IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 15 May 2009 (all times are UTC)
| 00:20 | <chupacabra> waterboard em.
| |
| 00:29 | hanthana has quit IRC | |
| 00:37 | <alkisg> Wow Moonlight (= free clone of Microsoft Silverlight) is a lot more LTSP-friendly than Adobe Flash Player...
| |
| 01:10 | ajohnson has quit IRC | |
| 01:10 | xvart has joined #ltsp | |
| 01:23 | <xvart> need some help with local apps on debian lenny, there is no ltsp-localapps script installed
| |
| 01:25 | xvart has quit IRC | |
| 01:35 | vvinet_ has joined #ltsp | |
| 01:46 | rcmysoft has joined #ltsp | |
| 01:47 | yanu has quit IRC | |
| 01:47 | yanu has joined #ltsp | |
| 01:48 | <rcmysoft> hey, my ltsp server has two nic one internet the other lan (Thin clients). To my surprise the thin clients can access the internet without any prob. Now how can i force then to use a proxy?
| |
| 01:50 | <sep> how do you check that the thin clients can access the internet ?
| |
| 01:51 | rcmysoft, since if you are in grafical you are realy accessing the internet from the ltsp server, and just displaying the result of the program on the screen of the thin clients
| |
| 01:52 | so if you want to force everything thru a proxy you realy should force the ltsp server thru the proxy.
| |
| 01:54 | <rcmysoft> so i have to make the system proxy settings in the ltsp server?
| |
| 01:55 | <sep> indeed
| |
| 01:55 | <rcmysoft> oki. do you know wich file it it? but i can google, too.
| |
| 01:56 | is it =:0)
| |
| 01:56 | <sep> depends but i do it in /etc/environment
| |
| 01:57 | vvinet has quit IRC | |
| 01:57 | <rcmysoft> oki. Thanks a lot. So when i use a printer from the internet lan the client will be able to print, too. Beeing in the other lan if so it is very cool.
| |
| 01:59 | byr
| |
| 01:59 | bye
| |
| 02:11 | rcmysoft has left #ltsp | |
| 02:49 | try2free has joined #ltsp | |
| 02:51 | try2free has left #ltsp | |
| 03:06 | Scormen has joined #ltsp | |
| 03:07 | <Scormen> good morning!
| |
| 03:07 | <X-TaZ> 'morning
| |
| 03:08 | <Scormen> i'm having a problem with the time of the thin clients
| |
| 03:08 | <X-TaZ> Does your ltsp server 's time is fine ?
| |
| 03:08 | <Scormen> the thin client is always taking its hardware time, it seems i can't push it via lts.conf
| |
| 03:09 | <Ryan52> on the login screen or after?
| |
| 03:09 | <Scormen> on the server, the time is fine
| |
| 03:09 | <Ryan52> the timezone is wrong or the time itself?
| |
| 03:09 | nevermind, dumb question..
| |
| 03:10 | <Scormen> ryan - on the login screen it is wrong and when I have loged in and look via a ltsp-localapps xterm
| |
| 03:10 | <Ryan52> set TIMESERVER
| |
| 03:10 | <X-TaZ> in ltsp-cluster, there is a lts.conf var named "TIMESERVER" it could be usefull
| |
| 03:10 | <Ryan52> Scormen: TIMESERVER=0.debian.pool.ntp.org
| |
| 03:10 | or, if you have a local ntp server, set it to that.
| |
| 03:10 | <Scormen> in the file "/var/lib/tftpboot/ltsp/amd64/ltsp.conf I have:
| |
| 03:10 | TIMEZONE=Europe/Brussels
| |
| 03:11 | TIMESERVER=91.189.94.4
| |
| 03:16 | <Ryan52> that server doesn't work.
| |
| 03:16 | afaict.
| |
| 03:16 | unless it's yours.
| |
| 03:16 | <Scormen> no, its the IP of ntp.ubuntu.com
| |
| 03:17 | <Ryan52> well, it don't work..
| |
| 03:17 | <Scormen> hmm oke
| |
| 03:17 | <Ryan52> wiat.
| |
| 03:17 | nevermind
| |
| 03:17 | <Scormen> I just tried 0.debian.pool.ntp.org, but it doesn't work to
| |
| 03:17 | <Ryan52> ugh, stupid internet problems.
| |
| 03:17 | it works for me now.
| |
| 03:17 | <Scormen> hehe :)
| |
| 03:21 | <alkisg> Scormen: so did it work for the TCs?
| |
| 03:21 | <Scormen> no, still nothing
| |
| 03:22 | it keeps taking the bios time
| |
| 03:22 | <alkisg> Hmmm well sudo ntpdate 91.189.94.4 works fine on the server, let's see why it isn't working on the clients...
| |
| 03:23 | <Scormen> when i do that on the server, i get:
| |
| 03:23 | 15 May 10:22:14 ntpdate[7579]: the NTP socket is in use, exiting
| |
| 03:23 | <alkisg> Hmm? who uses it? netstat -tulp
| |
| 03:25 | <X-TaZ> I'm trying to use LDAP with LTSP. It works for login, but i have to create home dirs manualy. I know pam_mkhomedir can do it, but i can't find how to use it. Someone here knows how to ?
| |
| 03:26 | <Scormen> alkisg: server.local:ntp, server.home.lan:ntp, localhost:ntp, *:ntp, .... all possible localhosts are using it
| |
| 03:27 | <alkisg> No pid? Maybe it's used by inetd?
| |
| 03:27 | <Scormen> indeed, it is in use by /usr/sbin/ntpd -p /var/run/ntpd.pid -u 115:125 -g
| |
| 03:28 | <alkisg> Scormen: so maybe if your server runs ntpd, it can't be used as an ntp client? (don't know, just wondering...)
| |
| 03:29 | <Scormen> so it seems... ntpdate on the server is working now: 15 May 10:27:36 ntpdate[7620]: adjust time server 91.189.94.4 offset 0.004507 sec
| |
| 03:29 | <alkisg> "An NTP daemon needs to be running on each host that is to have its clock accuracy controlled by NTP. The same NTP daemon is also used to provide NTP service to other hosts."
| |
| 03:30 | <Scormen> i have ntp and ntpdate installed on the server and the chroot image, is that wrong?
| |
| 03:30 | <alkisg> Scormen: an do your clients have direct access to the "outside world"? Or should you use your server as the ntp server?
| |
| 03:30 | <Scormen> they have to use the server as ntp
| |
| 03:30 | hmmm wait
| |
| 03:30 | <alkisg> I mean, can you ping ntp.ubuntu.com from a local terminal in a client?
| |
| 03:31 | <Scormen> no I cant
| |
| 03:31 | i'm try to define the server ip in lts.conf
| |
| 03:33 | if i define the server IP in lts.conf in the variable "timeserver", it is still not working
| |
| 03:34 | <alkisg> Get a shell on the client to test manually
| |
| 03:34 | SCREEN_02=shell or something.
| |
| 03:34 | <X-TaZ> Well i found how to make pam_mkhomedir working. Sorry :)
| |
| 03:37 | <Scormen> does that SCREEN_02=shell has to go into lts.conf?
| |
| 03:37 | <X-TaZ> Scormen, sure
| |
| 03:37 | there is a SCREEN_07=ldm in that file too
| |
| 03:38 | <Scormen> there is no SCREEN_07=ldm in lts.conf
| |
| 03:38 | <alkisg> Scormen: it's only used for debugging
| |
| 03:38 | <X-TaZ> There is in my
| |
| 03:38 | mine
| |
| 03:38 | <Scormen> the only variables I have in lts.conf are "TIMEZONE" "TIMESERVER" and now "SCREEN_02"
| |
| 03:38 | <alkisg> SCREEN_02=shell will make ldm *not* starting, though,
| |
| 03:39 | so better put both SCREEN_02=shell and SCREEN_07=ldm
| |
| 03:39 | <Scormen> oke
| |
| 03:39 | <X-TaZ> I don't think he really have to ad a shell. In his xdm, he can run a console. It'll have the same effect
| |
| 03:39 | <alkisg> Scormen: you do have a [default] stanza on top, right?
