | msignor_work (~test@24.115.67.19.res-cmts.sesp.ptd.net) joined #ltsp. | |
10:36 | Lns (~Lns@pdpc/supporter/professional/lns) joined #ltsp. | |
10:39 | <_UsUrPeR__> alkisg: go for it. Alexc says "that's cool"
| |
10:40 | <alkisg> _UsUrPeR__: unfortunately it needs to be written down somewhere to be included in a distro
| |
10:40 | Maybe put a small paragraph in the ltsp.org site somewhere
| |
10:40 | Or a .zip file that contains the .png or .svg an a small readme
| |
10:41 | <_UsUrPeR__> alkisg: oh, in that case, I'll let him know that he needs to ccl it. Use it forthwith. We will make that change shortly.
| |
10:41 | <alkisg> Nice. I'll try to make an ldm theme that has the logo on it
| |
10:42 | As the old logo with the yellow letters completely sucks :)
| |
10:42 | *the old theme
| |
10:48 | <highvoltage> alkisg: I almost finished a debian ldm theme yesterday (going to give it some polish tomorrow) and hopefully vagrant can sponsor it :)
| |
10:49 | jkinn2 (~jason@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) left irc: Ping timeout: 246 seconds | |
10:49 | <alkisg> highvoltage: cool!
| |
10:49 | Let us have a peek when you finish it
| |
10:51 | <msignor_work> Is it possible / advisable to install new packages when logged into a terminal as root?
| |
10:51 | <_UsUrPeR__> msignor_work, in a terminal chroot, or a booted terminal?
| |
10:51 | <msignor_work> booted terminal
| |
10:52 | i just want to try and avoid having to remake the image every time, heh
| |
10:52 | <alkisg> They'll get lost on reboot
| |
10:52 | <_UsUrPeR__> msignor_work: as long as the client will resolve correctly via DNS, the file system is temporary. If the program you are installing is small, it should work
| |
10:52 | though alkisg is correct: your changes will be lost on reboot
| |
10:53 | <alkisg> msignor_work: then switch to nfs ;)
| |
10:53 | <msignor_work> ah, ok. well seems reasonable enough. I figured that was the case :-)
| |
10:53 | lol
| |
10:53 | <alkisg> It'll be 3 times slower (if you have compression enabled, that is) but you won't have to rebuild the image on each change
| |
10:54 | <msignor_work> yeah, well i think the tradeoff for performance is better
| |
10:54 | my server is only 500mhz so it takes a while
| |
10:54 | although, i am sure I could do it on another machine and copy it
| |
10:54 | <alkisg> Mount it with nfs or sshfs and do it on a remove cpu
| |
10:55 | So that you don't have to transfer it
| |
10:55 | <msignor_work> wow, even better point
| |
10:55 | <alkisg> E.g. you can use your fat client cpu
| |
10:55 | ;)
| |
10:56 | *remote
| |
10:56 | <msignor_work> do I need to worry about any of the kernel or ssh key stuff? should just be a mount and then run the command correct?
| |
10:56 | <alkisg> with the ldm_sshoptions line you're using, you don't need to worry about keys
| |
10:56 | (and you wouldn't need to, anyway)
| |
10:56 | <msignor_work> ah, true. cool - sounds like a good solution then
| |
10:57 | I've got to say - this is pretty darn cool
| |
10:57 | <alkisg> Mount with nfs (don't forget no_root_squash) and you're ready to go
| |
10:59 | <msignor_work> yeap - just added that to all my exports
| |
10:59 | bbl - thanks again!
| |
10:59 | msignor_work (~test@24.115.67.19.res-cmts.sesp.ptd.net) left irc: Quit: Leaving | |
11:00 | drdata (~drdata@tw2.udallas.edu) joined #ltsp. | |
11:02 | roasted___ (4bc3a2d8@gateway/web/freenode/ip.75.195.162.216) joined #ltsp. | |
11:03 | <roasted___> Question - I have an old celeron box here with 128 RAM I'm trying to connect to the LTSP box. It succeeds, but if I open a lot of applications at once, it crashes and goes back to the log in screen. Likewise, I have a P4 box with 128 ram that does NOT exhibit this issue.
| |
11:04 | <alkisg> Try putting NBD_SWAP=True in lts.conf
| |
11:04 | <roasted___> Are you thinking it's due to it only having 128 ram?
| |
11:04 | <alkisg> Yes
| |
11:05 | <roasted___> I had 256 in it and it worked fine, but I wanted to give it as little RAM as possible to see how it works.
| |
11:05 | If it's cause of the RAM, why do you think the other P4 128mb box works fine?
| |
11:05 | <alkisg> Not shared video ram? Local swap partition? Opening different programs that don't require X pixmaps?
| |
11:05 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) left irc: Remote host closed the connection | |
11:06 | <roasted___> I'm opening the exact same programs. I have desktop icons for the same apps on each, I just select all and hit open. I just wanted to try and "bench" what this LTSP setup can handle with the limited hardware I have on the client ends.
| |
11:06 | LoveStorm (storm@thepcrepair.info) joined #ltsp. | |
11:06 | <roasted___> It's firefox, dictionary, calculator, libre office writer/calc/presentation.
| |
11:06 | <alkisg> I've seen firefox and openoffice needing 512 ram on a thin client
| |
11:07 | <roasted___> So each one I highlight and bam - open all at once. P4 box works fine, the other does not. Besides the proc differences (celeron vs P4) I wasnt sure why the P4 box NEVER failed yet the celly did.
| |
11:07 | <alkisg> So it crashed until I gave them 512 nbd swap
| |
11:07 | <roasted___> so it going back to the login screen is just a sign my thin client hardware is too limited
| |
11:07 | and its fallback is to bail out and get to login screen?
| |
11:08 | Yeah you called it. Looks like it's Firefox.
| |
11:08 | I can open dictionary, calc, etc one at a time. Once I hit FF it locks.
| |
11:08 | <alkisg> It depends on the page you visit
| |
11:08 | <roasted___> Just kind of surprised the other box works fine. But it IS different hardware.
| |
11:09 | It's just our district home page on both boxes.
| |
11:09 | ha, it loaded
| |
11:09 | just took forever
| |
11:09 | <Lns> roasted__, to verify, use xrestop on the client to monitor the local ram usage
| |
11:10 | <roasted___> Lns: I will. I just locked it up again...
| |
11:10 | <Lns> but FF/OOo are huge resource hogs, even on an ltsp client. trust me. i've dealt with it a lot ;)
| |
11:11 | <roasted___> Yeah. Even still, FF/ Libre OFfice writer/excel/powerpoint were opened on BOTH. and BOTH had 128 ram. that's what htrew me off.
| |
11:11 | <Lns> roasted__, what gets them to lock up is when you start to actually load the local xpixmap memory up with images.
| |
11:12 | such as loading websites with gfx, or inserting clipart into OOo document (and even just moving it around a bit)
| |
11:12 | lots of times it won't let go of the pixmap memory, clogging it up until the machine locks.
| |
11:12 | <roasted___> where within xrestop am I looking?
| |
11:14 | <Lns> roasted__, pxm mem i believe
| |
11:14 | or just total..for the procs you're monitoring
| |
11:15 | with usage and that terminal open, you'll see the usage climb steadily
| |
11:16 | <roasted___> I'm a little confused though. Is it loading this stuff to the local ram on the client?
| |
11:16 | I thought it used the server ram unles otherwise specified
| |
11:18 | <Lns> roasted__, for pixmaps, a lot of programs will use local video/memory to store images (imagine if every screen/pixel update required it to travel through the network)
| |
11:19 | <_UsUrPeR__> alkisg: about the nbd_swap=true option: does that require a file in /etc/ltsp/nbdswapd.conf?
