00:11 | Selveste1 has quit IRC | |
00:11 | gentgeen__ has quit IRC | |
00:11 | zamba has quit IRC | |
00:13 | Mava has quit IRC | |
00:14 | Mava has joined #ltsp | |
00:16 | highvoltage has quit IRC | |
00:16 | gentgeen__ has joined #ltsp | |
00:16 | zamba has joined #ltsp | |
00:21 | Selveste1 has joined #ltsp | |
00:36 | Selveste1 has quit IRC | |
00:39 | Faithful has joined #ltsp | |
00:39 | johnny has left #ltsp | |
00:40 | johnny has joined #ltsp | |
00:40 | alkisg has quit IRC | |
00:50 | _UsUrPeR_ has quit IRC | |
00:55 | _UsUrPeR_ has joined #ltsp | |
01:08 | alkisg has joined #ltsp | |
01:46 | alkisg has quit IRC | |
02:00 | johnny has quit IRC | |
02:02 | gnunux has joined #ltsp | |
02:05 | dobber has joined #ltsp | |
02:10 | johnny has joined #ltsp | |
02:16 | Selveste1 has joined #ltsp | |
02:51 | Selveste1 has quit IRC | |
02:53 | yanu has quit IRC | |
03:00 | Selveste1 has joined #ltsp | |
03:00 | sepski has joined #ltsp | |
03:01 | yanu has joined #ltsp | |
03:01 | yanu has joined #ltsp | |
03:31 | Selveste1_ has joined #ltsp | |
03:35 | Selveste1 has quit IRC | |
03:36 | Selveste1_ has quit IRC | |
03:53 | Egyptian[Home] has quit IRC | |
04:05 | Faithful has quit IRC | |
04:09 | Egyptian[Home] has joined #ltsp | |
04:32 | bobby_C has joined #ltsp | |
04:39 | alkisg has joined #ltsp | |
04:39 | alkisg has joined #ltsp | |
04:45 | Egyptian[Home] has quit IRC | |
04:47 | Egyptian[Home] has joined #ltsp | |
05:21 | alkisg has quit IRC | |
05:26 | highvoltage has joined #ltsp | |
05:45 | alkisg has joined #ltsp | |
05:55 | hersonls has joined #ltsp | |
06:15 | alkisg1 has joined #ltsp | |
06:17 | alkisg has quit IRC | |
06:17 | alkisg1 is now known as alkisg | |
06:22 | hersonls has quit IRC | |
06:32 | pasmen has joined #ltsp | |
06:38 | lucascoala_ has quit IRC | |
06:40 | pmatulis has joined #ltsp | |
07:05 | highvoltage has quit IRC | |
07:09 | AlTeeAsP has quit IRC | |
07:16 | mikkel has joined #ltsp | |
07:19 | _UsUrPeR_ has quit IRC | |
07:19 | _UsUrPeR_ has joined #ltsp | |
07:21 | bobby_C has quit IRC | |
07:26 | _UsUrPeR_ has quit IRC | |
07:47 | jammcq has quit IRC | |
07:53 | _UsUrPeR_ has joined #ltsp | |
07:58 | alkisg has quit IRC | |
08:06 | mgariepy has joined #ltsp | |
08:12 | Gadi has joined #ltsp | |
08:18 | <mgariepy> morning everyone
| |
08:25 | pasmen has quit IRC | |
08:26 | highvoltage has joined #ltsp | |
08:30 | ogra has quit IRC | |
08:30 | ogra has joined #ltsp | |
08:39 | cliebow has joined #ltsp | |
08:41 | wsuetholz has joined #ltsp | |
08:46 | anivair has joined #ltsp | |
08:47 | <anivair> does anyone know how ltsp thin clients can be made to work externally? I've got a few people who need to work remotely, and I'd way prefer they use thin clients.
| |
09:03 | highvoltage has quit IRC | |
09:07 | Remaille has joined #ltsp | |
09:09 | highvoltage has joined #ltsp | |
09:14 | highvoltage has joined #ltsp | |
09:15 | mikkel has quit IRC | |
09:15 | highvolt1ge has joined #ltsp | |
09:19 | highvoltage has quit IRC | |
09:34 | alkisg has joined #ltsp | |
09:36 | evil_root has joined #ltsp | |
09:45 | <sbalneav> Morning all
| |
09:49 | The_Code has joined #ltsp | |
09:50 | <anivair> morning.
| |
09:51 | you're here just in time for me to re-ask my question from an hour ago or so
| |
09:51 | (the short version): what do I need to do to get my thin clients running remotely? I have some people that need to work from home.
| |
09:51 | <alkisg> neatx :)
| |
09:53 | <anivair> is that required? I mean, i'm guessing that I can somehow set my firewall to X-forward and route the data, but I'm not sure how
| |
09:53 | is there maybe a page with a useful how to?
| |
09:54 | <evil_root> cant you just vpn your work at homes into your ltsp network?