| |
| 03:39 | <Scormen> yes, i have
| |
| 03:39 | <alkisg> X-TaZ: ltsp doesn't use xdm
| |
| 03:39 | <X-TaZ> gdm
| |
| 03:40 | sorry
| |
| 03:40 | <alkisg> Nor gdm. How is he going to run a console in ldm?
| |
| 03:40 | <X-TaZ> Don't know :x
| |
| 03:41 | I mean, my setup is SCREEN_07=ldm ... and it launched gnome
| |
| 03:41 | <alkisg> ldm is the ltsp display manager. After you login, you see the DE of the server
| |
| 03:41 | So gnome isn't running on the client, it runs on the server. But Scormen needs a *local* xterm.
| |
| 03:42 | <X-TaZ> Oh ok i didn't understand that point
| |
| 03:42 | Sorry :)
| |
| 03:42 | <Scormen> can't i just run a local one via "ltsp-localapps xterm"?
| |
| 03:42 | <alkisg> Scormen: sure :)
| |
| 03:42 | (didn't know you run 9.04)
| |
| 03:43 | <Scormen> yes i run 9.04 64bit
| |
| 03:43 | <alkisg> Scormen: I just tried TIMEZONE/TIMESERVER in my lts.conf, worked fine. :)
| |
| 03:43 | (thanks!)
| |
| 03:43 | <Scormen> damn, why not here ;)
| |
| 03:44 | I have "ntp" and "ntpdate" installed in the chroot of the image, that can't be a problem?
| |
| 03:44 | <alkisg> Scormen: your TCs don't have access in the outside world, so you either have to give them access (not recommented) or set up an ntpd server that works.... :)
| |
| 03:45 | Why ntp? Try removing it...
| |
| 03:45 | <Scormen> i set up ntpd on the server now, on the ltsp-server
| |
| 03:45 | but still the wrong time
| |
| 03:45 | <alkisg> Did you get a shell running?
| |
| 03:45 | <Scormen> yes
| |
| 03:45 | <alkisg> If so, try `ntpdate server-ip` from a client
| |
| 03:46 | <Scormen> ntpdate: command not found :/
| |
| 03:46 | <alkisg> Wow. I do have it, and I didn't install anything in the chroot.
| |
| 03:46 | Maybe ntp had conflicting packages and removed it?
| |
| 03:46 | <Scormen> i'll remove it from the chroot
| |
| 03:47 | <alkisg> Make sure you have ntpdate in the chroot.
| |
| 03:47 | <Scormen> ntp and ntpdate remove from the chroot?
| |
| 03:47 | so, only removing ntp, right?
| |
| 03:47 | <alkisg> Yes, and install ntpdate
| |
| 03:47 | <Scormen> oke
| |
| 03:47 | <alkisg> (in the chroot, right?)
| |
| 03:47 | <Scormen> yes
| |
| 03:48 | on the server: ntp and ntpdate where installed
| |
| 03:48 | in the chroot: removed ntp
| |
| 03:48 | in the chroot: ntpdate was already installed
| |
| 03:48 | <alkisg> If it was installed, then why did you say command not found?
| |
| 03:49 | <Scormen> thats what the client is says
| |
| 03:49 | <alkisg> Scormen, try: ls -lha /opt/ltsp/i386/usr/sbin/ntpdate
| |
| 03:49 | <Scormen> -rwxr-xr-x 1 root root 61K 2009-03-21 01:51 /opt/ltsp/amd64/usr/sbin/ntpdate
| |
| 03:49 | <alkisg> Hmmm? So why didn't the client see it? Wrong path?
| |
| 03:50 | echo $PATH in the local client terminal...
| |
| 03:50 | <Scormen> /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/game
| |
| 03:51 | no wait, that was on the server
| |
| 03:51 | on the client: /usr/local/bin:/usr/bin:/bin:/usr/games
| |
| 03:52 | ogra has quit IRC | |
| 03:52 | ogra_ has quit IRC | |
| 03:53 | ogra has joined #ltsp | |
| 03:54 | <alkisg> Scormen: ok, you need to copy environment, but before you do and out of curiosity, which method did you use for installation?
| |
| 03:55 | E.g. you didn't install with the alternate cd...
| |
| 03:55 | <Scormen> yes, installation with the alternate cd
| |
| 03:55 | but without the option "ltsp server"
| |
| 03:55 | just a clean installation
| |
| 03:55 | <alkisg> Ah, ok, that's why.
| |
| 03:56 | sudo cp /etc/environment /opt/ltsp/amd64/etc
| |
| 03:56 | <Scormen> with the option mode "ltsp server" it fails with me always on 50%
| |
| 03:56 | <alkisg> and then sudo ltsp-update-image
| |
| 03:56 | It needs a *lot* of time there, did you leave it for 20 minutes or so?
| |
| 03:56 | cyberorg has quit IRC | |
| 03:56 | <Scormen> i copied that
| |
| 03:57 | ltsp-update-image is ready to
| |
| 04:00 | eum.... I don't have "/etc/environment" on the server
| |
| 04:01 | cyberorg has joined #ltsp | |
| 04:01 | <alkisg> Hmmm well let's not try to solve everything, here, copy mine: :)
| |
| 04:01 | alkisg@alkis:~/tmp/ltsp-upstream$ cat /opt/ltsp/i386/etc/environment
| |
| 04:01 | PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games"
| |
| 04:03 | <Scormen> done
| |
| 04:03 | <alkisg> ok, ltsp-update-image, reboot client, get a local shell.
| |
| 04:05 | <Scormen> ready
| |
| 04:06 | <alkisg> what does ntpdate your-server-ip say now?
| |
| 04:06 | <Scormen> guess what... "ntpdate: command not found"
| |
| 04:07 | <alkisg> try with: /usr/sbin/ntpdate your-server-ip
| |
| 04:07 | ogra_ has joined #ltsp | |
| 04:07 | <Scormen> persmission denied for a normal user, become root?
| |
| 04:08 | <alkisg> In a moment. For now, try: cat /etc/environment
| |
| 04:08 | From the local xterm, and make sure it does have the /usr/sbin path in it.
| |
| 04:08 | Does it?
| |
| 04:08 | <Scormen> PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games"
| |
| 04:09 | it has "/usr/sbin" in it
| |
| 04:09 | <alkisg> ogra ogra_, (good morning), why would an amd64 chroot not respect the PATH that we put in /etc/environment?
| |
| 04:09 | login.defs or something overrides that?
| |
| 04:10 | Scormen: also try: echo $PATH, what do you get?
| |
| 04:11 | <Scormen> echo $PATH = "/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games"
| |
| 04:11 | <alkisg> Emmm what?
| |
| 04:11 | <Scormen> srr, I get:
| |
| 04:11 | /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/game
| |
| 04:12 | <alkisg> /usr/sbin is there, so why doesn't it run ntpdate?
| |
| 04:12 | <Scormen> i have it
| |
| 04:12 | because i wasn't root?
| |
| 04:12 | <alkisg> No, it shouldn't say "command not found"
| |
| 04:12 | <Scormen> if i become root, than i can run "ntpdate server-ip"
| |
| 04:13 | if i become a normal user again, i get "bash: ntpdate: command not found"
| |
| 04:16 | <alkisg> Did you run "echo $PATH" as root or as a normal user?