| |
11:19 | ...or is that just to expand the amount of swap space on the client?
| |
11:19 | I am unsure about what the standard swap space is...
| |
11:19 | <alkisg> _UsUrPeR__: yes, the default without that file is 64 I think
| |
11:19 | Used to be 32
| |
11:19 | * alkisg would like to raise that to 512, since it's sparse files anyway | |
11:20 | <alkisg> sparse==they don't take up space unless they're actually used
| |
11:20 | <_UsUrPeR__> roasted__: for the record, the nomenclature inside that file is: "SIZE=128" or whatever the size of the swap file you want in there is in megs
| |
11:20 | (without quotes)
| |
11:21 | * Lns waves to everyone in #ltsp as it's been a few months since he's been online here =) | |
11:21 | <_UsUrPeR__> hey dude. how's it going?
| |
11:21 | <Lns> It's going well _UsUrPeR_ =) how have you been?
| |
11:22 | <_UsUrPeR__> well enough. Things are weird over here at the moment for me at the moment. Bleh :P
| |
11:22 | VIA chipsets
| |
11:22 | <Lns> weird? hehe
| |
11:22 | if it weren't weird, it wouldn't be interesting.
| |
11:22 | nesusvet (~nesusvet@195.218.191.242) joined #ltsp. | |
11:23 | <_UsUrPeR__> Yeah! New, strange challenges.
| |
11:23 | like some sort of puzzle...
| |
11:23 | horrible, horrible puzzle ;)
| |
11:23 | <Lns> ha! =p
| |
11:24 | alkisg, how are you? =)
| |
11:24 | <alkisg> Hey Lns, long time no see
| |
11:24 | <Lns> i know =) everything going good for you?
| |
11:25 | <alkisg> Yup. Still working with LTSP? Or not so much anymore?
| |
11:25 | <nesusvet> hello guys :). I have one question. If i will build client with ---fat-client. Would an LTSP image be builded with X11 packages or not?
| |
11:25 | <alkisg> nesusvet: yes
| |
11:25 | <Lns> not a lot aside from my own personal setups. i'm hoping that will change, but atm I'm working on a prototype for a linux-ready mini-itx pc to offer the world
| |
11:25 | <nesusvet> How to disable X11 at all before installation?
| |
11:26 | <alkisg> nesusvet: you want a fat client without X?
| |
11:26 | <Lns> nesusvet, why install fat-client if you don't want x11?
| |
11:26 | seems strange to me
| |
11:26 | <alkisg> Like a cluster for computations or something?
| |
11:28 | <nesusvet> Because i need only linux environment. I am using LTSP in order to start big clusters :). But i change many packages under chroot environment, after installation. And X11 is a little annoy :)
| |
11:28 | <alkisg> fat clients == they login locally with ldm
| |
11:28 | Since you won't have X, you won't use ldm, so you can't use authentication from server with ldm+ssh
| |
11:29 | So you can just install any packages you want on a thin chroot
| |
11:29 | And handle authentication manually
| |
11:29 | <nesusvet> yes, I wrote to this chat 6 month ago, and we found out all solution which i need, but maybe something was changed.
| |
11:30 | alkisg, thanks, anyway.
| |
11:33 | by the way, how many machines i could load via NBD?
| |
11:34 | <alkisg> I'm not aware of any hard limits
| |
11:34 | <nesusvet> ok.
| |
11:35 | <roasted___> sorry was afk
| |
11:38 | So by default, if I boot a client with 128 ram, does it actually utilize that ram for local processing?
| |
11:38 | Like OOo and FF load right to that RAM and not the server RAM?
| |
11:41 | <Lns> roasted__, no. FF/OOo will still load on the server unless you specify it as a localapp. The difference is that those programs use local pixmap caching, which utilizes local client memory.
| |
11:42 | <roasted___> Lns: so realistically, if I make OOo and FF local apps, Id make everything worse. Right?
| |
11:42 | assuming I stay at 128 ram
| |
11:42 | <Lns> uhhh, yeah.
| |
11:42 | it probably won't work at all if you try that. =p
| |
11:42 | cVsup (~cVsup@189.12.14.211) left irc: Read error: Connection reset by peer | |
11:43 | <Lns> I've tried localapp firefox on an atom 1.6GHz cpu with a gig of ram, and it's still hard to use.
| |
11:43 | <roasted___> Lns: where do I specify local apps?
| |
11:43 | <Lns> roasted__, google 'ltsp localapps' ;)
| |
11:43 | <alkisg> Lns: hard to use with 1 gig ram? what do you mean?
| |
11:43 | <Lns> lots of info online
| |
11:43 | alkisg, well not the memory part but even the atom CPU couldn't handle firefox very well
| |
11:44 | <roasted___> Lns: with a standard array of apps: audacity, gimp, exaile, firefox, inkscape, libre office, is there anything that would make sense to use as local apps and others that make sense to use as server apps?
| |
11:44 | <alkisg> Youtube, pet society etc work fine on my daughter's atom
| |
11:44 | roasted___: with 128 mb ram you don't want any localapps at all :)
| |
11:45 | <Lns> roasted__, any app that needs 3D/heavy graphics support would be ideal in a localapp. but ^^^ too. ;)
| |
11:45 | <alkisg> If you have *more* than 256, you can start thinking about localapps
| |
11:45 | <roasted___> Lns: I certainly hope you are incorrect about that, as we are looking to get thin clients with atom procs and 1gb ram. This CANNOT Fail on us...
| |
11:45 | <Lns> roasted__, test test test is all i can advise!
| |
11:45 | <roasted___> alkisg: I was looking more along the lines of 512mb to 1gb of RAM. That being said, are there any apps that would make sense to run locally?
| |
11:45 | <alkisg> Sure, all of them :D
| |
11:45 | <roasted___> Lns: I am. But we almost dont have a choice but to go the thin client route. Everything else we tried sucks. LTSP is the only one that offers true hope...
| |
11:45 | * alkisg uses fat clients for clients with >= 512 RAM | |
11:46 | <Lns> atom CPUs are great but still not as good as core 2 duos ;)
| |
11:46 | <roasted___> school distrit budgets are crunching like no other.
| |
11:46 | <nesusvet> Guys, what i have to change within deb packages in order to disable X11 dependencies within ltsp-standalone-server?
| |
11:46 | <roasted___> alkisg: do you really think all apps would work locally with 512 ram?
| |
11:46 | <alkisg> nesusvet: the server doesn't depend on X. Are you talking about the client?
| |
11:46 | <roasted___> alkisg: it makes me nervous doing that considering what Lns said regarding 1gb RAM with firefox........
| |
11:46 | <Lns> roasted__, no disregard what i said about the 1gb ram
| |
11:47 | it was my CPU that was the bottleneck.
| |
11:47 | <alkisg> roasted___: 512 with atom is the limit I would put on "fat client better than thin client"
| |
11:47 | Anything better than that surely behaves better than as a thin client
| |
11:48 | So while 512 with local firefox is stretching it a bit, it's better than the same machine as thin
| |
11:48 | <nesusvet> alkisg, sorry, yes, i mean the client.
| |
11:48 | <Lns> With ram prices, 1GB is really ideal over 512 just to leave room for expansion..you want to focus on your CPU at that point to see if it can handle the apps with the performance you're looking for.