| |
09:54 | <alkisg> anivair, ltsp is designed for LAN, not for remote connections
| |
09:54 | The X traffic is to great for non-LAN connections
| |
09:54 | That's why nx got invented
| |
09:54 | <evil_root> can anyone say lag?
| |
09:54 | highvolt1ge has quit IRC | |
09:55 | <anivair> hrm. that does suck.
| |
09:55 | <alkisg> anivair: it's very easy to do remote X. It'll just be super slow.
| |
09:55 | highvoltage has joined #ltsp | |
09:56 | <alkisg> You can install neatx though and access your server from any OS and any connetion, with plain old ssh as the transport
| |
09:57 | <anivair> will neatx be faster, then? Or just easier?
| |
09:58 | <alkisg> Much, much faster
| |
09:59 | <anivair> okay ... I'll give that a shot (I'm, at the moment, 100% unsure how, but I'll poke around and see what I can find)
| |
09:59 | <alkisg> distro/version?
| |
10:03 | F-GT has quit IRC | |
10:03 | <anivair> ubuntu lucid
| |
10:04 | <dobber> oh goody
| |
10:04 | brand new fatserver just arived
| |
10:04 | michiel_e has joined #ltsp | |
10:04 | highvolt1ge has joined #ltsp | |
10:05 | <alkisg> anivair: sudo add-apt-repository ppa:freenx-team && sudo apt-get update && sudo apt-get install neatx-server
| |
10:05 | highvoltage has quit IRC | |
10:05 | <alkisg> ...and google for the proprietary nx client from nomachines, you can find .deb for ubuntu or .exe for windows
| |
10:06 | <anivair> and that needs to be installed on hte ltsp server?
| |
10:06 | <alkisg> Yes, the commands above go to the ltsp-server, while the nx client on the remote machines
| |
10:07 | The theory is there: https://help.ubuntu.com/community/FreeNX
| |
10:07 | <cliebow> neatx is an nx-y thing?
| |
10:07 | <alkisg> ...but freenx was replaced with neatx for lucid
| |
10:07 | <cliebow> ahhhh
| |
10:07 | <alkisg> cliebow: yeah, the google implementation of freenx-server
| |
10:08 | <cliebow> coolio..
| |
10:08 | <alkisg> It works out of the box - an apt-get install is all that it takes
| |
10:08 | <cliebow> out of the box is even better!
| |
10:09 | <anivair> So can i seta thin client to access the server that way? Won't I need to make a change to the thin client?
| |
10:09 | <alkisg> No, no thin clients. Regular standalone machines, like those that people have in their homes
| |
10:10 | I.e. you boot an os locally and run the nx client afterwards
| |
10:10 | You don't boot an os from the network, like thin clients do
| |
10:10 | So, LAN <=> LTSP, WAN <=> nx
| |
10:10 | <anivair> suck. I was really hoping not to have to deal with them having computers in their homes.
| |
10:10 | <alkisg> ?
| |
10:10 | <anivair> because then i'm either giving them systems or dealing with their crappy computers hat they load with spyware
| |
10:11 | <alkisg> You wanted them to pull a 500MB image from WAN every time they boot?
| |
10:11 | You can give them a live usb stick to boot from
| |
10:11 | <anivair> n, I just wanted not to have to deal with the headache of having a real system there instead of a thin client.
| |
10:11 | that's probably safer
| |
10:11 | Ghid0rah has quit IRC | |
10:12 | CAN-o-SPAM has joined #ltsp | |
10:12 | <anivair> I'll still have to give them a machine that can boot from the usb stick, though, since I'm positive that most of them cannot on the machines they have at home.
| |
10:12 | <alkisg> Thin client == pulls OS from the server == too big to get it from the internet
| |
10:12 | So you can't have thin clients at home, unless you got a 100mbps connection at home
| |
10:13 | (or if you use a cd or a usb stick or... but that's not LTSP)
| |
10:13 | <anivair> so I can give them a machine that is running some sort of blank ubuntu system and use that to connect to the nx server
| |
10:13 | <dobber> with cd/usb you need to give the user a network configuration utility
| |
10:13 | <alkisg> A live usb stick from `usb creator` is all that they'd need - with a few MB left so that they install any proprietary drivers
| |
10:14 | <dobber> not everyone is using dhcp at home
| |
10:14 | <alkisg> That's actually easier than doing dhcp at home...
| |
10:14 | <johnny> silliness..
| |
10:16 | do you actually have to support their home computers?
| |
10:16 | just make them install your environment in a virtual machine ..
| |
10:16 | virtualbox is free and easy to setup
| |
10:16 | <alkisg> nx client is easier than vbox, though
| |
10:16 | (if they can get their computers to boot)
| |
10:17 | <dobber> what? and leave them to work on an open windows with no antivirus and thousands of keyloggers ?
| |
10:17 | <evil_root> so.. anyone here get ltsp 5 to work with gentoo?
| |
10:19 | <johnny> sometimes..