| |
| 04:16 | mikkel has joined #ltsp | |
| 04:16 | <alkisg> Bah anyway if you can run it as root then that's enough
| |
| 04:16 | <Scormen> damn, as root, sorry
| |
| 04:16 | <alkisg> Maybe this perticular user has a user-defined path
| |
| 04:17 | <Scormen> as a local user, i get:
| |
| 04:17 | /usr/local/bin:/usr/bin:/bin:/usr/games
| |
| 04:17 | <alkisg> Scormen: anyway, leave it, the user path may be defined in .bashrc or somewhere, we're not interested in that.
| |
| 04:17 | <Scormen> so, no "/usr/sbin"
| |
| 04:17 | <alkisg> So login as root and we take it from there.
| |
| 04:17 | <Scormen> ok
| |
| 04:17 | <alkisg> As root: ntpdate your-server-ip
| |
| 04:18 | What does that tell you?
| |
| 04:18 | <Scormen> thats working, "... adjust time server ..."
| |
| 04:18 | <alkisg> OK, so it should be working from lts.conf now.
| |
| 04:18 | <Scormen> after that, the time is oke
| |
| 04:18 | great
| |
| 04:19 | <alkisg> try it...
| |
| 04:19 | <Scormen> restarting the client, you mean?
| |
| 04:21 | i restarted the client, the time is wrong again...
| |
| 04:24 | <alkisg> Do you have /opt/ltsp/amd64/usr/share/zoneinfo/Europe/Brussels?
| |
| 04:24 | and, upload your lts.conf to pastebot...
| |
| 04:26 | <Scormen> yes i have /opt/ltsp/amd64/usr/share/zoneinfo/Europe/Brussels
| |
| 04:28 | http://sial.org/pbot/36626
| |
| 04:29 | the "SCREEN_*" is still in it, from the start of this chat
| |
| 04:29 | <alkisg> OK. Try this:
| |
| 04:29 | Alt+Ctrl+F2 to switch to screen_02
| |
| 04:30 | Then: date
| |
| 04:30 | What time does it say?
| |
| 04:30 | Then: ntpdate your-server-ip
| |
| 04:30 | Then: date
| |
| 04:30 | What time does it say?
| |
| 04:30 | <Scormen> can't go to screen_02, just a blinking cursor left above
| |
| 04:31 | <alkisg> Erm... where did you put lts.conf ?
| |
| 04:31 | <Scormen> /var/lib/tftpboot/ltsp/amd64/lts.conf
| |
| 04:32 | <alkisg> OK... Again, run a local xterm, and in it: getltscfg -a
| |
| 04:32 | What does it say?
| |
| 04:33 | <Scormen> thats printing the lines from "lts.conf"
| |
| 04:33 | +
| |
| 04:33 | export TIMEZONE
| |
| 04:33 | export TIMESERVer
| |
| 04:33 | <alkisg> OK
| |
| 04:33 | <Scormen> export SCREEN_02
| |
| 04:33 | ...
| |
| 04:33 | <alkisg> How do you get a local root shell? You've unlocked the chroot root account?
| |
| 04:33 | <Scormen> indeed
| |
| 04:34 | <alkisg> OK, so another mystery, why SCREEN_02 doesn't work. Anyway. :)
| |
| 04:34 | Try the lines above (with the dates) with the local root account.
| |
| 04:35 | <Scormen> date: prints the wrong date
| |
| 04:35 | ntpdate server-ip: set time server ....
| |
| 04:35 | date: date and time oke
| |
| 04:36 | <alkisg> Please paste both of the hours here (to see if it's a zone mistake)
| |
| 04:36 | <Scormen> wrong date: Fri May 15 19:37:07 CEST 2009
| |
| 04:36 | right date: Fri May 15 11:35:07 CEST 2009
| |
| 04:37 | <alkisg> Wow, thats too wrong :P
| |
| 04:37 | * alkisg scratches head... | |
| 04:37 | <Scormen> hehe, just a few hours
| |
| 04:37 | <alkisg> (it's supposed to keep the correct date in the next reboot, yet it seems that it lost it)
| |
| 04:38 | <Scormen> indeed, it gets lost after the reboot of the client
| |
| 04:38 | I also tried to set the "option ntp-servers" and "option time-servers" into /etc/ltsp/dhcpd.conf
| |
| 04:38 | but still the wrong time
| |
| 04:39 | <alkisg> What do you have in /etc/timezone, both in the server and in the chroot?
| |
| 04:39 | <Scormen> server: Europe/Brussels
| |
| 04:39 | chroot: Europe/Brussels
| |
| 04:40 | <alkisg> And do you still have the ntp settings in dhcpd.conf ?
| |
| 04:40 | <Scormen> now I have yes, just rebooted the client, but again the wrong date
| |
| 04:41 | <alkisg> (I assume you did sudo invoke-rc.d dhcp3-server restart)
| |
| 04:41 | <Scormen> yes, i restarted dhcp3-server
| |
| 04:42 | but didn't build the image after that, but that isn't supposed?
| |
| 04:42 | <alkisg> No, no need.
| |
| 04:42 | I only have one last idea, about if the time is in UTC or not, let me find the file...
| |
| 04:44 | It's in /etc/default/rcS
| |
| 04:44 | <Scormen> UTC=yes
| |
| 04:44 | <alkisg> and in the chroot?
| |
| 04:45 | <Scormen> in chroot also UTC=yes
| |
| 04:45 | K_Wolf has quit IRC | |
| 04:46 | <alkisg> Scormen: Could you try this?
| |
| 04:46 | 1) Sync the time as we did before
| |
| 04:46 | 2) Stop the ntpd daemon in the server
| |
| 04:46 | <Scormen> we can try everything... ;)
| |
| 04:46 | <alkisg> 3) Reboot the client
| |
| 04:46 | If the time stays correct, it's a problem in your ntp server...
| |
| 04:47 | If it doesn't, I raise my hands :)
| |
| 04:47 | <Scormen> time synced
| |
| 04:47 | ntp on server stopped
| |
| 04:48 | rebooting
| |
| 04:49 | raise your hands.... again the wrong time on the client
| |
| 04:49 | Guest60180 has quit IRC | |
| 04:49 | <alkisg> Yup. Does it have any BIOS you can get into and set the time?
| |
| 04:50 | (and then reboot and see if it keeps it?)
| |
| 04:50 | <Scormen> in can cat in bios yes, but i would like to use automatic shutdown and WoL on the clients, so the same time on all clients should be nice
| |
| 04:51 | okay, i'll set the time in the bios by hand now
| |
| 04:52 | jammcq has quit IRC | |
| 04:53 | <Scormen> hmmm, the bios time is also quite different...
| |
| 04:53 | there it is 17:55:22
| |
| 04:53 | <alkisg> That's with UTC=no
| |
| 04:53 | <Scormen> ah, thats possible
| |
| 04:54 | <alkisg> I mean this: (1) Set the time in BIOS (2) Reboot and go into BIOS again, the time should be OK (3) Boot the client normally with ntpd stopped in your server. Notice the time. (4) Go into BIOS again. Again notice the time.
| |
| 04:54 | If the time is changed in (4), then something that we don't know about is changing your client's time...
| |
| 04:55 | <Scormen> 1 set time in bios: done
| |
| 04:55 | 2 reboot and look again in bios: done, same time what i've set
| |
| 04:56 | 3 server ntpd is stopped, time on client is wrong, it is 13:55
| |
| 04:56 | jammcq has joined #ltsp | |
| 04:57 | <Scormen> 4 rebooted again, in bios the time is still okay
| |
| 04:57 | <alkisg> What's the timezone in brussels? UTC+2 ?
| |
| 04:58 | jammcq_ has joined #ltsp | |
| 04:59 | <Scormen> utc+1, but it is summer now, utc+2 then
| |
| 04:59 | yoshi has joined #ltsp | |
| 05:00 | <alkisg> So the client time is OK, just the UTC=yes is the problem with the bios. Let's move on:
| |
| 05:00 | (1) boot the client, (2) start the ntp server, (3) manually sync and notice the time, (4) go into bios and notice the time.