| |
11:49 | JamesB192 (quassel@67-60-20-114.cpe.cableone.net) left #ltsp ("http://quassel-irc.org - Chat comfortably. Anywhere."). | |
11:49 | * Lns waits patiently for HTML5 to spread across and make flash a bit less necessary | |
11:49 | VectorX (~knight@unaffiliated/vectorx) joined #ltsp. | |
11:49 | cVsup (~cVsup@189.12.14.211) joined #ltsp. | |
11:50 | <roasted___> alkisg: I think Ineed to do a little reading on fat clients vs thin, because I'm having a hard time understanding how a fat client could run more efficiently than thin without feeding it a TON of processing power to the clients themselves.
| |
11:50 | <alkisg> nesusvet: I believe the easiest thing to do would be to build a chroot with X and then try to remove them. But do you *really* want to remove them? Or you can just leave them there and not use them?
| |
11:50 | roasted___: forget the fat. Think "thin vs local installations"
| |
11:51 | Which runs better, an atom as a thin client, or an atom with a locally installed linux?
| |
11:51 | <nesusvet> alkisg, I know the way how-to leave them there and not use, at the moment i want to remove :)
| |
11:51 | <alkisg> nesusvet: ok, I don't have any advice as the dependencies are complex
| |
11:52 | <nesusvet> alkisg, thanks.
| |
11:52 | got it
| |
11:52 | <alkisg> np
| |
11:52 | <roasted___> alkisg: so bottom line is, thin clients are for bare minimal use. Fat clients come with better performance when you have clients that have more onboard hardware.
| |
11:53 | <alkisg> roasted___: I'd put it as "thin clients are good when you don't want much UI"
| |
11:53 | Little scrolling, no youtube videos, no oo presentations
| |
11:53 | <roasted___> alkisg: I see. LTSP defaults entirely to thin from a base install, right?
| |
11:53 | <alkisg> No 3d apps etc. But if you want processing power, you got lots there
| |
11:54 | Yes, unless you use the --fat-client option.
| |
11:54 | https://help.ubuntu.com/community/UbuntuLTSP/FatClients
| |
11:54 | <Lns> roasted__, case-in-point: Thin clients are good for general purpose (think a library research station). For full-fledged computer labs, fat-clients/local installs are preferred.
| |
11:54 | I found this out the hard way ;)
| |
11:55 | <roasted___> Lns: I understand. I'm just trying to think a little bit here... we're using existing hardware to get LTSP off the ground. These boxes are Pentium 4s with 512MB RAM. I'm trying to think about which would work best in that case, and I'm also trying to think about what would work best when we purchase actual thin client boxes.
| |
11:55 | <Lns> BUT the bigger issue you will face in a school environment is staff resistance, govt. software requirements and ongoing support.
| |
11:55 | msaul_ (~administr@CPE001cf0f564ef-CM0011ae00c1f8.cpe.net.cable.rogers.com) joined #ltsp. | |
11:55 | <roasted___> LIke I'm trying to think... if we get thin client boxes, do we get 2gb of RAM in the boxes or would 1gb be enough to future proof the client boxes for a while?
| |
11:55 | and processing wise, is an atom enough for a thin client? etc.
| |
11:55 | <Lns> roasted__, for your current systems LTSP is a great fit (using P4s as thin clients)
| |
11:56 | <msaul_> I got my chroot running ssh, and I recall someone mentioned to set root password
| |
11:56 | was wondering how I do this chroot...
| |
11:56 | <Lns> CPU wise I would stay away from Atoms if you're looking to do a lot of local processing. A simple (I.E. not FF/OOo) localapp won't be an issue, but again, test.
| |
11:57 | <roasted___> Lns: good deal. Now if I say we are upgrading clients this summer... what spec would you aim for for the clients?
| |
11:57 | for fat clients
| |
11:57 | <Lns> msaul_, sudo chroot /opt/ltsp/<arch>, passwd
| |
11:57 | <alkisg> roasted___: with that specific boxes, and if you really have the time to invest, localapps would be the best. I.e. some apps local, some remote.
| |
11:57 | <roasted___> 1gb ram and naything better than an atom?
| |
11:57 | <msaul_> OK, thanks...
| |
11:58 | <roasted___> alkisg: yeah I'm just concerned about spending money and it backfiring. I want to make the best educated guess I can when I need to get new lcients.
| |
11:58 | alkisg: tax payer money is a funny thing. You need to spend it wisely as if your life depends on it, because in some cases, it just might.
| |
11:58 | <alkisg> roasted___: for fat clients, boot the client from an external usb disk an test.
| |
11:58 | vvinet (~vince@2001:470:8829:1000:227:eff:fe25:ee64) joined #ltsp. | |
11:58 | <alkisg> If the os runs fine locally, it will also run fine as fat
| |
11:58 | There's no difference there.
| |
11:59 | <roasted___> alkisg: I see. I'm just talking about purchasing hardware I don't have. Like if I look at a chart of vendors selling ultra small PCs/client boxes, which one spec wise would be the best.
| |
11:59 | <alkisg> Go to a store and boot netbooks with similar specs off a usb stick :)
| |
11:59 | <_UsUrPeR__> gadi: about exporting for xrandr: the only resolution shown by xrandr for the connected monitor is 640x480, and adding new monitor resolutions doesn't seem to have an effect
| |
11:59 | <Lns> roasted__, think of fat-clients as centrally administrated machines that do all their own processing. So you'll want modern hardware specs that your OS recommends.
| |
11:59 | <alkisg> Do you think an atom with 1 Gb runs fine? My personal view is "so and so". My daughter doesn't like it because it can't run modern games. But for firefox+flash with 1-2 tabs it's ok
| |
11:59 | <roasted___> alkisg: well I have a netbook at home. that might be a good test.
| |
12:00 | Lns: I understand. It's just you get to a certain point where I don't want these clients to have THAT much hardware that their price shoots up.
| |
12:00 | <highvoltage> alkisg: hmm, I run more tabs than that on my atom with 1GB RAM (even with flash) and it's quite ok
| |
12:00 | <roasted___> If we begin to flirt with the line of "were saving next to nothing with this LTSP setup" then I have to wonder why we would get it.
| |
12:01 | <alkisg> highvoltage: try pet society :D The damn thing is toooo heavy :D
| |
12:01 | <highvoltage> not as nice as one more powerful machine, but still
| |
12:01 | <Gadi> _UsUrPeR__: yeah, I actually think I have seen that case before - with openchrome drivers - having to set PanelSize option - I think on Lucid?
| |
12:01 | <roasted___> yeah my netbook runs ubuntu 10.10 great on it, 1gb of ram. I've had gimp, open office, and flash going on firefox with upwards of 8 tabs just fine
| |
12:01 | <highvoltage> alkisg: aah, my flash is limited to videos, not games. I guess those are a different kettle of fish :)
| |
12:01 | <Lns> roasted__, central administration is your killer feature if you're looking to do fat-client-only LTSP. Which is totally awesome in itself.
| |
12:01 | <alkisg> highvoltage: I don't know how, but kids those days select the most cpu-demanding web sites... like they have a talent for it
| |
12:01 | VectorX (~knight@unaffiliated/vectorx) left irc: Ping timeout: 240 seconds | |
12:01 | <highvoltage> alkisg: but surely you can only play one flash game at a time? :)
| |
12:02 | <alkisg> Well, they tend to leave 1 tab open, switch to the other to chat, then open a video the friend told them to, then return to the game...
| |
12:02 | <roasted___> Lns: yeah, but we're looking for electricity savings as well. the more hardware that goes into a box, the more watts it'll eat. So I'm trying to watch all angles. It's not about making my job easy. It's about making smart economic decisions.
| |
12:02 | <alkisg> at that point they scream "my pc is too slow" :D
| |
12:02 | <Lns> roasted__, it's a delicate balance. ;)
| |
12:03 | <_UsUrPeR__> Gadi: I am prepared to "call it" unless you can think of anything else I can try out. It's confusing how force panel works, but I can't seem to get a proper image out of the monitor otherwise.