| |
| 05:00 | yoshi is now known as Guest56099 | |
| 05:00 | jammcq has quit IRC | |
| 05:01 | <Scormen> 1 client booted, 2 ntpd started
| |
| 05:02 | 3: i get "no server suitable for synchronization found"
| |
| 05:03 | ntpd on the server is started well
| |
| 05:05 | <alkisg> Scormen: Maybe the basic problem is that you don't have hwclock in your client's path.
| |
| 05:06 | <Scormen> local client "hwclock" prints the right time
| |
| 05:08 | <alkisg> So if the root runs it, it's in the path?
| |
| 05:08 | (not a local user)
| |
| 05:11 | <Scormen> i mean, if i enter the command "hwclock" by root, it gives me the right time, the time i have set in the bios?
| |
| 05:11 | <alkisg> OK so it's in the path :)
| |
| 05:12 | (hwclock --utc would show the wrong time)
| |
| 05:13 | <Scormen> indeed, that prints 14:12
| |
| 05:13 | hwclock prints oke, 12:12
| |
| 05:14 | <alkisg> So now if you reboot normally with the ntp server on, what time does the client say?
| |
| 05:16 | <Scormen> rebooted, client says 14:15
| |
| 05:18 | <alkisg> That's correct (the bios time is wrong now).
| |
| 05:19 | Now, in a root shell, try these: ntpdate your-server-ip; hwclock --systohc
| |
| 05:19 | That's what the time init script does... let's try to do the same.
| |
| 05:20 | <Scormen> done
| |
| 05:20 | <alkisg> What's the time? :)
| |
| 05:20 | <Scormen> time is now 12:20
| |
| 05:20 | thats right
| |
| 05:20 | <alkisg> OK, please reboot and see the bios time...
| |
| 05:21 | Guest56099 has quit IRC | |
| 05:21 | jammcq_ has quit IRC | |
| 05:21 | gea has quit IRC | |
| 05:21 | nikosapi has quit IRC | |
| 05:21 | sutula has quit IRC | |
| 05:21 | dberkholz has quit IRC | |
| 05:21 | Egyptian[Home] has quit IRC | |
| 05:21 | <Scormen> *fingers crossed*
| |
| 05:21 | Guest56099 has joined #ltsp | |
| 05:21 | jammcq_ has joined #ltsp | |
| 05:21 | Egyptian[Home] has joined #ltsp | |
| 05:21 | nikosapi has joined #ltsp | |
| 05:21 | dberkholz has joined #ltsp | |
| 05:21 | sutula has joined #ltsp | |
| 05:21 | gea has joined #ltsp | |
| 05:21 | <Scormen> bios time is 12:21
| |
| 05:21 | <alkisg> Wow, that's wrong...
| |
| 05:22 | <Scormen> it had to be 10:21?
| |
| 05:22 | <alkisg> Yes
| |
| 05:23 | Please continue with the booting to see the time for one last time
| |
| 05:23 | <Scormen> booted, client time is now 14:22
| |
| 05:24 | <alkisg> OK. I think everything will be fixed if you put UTC=no in the client.
| |
| 05:24 | <Scormen> lets try
| |
| 05:24 | <alkisg> In /etc/default/rcS, and you'll need ltsp-update-image.
| |
| 05:25 | (in /opt/ltsp/amd64/etc/default/rcS I mean)
| |
| 05:25 | <Scormen> oke :)
| |
| 05:25 | rebooting
| |
| 05:26 | booted, client time is oke :) 12:25
| |
| 05:26 | <alkisg> You already did ltsp-update-image? Wow, you're server's fast :)
| |
| 05:26 | To confirm it's fixed, put a wrong time in the bios and see if it get's resynced ok
| |
| 05:27 | <Scormen> for trying i put 2 hdd's in raid 0 ;) quite fast then
| |
| 05:28 | putting a wrong time in bios now
| |
| 05:28 | rebooting again, in bios it is now 16:28
| |
| 05:29 | client booted, time is 16:28....
| |
| 05:30 | <alkisg> damn
| |
| 05:31 | <Scormen> maybe just a problem with ubuntu 9.04?
| |
| 05:31 | <alkisg> it works for me... :(
| |
| 05:31 | <Scormen> hmmm
| |
| 05:33 | strange things
| |
| 05:37 | pushing the time from the server to the clients isn't possible?
| |
| 05:40 | alkisg has quit IRC | |
| 05:45 | alkisg has joined #ltsp | |
| 05:48 | <Scormen> maybe i can best reinstall the whole server
| |
| 05:48 | it is just a one for testing, so it isn't a big problem
| |
| 05:49 | <alkisg> Scormen: if you do that, select "install an ltsp server" with F4, it's the most ...supported way to install ltsp in ubuntu.
| |
| 05:49 | <Scormen> i tried that already
| |
| 05:49 | but it fails always on 50%
| |
| 05:49 | i had the same problem with 8.04 and 8.10
| |
| 05:50 | <alkisg> Try with the server unplugged from the internet
| |
| 05:51 | <Scormen> aha oke, it was always connected with the internet
| |
| 05:51 | can that give the problem?
| |
| 05:51 | <ogra> how did it fail ?
| |
| 05:51 | and did you file a bug ?
| |
| 05:52 | <Scormen> it fails with a red screen
| |
| 05:52 | damn, what was that error.... *-)
| |
| 05:53 | <ogra> if you get something like that in the installer, hit alt-f4 ;) it has a lot of details about the actual error
| |
| 05:53 | a bug would have been nice though ...
| |
| 05:53 | now its to late to fix any installer issues
| |
| 05:53 | (for 9.04 at least)
| |
| 05:53 | <Scormen> it was something like "failed to build thin client image"
| |
| 05:54 | <ogra> yeah, that doesnt say much
| |
| 05:54 | <Scormen> no, "faild to build thin client system" i believe
| |
| 05:54 | indeed
| |
| 05:54 | oke, i'll try again, if i get that error, alf-f4
| |
| 05:54 | <ogra> a bug with the logs attached would have been helpful
| |
| 05:54 | <Scormen> with or without a internet connection?
| |
| 05:55 | <ogra> shouldnt matter at all
| |
| 05:55 | <Scormen> oke
| |
| 05:55 | thanks for your time guys, i'm going to do the reinstall
| |
| 06:05 | vvinet_ has quit IRC | |
| 06:08 | vvinet_ has joined #ltsp | |
| 06:10 | treyh has joined #ltsp | |
| 06:16 | treyh has quit IRC | |
| 06:19 | pmatulis has joined #ltsp | |
| 06:44 | rjune_home has quit IRC | |
| 06:47 | yoshi has joined #ltsp | |
| 06:47 | yoshi is now known as Guest65759 | |
| 06:55 | grami has joined #ltsp | |
| 06:56 | <grami> can anyone help me with a ltsp boot problem?
| |
| 06:57 | <alkisg> !ask
| |
| 06:57 | jammcq_ has quit IRC | |
| 06:57 | <ltspbot> alkisg: "ask" is Don't ask to ask a question, simply ask it, and if someone knows the answer, they'll respond. Please hang around for at least 15 minutes after asking a question, as not everybody constantly monitors the channel.
| |
| 06:58 | <grami> when booting a thin client it loads pxelinux.0 and starts to find the files in pxelinux.cfg but the time it takes to do this is over 30 mins
| |
| 07:00 | Guest56099 has quit IRC | |
| 07:02 | <Scormen> thats a long time, how fast is your connection?
| |
| 07:03 | ogra and alkisg: i'm reinstalling the server now, i'm a the failing step "failed to build ltsp chroot"
| |
| 07:04 | can i put this the error (alt-f4) on a usb disk or something, and how?
| |
| 07:06 | <grami> 100 mbit thinclient 1000 mbit server
| |
| 07:08 | <alkisg> grami: are you using tftpd-hpa?
| |
| 07:08 | <Scormen> what "speed" do you see when you do "ethtool your-interface" on the server?
| |
| 07:11 | vvinet_ has quit IRC | |
| 07:13 | <grami> alkisg: yes
| |
| 07:13 | i'm using tftp-hpa
| |
| 07:13 | <alkisg> grami: is it possible to try with a cross over cable instead of a switch?
| |
| 07:14 | <grami> have tried cross-over and still the same thing
| |
| 07:14 | <alkisg> grami: ok. Can you try it locally?
| |
| 07:14 | E.g. apt-get install tftp-hpa (the client),
| |
| 07:14 | and then tftp localhost and get /../...somefile
| |
| 07:14 | (all that on the server)
| |
| 07:15 | (or better yet on another machine)
| |
| 07:15 | <grami> will try
| |
| 07:16 | <alkisg> Scormen: try this from *another* console: cat /dev/vcs4
| |
| 07:17 | If it works, redirect it to a file and scp it somewhere...
| |
| 07:21 | pmatulis has quit IRC | |
| 07:25 | pmatulis has joined #ltsp | |
| 07:28 | <grami> localhost Fast, second PC fast but still on thin client slow
| |
| 07:28 | <Scormen> alkisg: the file was filled with data from mounting the usb stick
| |
| 07:30 | i have "repeated" the failing step
| |
| 07:30 | here is the output of the repeating failing step: http://sial.org/pbot/paste
| |
| 07:31 | srr, here it is: http://sial.org/pbot/36632
| |
| 07:31 | vvinet has joined #ltsp | |
| 07:37 | <alkisg> grami: try replacing pxelinux.0 with another version
| |
| 07:38 | Scormen: I've no idea what to do with this... maybe file a bug? I'd also try without internet connection...
| |
| 07:39 | <grami> pxelinux ver 2.13
| |
| 07:39 | will try now
| |
| 07:39 | <Scormen> "file a bug"?
| |
| 07:40 | <alkisg> Scormen: like ogra told you... I don't know anything about debian-installer. :)
| |
| 07:40 | <Scormen> aah oke :)
| |
| 07:40 | i'll try some other things, like no internet connection like you said
| |
| 07:41 | <ogra> that looks like you didnt reformat the disk or something
| |
| 07:42 | ltsp-client-builder is supposed to fail if the target dir (/opt/ltsp/$ARCH) exists
| |
| 07:42 | <alkisg> (that would be because he repeated the ltsp-build-client step)
| |
| 07:43 | (so Scormen if you're still on the same step, rm -rf the dir before repeating the step...)
| |
| 07:46 | <grami> Found the problem, Forgot to put "next-server " in the dhcpd.conf file thanks guys
| |
| 07:47 | <alkisg> grami: but then how did it download the files (even after 30 minutes)?
| |
| 07:47 | <ogra> yeah, next-server onbly matters if your tftp server is on a different machine
| |
| 07:49 | <grami> this is a unkown to me. alkisg
| |
| 07:51 | <ogra> especially next-server only matters *before* pxelinux.0 is downloaded
| |
| 07:51 | _Shane has joined #ltsp | |
| 07:53 | <grami> just one more thing I hope, when the image loads on this new thin client. it does a modprobe on 8139too and then kernel panic dew to the fact it can't find a module. any way to install the module?
| |
| 07:55 | * ogra wonders what distro that is, 8139too should be a default module in all supported ltsp5 distros | |
| 07:56 | <Scormen> its already reinstalling again. Yes, that dir was existing because of the repeating step
| |
| 07:58 | but the installer did something strange... i have 3 network interfaces on the server. eth0 is de outside one. but in that log (before putting it on the usb disk) i could see that he made only two interfaces: "eth0" and "eth1eth2". seems that something is going wrong with "eth1eth2"....
| |
| 07:59 | Egyptian[Home] has quit IRC | |
| 08:00 | Egyptian[Home] has joined #ltsp | |
| 08:00 | <Scormen> the installer is running again now, with a automatic ip deliver from the router. other times i was setting the ip manual... maybe that's a problem.... we'll see soon
| |
| 08:09 | alkisg: the install failed again
| |
| 08:10 | i put in my usb stick before the failing step, so this is the complete error:
| |
| 08:10 | http://sial.org/pbot/36637
| |
| 08:11 | <ogra> file a bug with that log snippet attached, looks like a bug in the interface detection logic
| |
| 08:11 | <Scormen> indeed
| |
| 08:11 | the third last line, you see the "eth1eht2"
| |
| 08:12 | <alkisg> Scormen: heh... so to bypass it you just need to remove one of the eth* interfaces? :)
| |
| 08:12 | <Scormen> yeah, but i don't want to break down my motherboard ;)
| |
| 08:13 | <alkisg> Wow, 3 embedded NICs? Nice...
| |
| 08:13 | Gadi1 has joined #ltsp | |
| 08:14 | Gadi1 is now known as Gadi | |
| 08:14 | <ogra> Scormen, well, given you likely dont want to serve amd64 clients anyway you should just run ltsp-build-client after installation
| |
| 08:14 | <Scormen> hmm but the server cd i don't have that problem... first time i see it
| |
| 08:14 | <ogra> (unless you actually *want* to serve amd64 clients) :)
| |
| 08:15 | the server cd doesnt have the ltsp mode in the installer
| |
| 08:15 | at least it would be news to me if it was added
| |
| 08:16 | <Scormen> i don't have to serve amd64, i386 is also good :)
| |
| 08:16 | <ogra> right
| |
| 08:16 | so dont even use the installer part
| |
| 08:16 | <Scormen> the server cd doesn't have the installer no, but i just mean that i didn't see that eth1eth2 error before
| |
| 08:16 | <ogra> just install normally and after install run sudo ltsp-build-client --arch i386 --copy-sourceslist
| |
| 08:17 | its *in* the ltsp part
| |
| 08:17 | <Scormen> thats what i did 2 days ago (but arch amd64)
| |
| 08:17 | <ogra> it runs a sed command over the output of th eip command
| |
| 08:17 | *the ip command
| |
| 08:18 | likely that sed command has a wrong match
| |
| 08:18 | <Scormen> aha
| |
| 08:19 | i'll first report the bug to ubuntu and after that, i'll reinstall the server
| |
| 08:19 | <ogra> yeah
| |
| 08:19 | <grami> what can you do if you get a nfs permission denied ( kernel panic)
| |
| 08:19 | <ogra> make sure your nfs server is configured properly
| |
| 08:19 | what distro is that ? and which release ?
| |
| 08:20 | <grami> i can mount for a second pc but not from the thin client, using gentoo ltsp 4.2
| |
| 08:20 | <ogra> ugh. ltsp 4
| |
| 08:21 | * ogra cant help with ltsp4 its dead | |
| 08:21 | <grami> thats not very nice of you....
| |
| 08:22 | <Gadi> grami: its only because it predates him
| |
| 08:22 | <ogra> grami, ?
| |
| 08:22 | <Gadi> :)
| |
| 08:22 | <grami> Gadi: lol
| |
| 08:22 | <ogra> grami, ltsp4 has not been developed in 4 years ... and has no security updates since than
| |
| 08:22 | you should definately not use it
| |
| 08:23 | <grami> i would love to use ltsp 5 on gentoo but for some reason it will not let me install.