| |
12:03 | <roasted___> Lns: what thin/fat client boxes would you recommend? I saw an ASus EEE Box that looked nice...
| |
12:03 | <Lns> Although I applaud your use of open source software, in whatever form, with education. Teaching our future generations what open source is about will help them think differently than those who created empires upon locked-down systems.
| |
12:03 | <Gadi> _UsUrPeR__: my brain cells are coming up with PanelSize and setting X_VIRTUAL
| |
12:03 | but, the brain cells are old an weak
| |
12:04 | I think X_VIRUTAL was to keep the monitor from panning
| |
12:04 | <alkisg> X_VIRUSAL?
| |
12:04 | * Gadi kicks alksig | |
12:04 | <Gadi> :)
| |
12:04 | <alkisg> :P
| |
12:05 | <Lns> roasted__, for thin, use your existing P4s with >=512MB RAM.
| |
12:05 | <_UsUrPeR__> :)
| |
12:05 | * Lns hands Gadi a beer | |
12:05 | <roasted___> Lns: again, were getting new stuff this summer
| |
12:05 | drdata (~drdata@tw2.udallas.edu) left irc: Read error: Connection reset by peer | |
12:05 | <Gadi> wow an unsolicited beer
| |
12:05 | <roasted___> Lns: that gear will only work as our test bed UNTIL june.
| |
12:05 | <Lns> Gadi, =p how goes it?
| |
12:05 | <Gadi> better with beer
| |
12:06 | <Lns> roasted__, well I've had success with both disklessworkstations.com thin clients (1220 PXE is what i use) and also Gadi can tell you about his thin clients
| |
12:06 | haha
| |
12:06 | * Lns waits for the mom jokes to start rolling in | |
12:06 | <roasted___> Lns: these are all Atom based.
| |
12:07 | Are there any without Atoms so I can see other alternatives?
| |
12:07 | <alkisg> roasted___: for new equipment, I'd go for dual core
| |
12:07 | <_UsUrPeR__> I actually did try xvirtual. Not sure what it did. the screen was still scrolling and garbled
| |
12:07 | <alkisg> Don't have any specific advice
| |
12:07 | <Lns> roasted__, what are all atom-based?
| |
12:07 | <roasted___> alkisg: dual core atoms or just anything dual core?
| |
12:07 | Lns: disklessworkstations.com under ltsp
| |
12:08 | <Lns> 1220PXEs are via
| |
12:08 | <roasted___> I thought VIA was to be avoided
| |
12:09 | <Lns> why?
| |
12:09 | works flawlessly for me as a tc
| |
12:09 | always has
| |
12:09 | <roasted___> I heard some negative opinions about VIA in here the other day...
| |
12:09 | <alkisg> roasted___: let's say something over 1000 in that scale: http://www.cpubenchmark.net/common_cpus.html
| |
12:09 | <Lns> opinions are like ....well..
| |
12:09 | =p
| |
12:10 | <roasted___> yeah, but when you have quite a few siding the same way, its hard to ignore
| |
12:10 | <_UsUrPeR__> that was me about an hour ago ;) 1220 doesn't have DVI so you'll be fine
| |
12:10 | I am working with an HP thin client at the moment
| |
12:10 | <roasted___> no, this was a few days ago
| |
12:11 | <_UsUrPeR__> oh. huh. Well we will support what we sell.
| |
12:11 | oh hi, my name is Joel and I work for disklessworkstations.com
| |
12:11 | * _UsUrPeR__ tips his hat | |
12:11 | <roasted___> someone found me an idea on newegg and someone else said no, its got VIA. and then the one who linked me was like, aw yeah I missed that, stay away from VIA, etc...
| |
12:11 | nice to meet you :
| |
12:11 | :)
| |
12:12 | I'm trying to see if we can find some ultra low powered small form factor dual core towers that arent too costly that could do the job too
| |
12:13 | <Lns> so, you want...everything. ;)
| |
12:13 | <roasted___> doesnt everybody?
| |
12:13 | I'm just trying to look at all scenarios.
| |
12:13 | I feel confident I could get a few hundred EEE Box 1007's (1.66 atom 1gb ram) and do just fine with a few servers.
| |
12:14 | but if I can maximize it while I got the chance, why not?
| |
12:14 | <alkisg> roasted___: for fat clients you don't need CPU on the server
| |
12:14 | So fast network + fast disk could serve lots of fat clients
| |
12:15 | <roasted___> yeah. I'm just trying to weigh some of our options here.
| |
12:16 | I think I'm going to set up a test bed right now with a few P4 boxes with 512 ram
| |
12:16 | and compare them, fat vs thin
| |
12:16 | off my dual core laptop as the server and see wuz up
| |
12:16 | I just have to add the fat client tag when building the client image, right?
| |
12:17 | <alkisg> P4 at how many MHz/GHz?
| |
12:17 | --fat-client, right
| |
12:17 | <roasted___> uh, I'm not positive actually. one second.
| |
12:17 | vvinet (~vince@2001:470:8829:1000:227:eff:fe25:ee64) left irc: Remote host closed the connection | |
12:17 | <roasted___> so I would do sudo ltsp-build-client --arch i386 --fat-client
| |
12:17 | <alkisg> Yup
| |
12:18 | <roasted___> and if I want to bounce back to thin, just --thin-client ??
| |
12:18 | <_UsUrPeR__> gadi: is this the option you are describing? X_OPTION_01 = "\"Virtual\" \"1920 1200\""
| |
12:18 | <roasted___> alkisg: I think these are 2.4's
| |
12:18 | <alkisg> roasted___: in lts.conf, you specify LTSP_FATCLIENT=False
| |
12:18 | <roasted___> alkisg: though some older compaq boxes we have are P4s at 1.8. they'd be the slowest we have.
| |
12:19 | <alkisg> No need to build 2 chroots, one can serve both thin + fat
| |
12:19 | <roasted___> alkisg: no kidding?
| |
12:19 | <alkisg> Nope
| |
12:19 | <roasted___> how does the server know which image to load on the client?
| |
12:19 | <Gadi> _UsUrPeR__: no, just: X_VIRUTAL = "1280 1024"
| |
12:19 | damn
| |
12:19 | X_VIRTUAL
| |
12:19 | (if only I could type)
| |
12:19 | <alkisg> roasted___: the fat image is always used. It doesn't need more ram than the thin image, until the login screen (ldm)
| |
12:20 | And then LTSP_FATCLIENT tells the client to login locally or on the server
| |
12:20 | <roasted___> alkisg: the fat image is only always used when I initiate the --fat-client command, right? I wouldn't be using it now I wouldnt think...
| |
12:21 | <_UsUrPeR__> lol
| |
12:21 | <alkisg> roasted___: yes
| |
12:22 | <roasted___> I guess I just dont want to initiate the fat setup if I'm going to have some clients that wont be able to handle it
| |
12:22 | <alkisg> roasted___: any client that can boot from a thin image, can also boot from a fat image
| |
12:22 | Actually the fat image requires 2-3 MB *less* RAM
| |
12:22 | (don't ask my why)
| |
12:23 | <roasted___> The difference (correct me if I'm wrong) lies in HOW the processing of applications takes place, based on the image. If it's thin, it'll rely on the server. If it's fat, it'll rely on local resources.