| |
| 08:24 | <Gadi> grami: if johnny is around this room, he can help with ltsp5 and gentoo
| |
| 08:24 | <ogra> yeah, he made the gentoo port
| |
| 08:24 | <Gadi> as can a few others, im sure
| |
| 08:24 | otherwise, if you need to use 4.2 for now,
| |
| 08:24 | * ogra thinks knipwim works with ltsp5 on gentoo too | |
| 08:25 | <Gadi> your error is most likely a network driver error
| |
| 08:25 | <grami> i need to use it now and will upgrade later
| |
| 08:25 | <Gadi> do you know what NIC is in the client?
| |
| 08:25 | <ogra> there is no upgrade path from 4.x to 5 ... you will have to redo it for ltsp5
| |
| 08:25 | <grami> the NIC is realtek 8139too
| |
| 08:26 | <Gadi> hmm...
| |
| 08:26 | that should have been supported
| |
| 08:26 | <grami> i can find the module in /lib/modules
| |
| 08:26 | <Gadi> well, you only care that they are in the initrd
| |
| 08:26 | you can gunzip the initrd and mount it on a loopback
| |
| 08:27 | and check it that way
| |
| 08:27 | otherwise, you can specify NIC=8139too as a kernel command line option
| |
| 08:27 | <grami> ok you lost me abit there
| |
| 08:28 | <Gadi> maybe try that firs
| |
| 08:28 | is it pxe?
| |
| 08:29 | tadeu_ has joined #ltsp | |
| 08:29 | <Scormen> alkisg: any idea what that "error code 255" means? something of ubuntu itself, or from ltsp?
| |
| 08:29 | "May 15 13:04:27 main-menu[804]: WARNING **: Configuring 'ltsp-client-builder' failed with error code 255"
| |
| 08:30 | <ogra> that just means that the 'ltsp-client-builder' shellscript exited with code 255
| |
| 08:30 | tadeu_ has quit IRC | |
| 08:30 | * Gadi wonders what ltsp-client-builder is | |
| 08:30 | <ogra> i.e. it had an error
| |
| 08:30 | the debian-installer udeb
| |
| 08:31 | that gets run on ubuntu if you select the option to install an out of the box ltsp server from the CD
| |
| 08:31 | <Gadi> craziness
| |
| 08:31 | :)
| |
| 08:32 | <Scormen> lol
| |
| 08:32 | <ogra> well, it just calls ltsp-build-client and sets up your second NIc
| |
| 08:32 | when i developed it i never had a three NIC setup ... which means the three NIC situation isnt much tested
| |
| 08:33 | <Gadi> thats worth its weight in gold
| |
| 08:33 | <Scormen> yes, but any idea what "code 255" is? shouldn't there be a list of these codes or something i mean?
| |
| 08:33 | * ogra wonders how much 100 lines of code weight though | |
| 08:33 | <Gadi> so, it works on systems with n NICs where n is 2
| |
| 08:33 | :)
| |
| 08:33 | <Scormen> lol, great
| |
| 08:33 | <ogra> Scormen, google for bash exit code 255 i guess
| |
| 08:33 | Gadi, 1 or 2 nics
| |
| 10:28 | ltspbot has joined #ltsp | |
| 10:28 | ajohnson has joined #ltsp | |
| 10:29 | rjune has joined #ltsp | |
| 10:29 | <Scormen> alkisg, I have installed the server again
| |
| 10:29 | just a clean install, without the "ltsp" mode
| |
| 10:30 | <alkisg> Scormen: everything ok?
| |
| 10:30 | <Scormen> after that, i installed ltsp server standalone
| |
| 10:30 | installed ntp on the server itself
| |
| 10:30 | booted the client
| |
| 10:30 | wrong time
| |
| 10:30 | on the client did a ntpdate via root account
| |
| 10:30 | time oke
| |
| 10:30 | rebooted to check
| |
| 10:30 | time still oke :)
| |
| 10:31 | <alkisg> Scormen: but that looks like it doesn't actually update, so your ntp server might be misconfigured
| |
| 10:31 | I mean, from what you're saying I think that if you change the time in BIOS, it'll still be wrong when the client boots
| |
| 10:31 | <Scormen> i'll change the bios time
| |
| 10:32 | <alkisg> Scormen: only change it by 10 minutes
| |
| 10:32 | <Scormen> set bios time to 20:31, rebooting
| |
| 10:32 | oei
| |
| 10:33 | rebooted
| |
| 10:33 | thin client time is 17:32
| |
| 10:33 | mikkel has quit IRC | |
| 10:34 | <alkisg> So the time was correctly adjusted?
| |
| 10:34 | <Scormen> rebooted - bios time is 17:33 again
| |
| 10:34 | <Gadi> Scormen: I wonder if you have the code that was prebugfix
| |
| 10:34 | <Scormen> so, it seems to work now?
| |
| 10:34 | <Gadi> can you paste: /opt/ltsp/i386/usr/share/ltsp/ltsp-init-common?
| |
| 10:34 | <alkisg> !pastebot
| |
| 10:34 | <ltspbot> alkisg: "pastebot" is The LTSP pastebot is at http://pastebot.ltsp.org. Please paste all text longer than a line or two to the pastebot, as it helps to reduce traffic in the channel. A link to the content will be pasted in the channel.
| |
| 10:35 | <Scormen> oke
| |
| 10:35 | <Gadi> line #193 should read: (ntpdate $TIMESERVER; hwclock --systohc) &
| |
| 10:35 | <ltsppbot> "scormen" pasted "/opt/ltsp/i386/usr/share/ltsp/ltsp-init-common" (196 lines) at http://pastebot.ltsp.org/343
| |
| 10:36 | <Scormen> http://pastebot.ltsp.org/343
| |
| 10:36 | <Gadi> ah, thats right
| |
| 10:36 | you're good
| |
| 10:36 | what time zone do you have?
| |
| 10:37 | <alkisg> Scormen: so everything's fine even though UTC=yes in the chroot?
| |
| 10:37 | <Scormen> Europe/Brussels
| |
| 10:37 | yes, UTC=yes on server and chroot
| |
| 10:38 | <Gadi> so whats problem do you have?
| |
| 10:38 | just that the thin client is setting the hardware time?
| |
| 10:38 | <alkisg> That's strange... if the BIOS says it's 17:33, then when the client boots it should say 15:33
| |
| 10:39 | <Gadi> not if you set the timezone
| |
| 10:39 | <Scormen> in "/var/lib/tftpboot/ltsp/i386/lts.conf" I defined the TIMEZONE and TIMESERVER variable to
| |
| 10:39 | <Gadi> er, I mean if you set the timezone
| |
| 10:40 | <alkisg> The timezone is Europe/Brussels in the server, in the chroot, and in lts.conf...
| |
| 10:40 | <Scormen> (and now I have i386 for clients, not amd64 anymore)
| |
| 10:40 | japerry has joined #ltsp | |
| 10:40 | <alkisg> Well, if it works, all is fine, even though it's strange why it didn't work on the first boot.
| |
| 10:41 | <Scormen> indeed, timezone is "Europe/Brussels" in the lts.conf, /etc/timezone on the server and /etc/timezone in the chroot
| |
| 10:41 | <alkisg> (and on the previous amd64 installation :))
| |
| 10:41 | <Scormen> on the first boot i didn't specified the variables in lts.conf, maybe because of that?
| |
| 10:41 | <alkisg> Ah, sure
| |
| 10:41 | <Gadi> hmm... maybe it is a race condition to set the time in the background
| |
| 10:41 | <Scormen> (lol ;))
| |
| 10:41 | <Gadi> oh, or that
| |
| 10:42 | <alkisg> Scormen: so... I guess that was an amd64 bug?
| |
| 10:42 | (the previous installation)
| |
| 10:42 | <Scormen> i wish i could confirm...
| |
| 10:42 | rjune has quit IRC | |
| 10:43 | rjune has joined #ltsp | |
| 10:43 | <Scormen> i can't say i have done something else this time
| |
| 10:44 | <alkisg> Anyway, one bug was reported, another wasn't :), all fine now.
| |
| 10:47 | <Scormen> indeed
| |
| 10:47 | thanks for al the support today guys
| |
| 10:51 | <alkisg> Heh... first result in google, already: http://lmgtfy.com/?q=ltsp-client-builder.postinst
| |
| 10:53 | <Scormen> hehe
| |
| 10:53 | no, second ;)
| |
| 10:54 | <alkisg> Hmmm we go in different google servers :)
| |
| 10:54 | <Scormen> yeah i think so
| |
| 10:54 | hé, also first now :)
| |
| 10:54 | the first two switched here
| |
| 11:06 | rjune_ has joined #ltsp | |
| 11:14 | cliebow has quit IRC | |
| 11:19 | rjune has quit IRC | |
| 11:20 | lucascoala has joined #ltsp | |
| 11:26 | Remaille has joined #ltsp | |
| 11:29 | runout1 has joined #ltsp | |
| 11:31 | Remaille has quit IRC | |
| 11:37 | japerry has quit IRC | |
| 11:44 | vagrantc has joined #ltsp | |
| 11:58 | alkisg has quit IRC | |
| 12:07 | <pmatulis> i'm bewildered by the ouput to the mount command issued on the console of a client. where can i learn about this?
| |
| 12:07 | *output
| |
| 12:09 | <ltsppbot> "pmatulis" pasted "client mount output" (16 lines) at http://pastebot.ltsp.org/344
| |
| 12:20 | <vagrantc> pmatulis: what are you confused about?
| |
| 12:22 | <pmatulis> vagrantc: none on..., /cow?
| |
| 12:23 | <vagrantc> ??
| |
| 12:24 | i don't really understand your confusion
| |
| 12:25 | <pmatulis> vagrantc: alright, disregard
| |
| 12:34 | vagrantc: i have a more specific question regarding NBD
| |
| 12:34 | <vagrantc> ask away, and people will do what they can to answer
| |
| 12:35 | <pmatulis> vagrantc: the image (i386.img) is a compressed filesystem
| |
| 12:35 | vagrantc: i'm just not sure how this thing gets uncompressed
| |
| 12:36 | <Gadi> it doesnt
| |
| 12:36 | <vagrantc> it's mounted as a squashfs filesystem
| |
| 12:36 | it gets compressed on the fly when data is requested from it
| |
| 12:36 | <Gadi> the kernel has drivers to talk to compressed filesystem
| |
| 12:36 | <pmatulis> that's what i was looking for!
| |
| 12:36 | and that's cool
| |
| 12:36 | <vagrantc> er, decompressed ... the whole image doesn't ever get decompressed
| |
| 12:36 | it brings some headaches
| |
| 12:37 | but overall has seemed to work well for most cases
| |
| 12:37 | <pmatulis> since it remains remote what is the rationale for compression?
| |
| 12:37 | <Gadi> speed
| |
| 12:37 | <vagrantc> reduced network traffic
| |
| 12:37 | <pmatulis> ah yes
| |
| 12:38 | <Gadi> leading to speed
| |
| 12:38 | :)
| |
| 12:38 | <vagrantc> the need for speed
| |
| 12:38 | i blame ebox2300
| |
| 12:38 | * Gadi hugs nbd | |
| 12:38 | <Gadi> and squashfs
| |
| 12:38 | and the collective unionfs spin-offs
| |
| 12:38 | <vagrantc> nbd is where it's at. squashfs is negligable
| |
| 12:38 | <Gadi> :)
| |
| 12:39 | <pmatulis> Gadi: can you expound on these spin-offs?
| |
| 12:39 | <Gadi> squashfs is gold - nbd is nothing without it
| |
| 12:39 | :P
| |
| 12:39 | rjune_ has quit IRC | |
| 12:39 | <vagrantc> it is the lust for gold that makes people compromise their core values
| |
| 12:39 | <Gadi> pmatulis: the open source community seemd hell bent on implementing unionfs as many times as it can
| |
| 12:39 | *seems
| |
| 12:39 | alkisg has joined #ltsp | |
| 12:39 | <Gadi> they all pretty much do the same thing
| |
| 12:40 | in the spirit of open source
| |
| 12:40 | :)
| |
| 12:40 | <laga> yeah, but aufs actually works
| |
| 12:40 | ;)
| |
| 12:40 | <vagrantc> and they all do it *wrong*, so *I* am going to do it *Right*
| |
| 12:40 | laga: just the other day, ogra asked me to join a conversation regarding ditching aufs ....
| |
| 12:41 | <laga> vagrantc: yeah, i know it's being discussed.
| |
| 12:41 | <pmatulis> and this image is read-only right?
| |
| 12:41 | <vagrantc> yaufs
| |
| 12:42 | yet another unionfs
| |
| 12:42 | rjune_ has joined #ltsp | |
| 12:43 | * vagrantc should eat breakfast before dreaming up new unionfses | |
| 12:45 | <pmatulis> so read-only?
| |
| 12:46 | <alkisg> Couldn't ltspfs be modified a little so that it would support exporting the whole uncompressed /opt/ltsp/i386 ? (along with a tmpfs somewhere)?
| |
| 12:47 | runout1 has left #ltsp | |
| 12:49 | <epsas> i have been having problems using aufs with nfs directories
| |
| 12:49 | we have our home directories mounted over nfs -- i am unable to do an aufs overlay on these directories
| |
| 12:49 | it just hangs...
| |
| 12:49 | dunno why
| |
| 13:05 | cliebow has joined #ltsp | |
| 13:15 | litlebuda has joined #ltsp | |
| 13:19 | hanthana has quit IRC | |
| 13:23 | <pmatulis> what parses the lts.conf file during client boot?
| |
| 13:24 | <alkisg> getltscfg, it's a bison/flex/c program
| |
| 13:25 | <pmatulis> alkisg: thanks
| |
| 13:28 | alkisg: is that done before ltsp-client-setup
| |
| 13:29 | <alkisg> pmatulis: I think so.
| |
| 13:32 | <Gadi> pmatulis: it is called by /usr/share/ltsp/ltsp_config
| |
| 13:32 | <pmatulis> Gadi: nice, thanks
| |
| 13:35 | <vagrantc> epsas: for some reason, NFS and various unionfs implementations tend to have problems... though i had the impression there were maybe some mount options to specify fro aufs+nfs
| |
| 13:36 | CAN-o-SPAM has quit IRC | |
| 13:36 | litlebuda has quit IRC | |
| 13:36 | spectra has joined #ltsp | |
| 13:38 | <epsas> oh really now
| |
| 13:38 | i will need to take a look at the aufs manpage
| |
| 13:39 | although it is (understandably) hard to read
| |
| 13:43 | Scormen has quit IRC | |
| 13:54 | cliebow has quit IRC | |
| 14:19 | japerry has joined #ltsp | |
| 14:26 | Remaille has joined #ltsp | |
| 14:40 | rjune__ has joined #ltsp | |
| 14:58 | rjune_ has quit IRC | |
| 14:59 | X-TaZz has joined #ltsp | |
| 15:03 | Remaille has quit IRC | |
| 15:04 | litlebuda has joined #ltsp | |
| 15:15 | pmatulis has quit IRC | |
| 15:28 | rjune__ has quit IRC | |
| 15:34 | Mip5 has joined #ltsp | |
| 15:36 | <Mip5> Hey Gang - Are there any known issues with a raid1 install of ltsp on 9.04? I'm getting md0: unknown partition table. The last things I see are md:bind<sdc1> (which is my / partition) and md: bind<sdc5> (which is swap)
| |
| 15:37 | Then the machine just hangs. I've tried booting into the rescue mode, but the didn't see anything wrong there.