| |
12:23 | <_UsUrPeR__> Gadi: X_VIRTUAL = "1024 768" same results. Screen garbled, looks like a bad refresh rate
| |
12:24 | strangely enough, the force panel option works, but only up to 1400x1050
| |
12:24 | which, according to openchrome docs, is the highest resolution ever ;)
| |
12:24 | <alkisg> roasted___: No. A fat image can be used for both thin and fat clients. Thin clients booted from a fat image will do the processing on the server.
| |
12:25 | cliebow (~cliebow@WatchGuard.ellsworth-hs.ellsworth.k12.me.us) left irc: Ping timeout: 240 seconds | |
12:25 | <alkisg> A fat image can do whatever a thin image can do (i.e. serve thin clients). In addition, it can also serve fat clients.
| |
12:25 | <Gadi> _UsUrPeR__: that is very high - for a panel
| |
12:25 | ;)
| |
12:25 | <roasted___> alkisg: so a fat image is just more versatile to use.
| |
12:25 | fat - fat/thin. thin - only thin
| |
12:25 | <alkisg> roasted___: yes. It's also bigger, which makes "ltsp-update-image" slower. And that's about all its drawback.
| |
12:26 | <roasted___> is the update-image command the same for fat and thin?
| |
12:26 | <alkisg> Yes
| |
12:26 | <_UsUrPeR__> Gadi: haha
| |
12:27 | <roasted___> I guess I'm just having trouble understanding how the setup knows the difference. If I use a fat client image, that means thin and fat clients can boot to the LTSP server. Okay, great... but I'm just tyring to grasp how the actual clients know WHERE to do their processing. AFter all, they're getting the same image. How would the client know to do local processing vs server processing?
| |
12:28 | <alkisg> roasted___: I told you above, with the LTSP_FATCLIENT variable from ltsconf
| |
12:28 | <Lns> roasted__, you need to read up on the wikis/docs/etc about all of this, it takes time to learn the ins and outs of it all
| |
12:28 | knowledge is power!
| |
12:28 | <roasted___> the thing is, I have read it.
| |
12:28 | <alkisg> It's a bit more complicated actually. LTSP_FATCLIENT defaults to true for fat chroots, and to false for thin chroots
| |
12:29 | <roasted___> I've been over this hundred page pdf a dozen times
| |
12:29 | <alkisg> So when a client boots from a fat chroot, it prefers to be fat
| |
12:29 | But there's another variable, FAT_RAM_THRESHOLD
| |
12:29 | With a default value of 500, if I remember well
| |
12:29 | <roasted___> so now that I ran --fat-client on my laptop and its currently building a fat image, do I need to change the FATCLIENT varaible in lts.conf or is that only if I have a mix of fat/thin clients?
| |
12:29 | <alkisg> So any clients with <500 RAM will by default be booted as thin, while > 500 as fat, IF they are booted from a fat chroot
| |
12:30 | <roasted___> fat chroot - the fat image itself. right.
| |
12:30 | * Lns is now attempting to install nvidia drivers on newly upgraded lenny-->squeeze box. will be back soon (hopefully) ;) | |
12:30 | <alkisg> So. You run ltsp-build-client --fat-client
| |
12:30 | nesusvet (~nesusvet@195.218.191.242) left irc: Ping timeout: 255 seconds | |
12:30 | <roasted___> yes
| |
12:30 | <alkisg> At that point any clients with <500 ram will boot as thin, and with >500 ram as fat
| |
12:30 | <roasted___> (plus --arch i386)
| |
12:30 | Lns (~Lns@pdpc/supporter/professional/lns) left irc: Quit: Leaving | |
12:30 | <alkisg> Now suppose you want a specific client with 700 RAM to be booted as thin
| |
12:30 | You'd put LTSP_FATCLIENT=False in lts.conf for that client
| |
12:31 | <roasted___> alkisg: so I only add the variable in lts.conf if I want to change the default behavior? (the 500 meg portion)
| |
12:31 | <alkisg> Right
| |
12:31 | <roasted___> how does false know to go to 700?
| |
12:32 | <alkisg> [mac:address] section in lts.conf
| |
12:32 | [01:02:03:04:05:06]
| |
12:32 | LTSP_FATCLIENT=False
| |
12:32 | <roasted___> is there no way to set it to all?
| |
12:32 | <alkisg> [Default]
| |
12:32 | LTSP_FATCLIENT=False
| |
12:32 | Standard lts.conf methods
| |
12:33 | <roasted___> So if I want to be selective on which clients have the newfound 700mb threshold, I do it by MAC. If I want EVERYBODY to be treated equal, I set it in default.
| |
12:33 | <alkisg> Right
| |
12:33 | <roasted___> last question
| |
12:33 | How does it know to threshold at 700? Is it designed to be 500 first, and 700 if false?
| |
12:34 | <alkisg> No that was an example
| |
12:34 | The threshold is at 500
| |
12:34 | So *for example* a client with 700 will be fat by default
| |
12:34 | <roasted___> oh okay. I was just trying to figure out where the numbers came in at
| |
12:34 | OH
| |
12:34 | <alkisg> And if you want it to be thin, you specify LTSP_FATCLIENT=False
| |
12:34 | <roasted___> I thought you meant changing that threshold to 700
| |
12:35 | so anything below 700 is thin, anything above is fat
| |
12:35 | <alkisg> If you want to, you can do that too
| |
12:35 | [Default]
| |
12:35 | FAT_RAM_THRESHOLD=700
| |
12:35 | <roasted___> awesome
| |
12:35 | taking these notes down quick...
| |
12:35 | <msaul_> hi, I set root passwd in chroot, but I think I need to perform something to allow me access to other users I created (i.e. I get "Permission Denied")..
| |
12:35 | <alkisg> roasted___: or just use `man lts.conf` :)
| |
12:36 | <msaul_> like ssh testa@192.168.1.20 (when I issue from server)
| |
12:36 | <roasted___> alkisg: I hear ya. :P
| |
12:37 | alkisg: I just like to take notes, because there's a LOT I do here... I'd hate to bounce out and someone come in trying tofigure out what the hell I was doing.
| |
12:37 | alkisg: the 500 meg limit wasn't an example, was it? That's ACTUALLY what LTSP defaults as. right?
| |
12:37 | <alkisg> Right
| |
12:38 | roasted___: erm, nope
| |
12:38 | As I said "if I remember well"
| |
12:38 | I didn't - it's actually 300
| |
12:38 | ltsp-trunk/client/ltsp_config: FAT_RAM_THRESHOLD=${FAT_RAM_THRESHOLD:-300}
| |
12:38 | <roasted___> so it is 300?
| |
12:38 | <alkisg> Yes
| |
12:41 | <msaul_> Hi alkisg, was wondering how to setup in chroot access to account in thin client in ssh (ssh is running in chroot and set root password)..
| |
12:42 | I created another account "testa" and get permission denied
| |
12:42 | <alkisg> msaul_: you already asked in the channel :) I'll let others answer that one while I get a drink... ;)
| |
12:42 | <msaul_> Fair enough :) sorry...
| |
12:46 | <roasted___> I understand RAM is the most important thing with an LTSP setup. For fat clients, is the same true? Or would processing power for fat clients suddenly be more important, given the need to balance things out??
| |
12:48 | <alkisg> Thin clients don't need much ram. 64 is the bare minimum. Fat clients have the same specs as local installations. So, read the ubuntu page (or any distro page) for the recommented specs
| |
12:49 | For ubuntu lucid I think the recommentation was 1 Gb RA
| |
12:49 | <roasted___> ahhh
| |
12:49 | you said that
| |
12:49 | what, really?
| |
12:50 | looks like 512
| |
12:50 | 1ghz proc
| |
12:50 | bobby_C (~bobby@85-124-22-227.teleworker.xdsl-line.inode.at) joined #ltsp. | |
12:51 | <alkisg> https://help.ubuntu.com/community/Installation/SystemRequirements
| |
12:51 | 1 GHz x86 processor 1GB of system memory (RAM)
| |
12:51 | <roasted___> seems as if wikipedia is lying then
| |
12:51 | <alkisg> That's the recommented. Below, as minimum, it says 512
| |
12:51 | <roasted___> ahh
| |
12:52 | <alkisg> So anyway all this falls outside the ltsp scope - best asked in #ubuntu
| |
12:52 | <roasted___> naw, that's all I was curious about :P
| |
12:52 | <alkisg> msaul_: you can't ssh as user@ltsp14 because there's no such user in the chroot
| |
12:53 | msaul_: people that install ssh in the chroot usually ssh as root@ltsp14
| |
12:53 | To troubleshot/shutdown/reboot the client, nothing much
| |
12:54 | <msaul_> Ok, it was instructions on scp ing keys to client for italc
| |
12:54 | They mentioned scp - r <keypath> user@IPADDRESS:/tmp
| |
12:56 | I was able to ssh into root via chroot though - I see your point...
| |
12:57 | <alkisg> msaul_: there's no point to copying keys with ssh
| |
12:57 | They'll get lost on reboot
| |
12:57 | Put them to chroot. You don't need to unlock the root account for that.
| |
12:57 | I.e. sudo cp key /opt/l/tsp/i386/path/to/key
| |
12:59 | <msaul_> OK, that's good advice - I'll do that to see what happens - thanks
| |
12:59 | Do you know of any one that was setup and used italc? Just curious...
| |
13:01 | vagrantc (~vagrant@c-24-21-191-93.hsd1.or.comcast.net) joined #ltsp. | |
13:01 | komunista (~slavko@adsl-195-098-015-248.dynamic.nextra.sk) left irc: Quit: Leaving. | |
13:02 | komunista (~slavko@adsl-195-098-015-248.dynamic.nextra.sk) joined #ltsp. | |
13:02 | Nick change: evil_root -> zz_evil_root | |
13:11 | <alkisg> msaul_: I used to use italc, until it proved to crash too often to be usable in classrooms
| |
13:20 | cVsup (~cVsup@189.12.14.211) left irc: Ping timeout: 252 seconds | |
13:21 | andygraybeal (~andy.gray@obsidian.casanueva.com) left irc: Remote host closed the connection | |
13:22 | drdata (~drdata@tw2.udallas.edu) joined #ltsp. | |
13:23 | <_UsUrPeR__> hey all
| |
13:24 | quick question: I am missing notes, and I swear this was discussed in here previously
| |
13:24 | There is a command that can be used to set BIOS times on clients based on an NTP server.
| |
13:24 | what is that command?
| |
13:24 | cVsup (~cVsup@189.12.14.211) joined #ltsp. | |
13:25 | <roasted___> do I have to update-image when I make changes to lts.conf?
| |
13:25 | <_UsUrPeR__> roasted__: no you do not
| |
13:25 | <roasted___> what about rebooting after build-image?
| |
13:25 | <_UsUrPeR__> you only need to update the image when you make a change to the /opt/ltsp/i386 directory
| |
13:25 | <roasted___> I thought so
| |
13:25 | wasnt positive though
| |
13:26 | <_UsUrPeR__> if you have rebuilt an image, you will need to reboot the clients so they may acquire the new image upon reboot
| |
13:26 | <roasted___> but not the server?
| |
13:26 | wait
| |
13:26 | I can rebuild while clients are logged in??
| |
13:26 | <alkisg> _UsUrPeR__: if you set the ntp server in lts.conf, then the cmos time is automatically updated... are you looking for a manual way to do it?
| |
13:27 | roasted___: yes
| |
13:27 | <roasted___> wow
| |
13:27 | that just seems strange :P
| |
13:27 | alkisg: how long have you worked with ltsp?
| |
13:27 | <alkisg> 3 years with linux/ltsp
| |
13:27 | <roasted___> have you done actual deployments for businesses/districts?
| |
13:27 | or do you just do backend work to make it happen?
| |
13:28 | <_UsUrPeR__> alkisg: oh, huh. What's the ntp server nomenclature? NTP_SERVER=<ipaddress>?
| |
13:28 | <alkisg> roasted___: deployed a few labs, but mostly backend work. About 200 schools so far switched to ltsp in greece: http://goo.gl/maps/nOoQ
| |
13:29 | <_UsUrPeR__> alkisg: you are actually living in greece right?
| |
13:29 | <alkisg> _UsUrPeR__: TIMESERVER
| |
13:29 | Yup
| |
13:29 | <_UsUrPeR__> so that makes it... what, 10pm for you right now?
| |
13:29 | <roasted___> nice. alkisg. I thought we were the only district in the area considering ltsp, but theres a neighboring district who's doing the same thing. in fact, we're meeting up next week to exchange friendly info.
| |
13:29 | <_UsUrPeR__> you are past GMT right?
| |
13:29 | roasted__: where are you located?
| |
13:29 | <roasted___> oddly enough I interviewed with that district, so it'll be weird to meet with them again
| |
13:29 | <alkisg> GMT+2
| |
13:29 | <roasted___> pennsylvania, USA
| |
13:29 | <_UsUrPeR__> ok, cool :)
| |
13:29 | <roasted___> the education system is hurting badly now
| |
13:30 | budget cuts are going like crazy
| |
13:30 | that's why this LTSP thing is like a godsend...
| |
13:30 | <_UsUrPeR__> yeah, saving bucks on clients is a good idea
| |
13:30 | I am surprised more schools don't use it
| |
13:30 | <alkisg> And administration, once you get it going the first time...
| |
13:30 | <roasted___> electricity savings for light power consuming clients AND gear that lasts longer?
| |
13:30 | AND the gear is cheaper?
| |
13:30 | I'm a huge linux fanboy... I've used it for 8 or 9 years now. But I've always been the oddball in the tech department here.
| |
13:31 | <_UsUrPeR__> alkisg: so TIMESERVER="ntp.ubuntu.com" is a good entry then? Does the client need to have access to DNS, or is that handled via server?
| |
13:31 | <alkisg> _UsUrPeR__: the client needs to have access to DNS and to the internet (nat on the server, or single nic setup, etc)
| |
13:31 | <roasted___> However, over the last year we've ported to a few linux based alternatives. So it began to win over the department. However, to the end user, we still have windows software we rely on. I'm doing a presentation next week to the board to prove to them we MUST act or we'll be in Windows-vendor-lock-in forever.
| |
13:31 | <alkisg> Otherwise you can install an ntp server on the server
| |
13:31 | <_UsUrPeR__> alkisg: gotcha
| |
13:32 | roasted__: it's unfortunate, but school districts in the US are typically gunshy when it comes to linux installations. I believe they think it's insecure, and the professionals aren't easlily found to help out.
| |
13:32 | <roasted___> I agree.
| |
13:32 | <_UsUrPeR__> you would have first-hand knowledge of these circumstances though :)
| |
13:32 | <roasted___> my boss's fear is we don't have the money to upgrade to windows 7
| |
13:32 | we just don't
| |
13:33 | but XP won't last forever
| |
13:33 | what do you do?
| |
13:33 | we get some CRAZY discounts with microsoft being we're an educational institution. But at 2,000 computers, it doesn't matter if a license is 10 bucks. that's a chunk of chnage just fo rupgrading your OS
| |
13:34 | <_UsUrPeR__> roasted__: I work for disklessworkstations.com . I do a lot of support and "special implementation" support.