| |
| 15:37 | I did have a previous install on these same disks before, and didn't do anything to that other than trying a fresh install of 9.04
| |
| 15:38 | I would think that would have wiped the disks.
| |
| 15:41 | I should add that this occurred during the first boot after kicking out the install cd. I'm not wedded to this install, so could do it again, I'm just not sure what I'd choose differently.
| |
| 15:45 | <litlebuda> Mip5, im not sure its a ltsp problem , are you sure the raid is set up ok ?
| |
| 15:46 | <Mip5> I'm not sure how to verify the raid
| |
| 15:47 | I'm in the busybox environment now, and just mounted /dev/md0, and was able to do that
| |
| 15:47 | I am able to look around and see stuff, and it *apears* to all be there
| |
| 15:47 | <litlebuda> how many md do you have ?
| |
| 15:48 | <Mip5> 2 - one for / and the other for swap (which I've heard is debatable, but I figured wouldn't hurt, and have done before)
| |
| 15:48 | md0 is /
| |
| 15:48 | <vagrantc> i have heard that swap doesn't like software raid
| |
| 15:49 | <litlebuda> can you coment out md1 on fstab ?
| |
| 15:49 | <Mip5> I've heard that, but I've been running it this way on another box
| |
| 15:49 | litlebuda - good idea, I can try that. Maybe something changed in jaunty
| |
| 15:50 | <litlebuda> vagrantc, im using it here with no problems too :)
| |
| 15:50 | <vagrantc> what doesn't give you a backache may give you a headache
| |
| 15:50 | <Mip5> ;-)
| |
| 15:50 | <litlebuda> :)
| |
| 15:50 | <Mip5> Um.... how can I edit a file in initramfs? vi doesn't appear to exist
| |
| 15:50 | <vagrantc> i'm not saying it can't be done, but there may be particular situations that trigger the problem
| |
| 15:51 | <litlebuda> sure vagrantc
| |
| 15:51 | <vagrantc> Mip5: chroot into the filesystem and use your tools
| |
| 15:51 | <Mip5> ok - thanks
| |
| 15:52 | I can't seem to do that - I lost my initramfs prompt, but didn't get anything in return. I think I'll have to boot into a recovery cd
| |
| 15:53 | <litlebuda> vagrantc, i have installed ltsp on ubuntu 8.4 and i cant seem to access the dvd drive or a flash drive . any sugestions for some reading ?
| |
| 15:53 | <vagrantc> not my cup of tea, so much
| |
| 15:53 | <litlebuda> ok
| |
| 15:53 | thanks anyway
| |
| 15:53 | <Mip5> litlebuda - have you enabled it in lts.conf?
| |
| 15:54 | <vagrantc> oh, wait, you're talking about LOCALDEV stuff?
| |
| 15:54 | should be enabled by default...
| |
| 15:54 | LOCALDEV=true in lts.conf
| |
| 15:54 | !localdev
| |
| 15:54 | <ltspbot> vagrantc: "localdev" is (#1) Information regarding using pluggable devices and cdroms with LTSP can be found at http://wiki.ltsp.org/twiki/bin/view/Ltsp/LTSP-42-LocalDev, or (#2) for ubuntu: https://wiki.ubuntu.com/EnableLTSP5LocalDevices, or (#3) for debian-edu: http://wiki.debian.org/DebianEdu/HowTo/LocalDeviceLtspfs, or (#4) http://wiki.ubuntu.com/DebugLocalDev
| |
| 15:55 | <vagrantc> some of those links may be a bit out of date, though
| |
| 15:55 | <litlebuda> will take a look
| |
| 15:55 | <alkisg> litlebuda: localdev is enabled by default, are your client in the fuse group?
| |
| 15:56 | <litlebuda> yes
| |
| 15:56 | <alkisg> (the user, I mean)
| |
| 15:56 | * vagrantc is only mildly familiar with ubuntu | |
| 15:56 | <Mip5> vagrantc - how would you recommend setting up an ltsp server? Would you raid / ?
| |
| 15:56 | <litlebuda> it was a fresh install from the alternate cd with the option to install ltsp
| |
| 15:56 | <vagrantc> Mip5: depends on the needs ...
| |
| 15:57 | <Mip5> high availability - it will be the main lab for a middle school w/300 students
| |
| 15:57 | 25 stations
| |
| 15:57 | They are planning to use google docs, so that kids can get to their work from anywhere
| |
| 15:57 | So I don't need large home directory for them
| |
| 15:58 | lucascoala has quit IRC | |
| 15:58 | <vagrantc> i'd probably go with raid10 /, raid1 /boot, and raid10 /home or something along those lines...
| |
| 15:59 | and make sure to mount / and /home with relatime or noatime
| |
| 15:59 | that'd give you good redundancy and reliable raid recovery...
| |
| 16:03 | <Mip5> !raid
| |
| 16:03 | <ltspbot> Mip5: Error: "raid" is not a valid command.
| |
| 16:04 | <Mip5> Oh -well, worth a shot. Is Raid10 mirrored and stripping? Really I just want high availability. If speed gains are only marginal, they are not nearly as important as uptime.
| |
| 16:07 | elisboa has quit IRC | |
| 16:08 | <vagrantc> raid1 is slow writes, fast reads.
| |
| 16:08 | it's the simplest and most reliable form of raid, if you don't mind the slow writes.
| |
| 16:08 | <Mip5> vagrantc or anyone - can you tell me how to make sure my disks are clean? How do I get rid of any residual info on them regarding previous raid identities?
| |
| 16:08 | thanks - I don't mind the slow writes
| |
| 16:09 | <vagrantc> using relatime or noatime really helps a lot with raid1, otherwise all reads are effectively writes.
| |
| 16:10 | <Mip5> okay - I think those options were already in fstab, but I'll make sure next time around (I'm trying a clean install as I write this)
| |
| 16:20 | spectra has quit IRC | |
| 16:21 | <Mip5> FTR - I'm pretty sure I didn't erase the disks prior to last night's install. I wouldnt' have thought that would make a difference, but maybe it did. Doing that now
| |
| 16:26 | vvinet has quit IRC | |
| 16:28 | <vagrantc> "dd if=/dev/zero of=/dev/NNN bs=1024k count=1" on each partition ought to work wonders.
| |
| 17:01 | Gadi has left #ltsp | |
| 17:09 | alkisg has quit IRC | |
| 17:12 | shogunx has quit IRC | |
| 17:12 | Mip5 has left #ltsp | |
| 17:17 | Mip5 has joined #ltsp | |
| 17:18 | strattog has quit IRC | |
| 17:23 | strattog has joined #ltsp | |
| 17:26 | evilx has quit IRC | |
| 17:27 | pmatulis has joined #ltsp | |
| 17:39 | jammcq has joined #ltsp | |
| 17:40 | vagrantc has quit IRC | |
| 18:03 | Mip5 has quit IRC | |
| 18:09 | artista_frustrad has joined #ltsp | |
| 18:18 | X-TaZz has quit IRC | |
| 19:04 | litlebuda has quit IRC | |
| 19:25 | vvinet has joined #ltsp | |
| 19:35 | shogunx has joined #ltsp | |
| 19:41 | Egyptian[Home] has quit IRC | |
| 19:55 | Egyptian[Home] has joined #ltsp | |
| 20:10 | yanu has quit IRC | |
| 20:14 | yanu has joined #ltsp | |
| 20:30 | try2free has joined #ltsp | |
| 21:18 | lucascoala has joined #ltsp | |
| 21:32 | pmatulis has quit IRC | |
| 21:52 | zamba_ has joined #ltsp | |
| 21:53 | zamba has quit IRC | |
| 22:06 | hanthana has joined #ltsp | |
| 22:09 | artista_frustrad has quit IRC | |
| 22:09 | artista_frustrad has joined #ltsp | |
| 23:50 | hanthana_ has joined #ltsp | |