| |
13:34 | <roasted___> We ported to open office last year and it was a hit. Couple that with firefox, audacity, gimp, and several other cross-platform apps and you begin to wonder, why do we even have windows? Sure we have our windows photoshop lab, but a LOT of our 25 labs are basic usage where we don't need windows specific software.
| |
13:34 | <_UsUrPeR__> also school installations ;)
| |
13:35 | jkinn2 (~jason@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) joined #ltsp. | |
13:35 | <roasted___> sweet deal man
| |
13:35 | I've been so happy with the support and answers I've gotten from everyone in this chat. It really make sme think we can pull this off without a hitch.
| |
13:35 | <_UsUrPeR__> I'd wager you'll be alright.
| |
13:36 | <roasted___> _UsUrPeR__: have you used ncomputing?
| |
13:36 | <_UsUrPeR__> roasted__: does not ring a bell, no.
| |
13:36 | <roasted___> ncomputing is a windows based thin client type setup...
| |
13:36 | proprietary hardware, windows server, etc.
| |
13:36 | we have it in our one lab.
| |
13:36 | Windows Server 2008, NComputing thin client boxes, 4 gigabit network cards, two six core 3.3ghz processors, and 32gb of RAM.
| |
13:36 | Guess how many times we have to reboot it in a week's time for it to function. :)
| |
13:39 | ogra (~ogra@ubuntu/member/ogra) left irc: Read error: Operation timed out | |
13:41 | <roasted___> Let's say I were to walk into an existing LTSP install. I have access on the server. Is there any simple way to determine if the server admin had set up fat clients on it?
| |
13:41 | <_UsUrPeR__> I am afraid to show my bias by answering that question
| |
13:41 | <roasted___> _UsUrPeR__: go ahead. Tell me what youre thinking :)
| |
13:42 | <_UsUrPeR__> roasted__: I would say at least once a day, but I am Windows Server dubious.
| |
13:42 | <roasted___> _UsUrPeR__: bingo
| |
13:42 | minimal of once a day, sometimes twice.
| |
13:42 | ogra (~ogra@p5098ed03.dip0.t-ipconnect.de) joined #ltsp. | |
13:42 | Nick change: ogra -> Guest95250 | |
13:42 | <roasted___> granted, that wasnt MY idea to put that in there. I was heavily skeptical of it. But that's why I'm doing so much testing...
| |
13:43 | <_UsUrPeR__> roasted__: about fatclients: you should not be able to tell, aside from the time it takes to start some programs
| |
13:43 | <matrix3000> how do I disable the dhcp running on ltsp
| |
13:43 | <roasted___> when you update the image, why would it take hdd space?
| |
13:43 | I'm on 200mb remaining
| |
13:43 | on a friday at almost 5 pm
| |
13:43 | stab me
| |
13:44 | <vagrantc> roasted___: clients still using the old image will continue to use it ... so even after the file is deleted, the inode is still in use
| |
13:44 | <roasted___> no clients are booted though
| |
13:44 | <vagrantc> so if the clients don't get rebooted, you end up with phantom disk images
| |
13:45 | any running nbd processes?
| |
13:45 | <alkisg> roasted___: maybe clean the apt cache? sudo apt-get clean
| |
13:45 | this should get you some free space
| |
13:45 | <roasted___> none that I see
| |
13:45 | its taking up hdd space
| |
13:45 | <vagrantc> i *think* it doesn't delete the old image until the new image is sucessfully created
| |
13:45 | <roasted___> Im at 70% and 250mb remaining
| |
13:45 | I removed some .debs I downloaded
| |
13:46 | 173 remaining
| |
13:46 | 78%
| |
13:46 | <alkisg> About fat clients, you can see it from /opt/ltsp/image/i386.img, which will be > 1 Gb, and from /opt/ltsp/i386/etc/ltsp_fat_chroot
| |
13:46 | <roasted___> I'm screwed arent I?
| |
13:46 | <vagrantc> roasted___: ltsp-chroot apt-get clean
| |
13:46 | <_UsUrPeR__> matrix3000: stop the process /etc/init.d/dhcp3-server
| |
13:46 | with the following command: /etc/init.d/dhcp3-server stop
| |
13:46 | <vagrantc> roasted___: pause ltsp-update-image with ctrl-z
| |
13:46 | <roasted___> 17 mb
| |
13:46 | paused at 2.7 remaining
| |
13:46 | lol
| |
13:46 | <alkisg> Haha
| |
13:47 | <vagrantc> then figure out how to free up some space ... and resume with "fg"
| |
13:47 | <roasted___> I dont think I can
| |
13:47 | this is a fresh install
| |
13:47 | its a 15gb partition...
| |
13:47 | I thought thatd be enough
| |
13:47 | <Gadi> roasted__: sudo apt-get clean is usually a good way to free some space
| |
13:47 | <roasted___> (this is just for testing)
| |
13:47 | <Gadi> :)
| |
13:47 | <vagrantc> or let it fail, delete the existing image ... you can always recreate it with ltsp-update-image
| |
13:47 | <alkisg> heh, my fat chroot is 15 Gb
| |
13:47 | <roasted___> Gadi: that command does nothing
| |
13:47 | * _UsUrPeR__ jokes -- rm -rv /var | |
13:47 | <Gadi> check your free space
| |
13:47 | <roasted___> vagrantc: /opt/ltsp/i386?
| |
13:47 | <vagrantc> roasted___: ltsp-chroot apt-get clean
| |
13:47 | ??
| |
13:47 | <_UsUrPeR__> DO NOT ENTER THAT
| |
13:48 | <roasted___> I know
| |
13:48 | :P
| |
13:48 | <vagrantc> roasted___: delete /opt/ltsp/images/*
| |
13:48 | <_UsUrPeR__> what kind of space does your serve rhave? df -h ?
| |
13:48 | <vagrantc> roasted___: yes, clean the apt cache in your chroot
| |
13:48 | <roasted___> 160gb hd, but I partitioned edubuntu to have 15gb
| |
13:48 | <alkisg> roasted___: like vagrantc said, sudo chroot /opt/ltsp/i386 apt-get clean
| |
13:48 | <vagrantc> roasted___: ltsp-chroot apt-get clean
| |
13:48 | <roasted___> k
| |
13:48 | got 600 meg now
| |
13:49 | anything else?
| |
13:49 | <alkisg> I don't think ltsp-chroot is available on ubuntu yet
| |
13:49 | <vagrantc> what?!
| |
13:49 | <alkisg> (not in the path)
| |
13:49 | <vagrantc> it's been there for years
| |
13:49 | <alkisg> Yup. But not in the system path in ubuntu :)
| |
13:49 | <roasted___> if I run update-image will I eat up more than 600 meg?
| |
13:49 | <vagrantc> if ubuntu can't keep up with changes in debian ... somethings going kind of funny :)
| |
13:49 | <alkisg> vagrantc: https://bugs.launchpad.net/ubuntu/+source/ltsp/+bug/589833
| |
13:50 | <vagrantc> we agreed to start using it in like 2007
| |
13:50 | <alkisg> I filed that last year, I think it'll be there on natty
| |
13:50 | <_UsUrPeR__> alkisg: for the lts.conf TIMEZONE =, what is the nomenclature? would it be gmt-5, or just -5?
| |
13:50 | <roasted___> yeah this will use more than 600
| |
13:50 | * vagrantc wonders why alkisg doesn't have upload rights to ubuntu | |
13:50 | <alkisg> _UsUrPeR__: ls /usr/share/zoneinfo/
| |
13:50 | <roasted___> 2% in and 120mb is gone
| |
13:50 | <_UsUrPeR__> thx
| |
13:51 | <alkisg> vagrantc: no idea, but I'm also not sure that I'd want that for now, with my phd. After 2 years I'll look into it, but I may switch to debian by that time with all the changes to come in ubuntu :D
| |
13:52 | <roasted___> am I screwed?
| |
13:52 | <alkisg> _UsUrPeR__: e.g. I'm using TIMEZONE=Europe/Athens (so you can have slashes there)
| |
13:52 | <_UsUrPeR__> oh OH
| |
13:52 | ok
| |
13:52 | so America/Detroit then
| |
13:52 | gotcha
| |
13:52 | Gadi (romm@ool-18bbe47a.static.optonline.net) left #ltsp. | |
13:53 | <roasted___> screw this. it's beer time.
| |
13:53 | <matrix3000> _UsUrPer_: so like normal...i just wasn't sure, I knew the config file was stored in a diff directory and didn't know if there was a specific config
| |
13:53 | <alkisg> roasted___: you can bind-mount from another partition
| |
13:53 | <roasted___> alkisg: already powered it off :P
| |
13:53 | <_UsUrPeR__> matrix3000: yeah, config file is stored in /etc/ltsp/
| |
13:53 | <alkisg> Heh, beer time indeed
| |
13:53 | <roasted___> got a 2nd laptop here on irc
| |
13:54 | <_UsUrPeR__> matrix3000, but like normal, yeah
| |
13:54 | <vagrantc> alkisg: oh, well, lucky for debian, then :)
| |
13:57 | heh. the main thing holding back an ltsp 5.2.6 upload to debian is editing the changelog ...
| |
13:58 | roasted___ (4bc3a2d8@gateway/web/freenode/ip.75.195.162.216) left irc: Ping timeout: 252 seconds | |
14:01 | alkisg (~alkisg@ubuntu/member/alkisg) left irc: Quit: Leaving. | |
14:02 | _UsUrPeR__ (~jsass@fw.acurrus.com) left irc: Ping timeout: 255 seconds | |
14:05 | <matrix3000> _UsUrPeR_: now can I just do a sudo apt-get remove dhcp3? and just remove it from the server since we dont need it?
| |
14:06 | or should i just remove it from rd.d
| |
14:06 | rc.d*
| |
14:14 | <msaul_> alkisg: Are there any alternative to italc that you would recommend?
| |
14:49 | Kicer86 (~Kicer86@host-5db0eeee.sileman.net.pl) left irc: Quit: KVIrc Insomnia 4.0.2, revision: 4740, sources date: 20100627, built on: 2010-08-08 18:29:00 UTC http://www.kvirc.net/ | |
14:54 | matrix3000 (matrix@rrcs-70-61-255-226.central.biz.rr.com) left #ltsp. | |
14:56 | bobby_C (~bobby@85-124-22-227.teleworker.xdsl-line.inode.at) left irc: Quit: Goin' down hard | |
15:02 | drdata (~drdata@tw2.udallas.edu) left irc: Ping timeout: 250 seconds | |
15:02 | jkinn2 (~jason@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) left irc: Ping timeout: 252 seconds | |
15:06 | bobby_C (~bobby@85-124-22-227.teleworker.xdsl-line.inode.at) joined #ltsp. | |
15:16 | Trixboxer (~Trixboxer@office.supportdepartment.net) left irc: Quit: "Achievement is not the end, its the beginning of new journey !!!" | |
15:18 | msaul_ (~administr@CPE001cf0f564ef-CM0011ae00c1f8.cpe.net.cable.rogers.com) left irc: Quit: Ex-Chat | |
15:22 | Lns (~Lns@pdpc/supporter/professional/lns) joined #ltsp. | |
15:36 | Selveste1 (~Selveste1@88.83.86.20.static.dong.customer.smilecontent.dk) left irc: Quit: Server does not respond. | |
15:37 | Selveste1 (~Selveste1@88.83.86.20.static.dong.customer.smilecontent.dk) joined #ltsp. | |
15:38 | Guest95250 (~ogra@p5098ed03.dip0.t-ipconnect.de) left irc: Changing host | |
15:38 | Guest95250 (~ogra@ubuntu/member/ogra) joined #ltsp. | |
15:38 | Nick change: Guest95250 -> ogra | |
15:54 | garymc (~chatzilla@host86-145-43-73.range86-145.btcentralplus.com) joined #ltsp. | |
15:55 | garymc (~chatzilla@host86-145-43-73.range86-145.btcentralplus.com) left irc: Client Quit | |
16:16 | alkisg (~alkisg@ubuntu/member/alkisg) joined #ltsp. | |
16:23 | roasted__ (~Jason@c-174-54-217-48.hsd1.pa.comcast.net) left irc: Read error: Connection reset by peer | |
16:41 | <vagrantc> wheee. finally got new ltsp uploaded to debian.
| |
16:41 | was a fairly minor change, but somehow took a whole week.
| |
16:43 | <alkisg> vagrant++
| |
16:46 | * vagrantc rolls off into the sunset | |
16:46 | vagrantc (~vagrant@c-24-21-191-93.hsd1.or.comcast.net) left irc: Quit: leaving | |
17:09 | komunista (~slavko@adsl-195-098-015-248.dynamic.nextra.sk) left irc: Quit: Leaving. | |
17:27 | Lns (~Lns@pdpc/supporter/professional/lns) left irc: Ping timeout: 240 seconds | |
17:30 | alkisg (~alkisg@ubuntu/member/alkisg) left irc: Quit: Leaving. | |
17:54 | bobby_C (~bobby@85-124-22-227.teleworker.xdsl-line.inode.at) left irc: Quit: Goin' down hard | |
18:08 | drdata (~drdata@pool-173-57-189-158.dllstx.fios.verizon.net) joined #ltsp. | |
18:50 | rjune (~rjune@oh-67-77-122-143.sta.embarqhsd.net) left irc: Read error: Operation timed out | |
19:03 | msignor (~msignor@24.115.67.19.res-cmts.sesp.ptd.net) left irc: Ping timeout: 248 seconds | |
21:19 | Lns (~Lns@76-191-128-251.dsl.dynamic.sonic.net) joined #ltsp. | |
21:19 | Lns (~Lns@76-191-128-251.dsl.dynamic.sonic.net) left irc: Changing host | |
21:19 | Lns (~Lns@pdpc/supporter/professional/lns) joined #ltsp. | |
21:48 | alkisg (~alkisg@ubuntu/member/alkisg) joined #ltsp. | |
21:56 | artista-frustrad (~artista_f@187.7.151.137) left irc: Ping timeout: 250 seconds | |
21:59 | alkisg (~alkisg@ubuntu/member/alkisg) left irc: Quit: Leaving. | |
22:01 | msignor (~msignor@24.115.67.19.res-cmts.sesp.ptd.net) joined #ltsp. | |
22:09 | artista-frustrad (~artista_f@187.7.151.137) joined #ltsp. | |
22:11 | <msignor> hello - anyone notice a bug in firefox where it says a SSL site is not trusted even though its from a valid CA? just curious..
| |
22:20 | cVsup (~cVsup@189.12.14.211) left irc: Ping timeout: 250 seconds | |
22:30 | MorningSon (~MorningSo@cpe-70-114-21-95.satx.res.rr.com) left irc: Quit: WeeChat 0.3.0 | |
22:47 | spectra (~spectra@debian/developer/spectra) left irc: Ping timeout: 248 seconds | |
22:55 | spectra (~spectra@debian/developer/spectra) joined #ltsp. | |
23:09 | msignor (~msignor@24.115.67.19.res-cmts.sesp.ptd.net) left irc: Ping timeout: 250 seconds | |
00:00 | --- Sat Mar 19 2011 | |