| |
10:19 | staffencasa has joined #ltsp | |
10:19 | <johnny> there is a howto on gentoo wiki.. i hope somebody didn't break it
| |
10:19 | F-GT has joined #ltsp | |
10:20 | <evil_root> yea i know, i got all the way to ldm loading and when i try to log in, the xsession starts and dies in the same second and goes back to the ldm login screen
| |
10:20 | <johnny> did you install gnome|kde or similiar on the server?
| |
10:21 | and install ldm-server package (i think)
| |
10:21 | <evil_root> xfce and fluxbox
| |
10:21 | <johnny> and then start xinetd ?
| |
10:21 | <anivair> okay, so my plan right now is to install nx on the local server and to give the people working from home old machines that have a blank linux install running on them. they can use those to get things running. that way I don't have to worry about too much crap and I can pull anything that they could use to break the system off the system.
| |
10:21 | <johnny> /me isn't sure how the package is now
| |
10:21 | <evil_root> yep, i can get x to run localy on the thin client just fine, cant get it to run threw the ssh tunnel
| |
10:21 | <johnny> yep what?
| |
10:22 | check to make sure the xinetd file has a reference to ldminfod or ldm-server
| |
10:22 | and that that file exists
| |
10:22 | also. make sure you can ssh <server> without warnings from the client
| |
10:23 | or without the you haven't connected before message
| |
10:23 | if you get that, you can't login
| |
10:23 | and you must run ltsp-update-sshkeys
| |
10:26 | <evil_root> the ssh works just fine, i can even ssh to the thin client and then from the thin client back to the server, i am checking on the xinetd file right now
| |
10:29 | anivair has quit IRC | |
10:30 | <johnny> does xfce put the requisite session file in the right place? i hope so
| |
10:30 | anyways.. you might wanna look at the ldm log on the thin client /var/log/ldm.log
| |
10:39 | dobber has quit IRC | |
10:42 | <evil_root> the ldm.log shows logged in successfully then it shows
| |
10:43 | Waiting for process 11479
| |
10:43 | Process 11479 exited with status 0
| |
10:43 | Established ssh session.
| |
10:43 | Executing rc files.
| |
10:43 | rc_files: /bin/sh /usr/share/ldm/ldm-script start
| |
10:43 | ldm_spawn: pid = 11494
| |
10:43 | Waiting for process 11494
| |
10:43 | Process 11494 exited with status 0
| |
10:43 | Beginning X session.
| |
10:43 | rc_files: /bin/sh /usr/share/ldm/ldm-script xsession
| |
10:43 | ldm_spawn: pid = 11503
| |
10:43 | Waiting for process 11503
| |
10:43 | Process 11503 exited with status 0
| |
10:43 | X session ended.
| |
10:43 | sorry about the spam everyone
| |
10:44 | <Gadi> evil_root: make sure your ssh server allows ssh -X connections
| |
10:44 | <evil_root> it does
| |
10:44 | <Gadi> also make sure that the server has an : /etc/X11/Xsession
| |
10:45 | file
| |
10:45 | Faithful has joined #ltsp | |
10:45 | gnunux has quit IRC | |
10:51 | <evil_root> i do have a /etc/X11/Sessions/Xsession file, i created a link for /etc/X11/Xsession for it just incase it needed to be in that specific location and still no luck
| |
10:52 | i get the login screen, can successfully authenticate, mouse turns into an x for about half a second then flashes back to the login screen
| |
10:59 | michiel_e has quit IRC | |
11:01 | <johnny> but do you have the desktop file for the session?
| |
11:02 | the one that links between the sessions
| |
11:02 | i forgot what it is called tho
| |
11:02 | knipwim, are you here?
| |
11:24 | Remaille has quit IRC | |
11:27 | <cliebow> mistik1 would say "OIhhh my eyes!"
| |
11:38 | evilx_ has quit IRC | |
12:00 | Ghid0rah has joined #ltsp | |
12:02 | lucascoala has joined #ltsp | |
12:11 | Remaille has joined #ltsp | |
12:11 | <lfaraone> alkisg: well, we have DHCP "next-server" in our router, but the DHCP server on our LTSP server is off.
| |
12:12 | <alkisg> lfaraone: sorry I must have missed the start, were we talking about something?
| |
12:13 | <lfaraone> alkisg: yesterday I mentioned I was getting "Hi, I upgraded to Ubuntu Lucid and now I get "Failed to connect to NBD server" when booting from the chroot. Any idea why?"
| |
12:13 | *yesterday I mentioned "
| |
12:13 | <alkisg> Ah, ok, it's been a long time since then, I forgot :)
| |
12:13 | <lfaraone> alkisg: np. problem persists to today though :(
| |
12:14 | <alkisg> lfaraone: try setting SERVER=<ip of your server> in lts.conf
| |
12:17 | <lfaraone> alkisg: done, nothing happened. (/var/lib/tftpboot/ltsp/i386/lts.conf, right?)
| |
12:17 | <alkisg> Right
| |
12:18 | What happens if you put "break=mount" right beside of "quiet splash" in /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default ?
| |
12:18 | Do you get a busybox shell?
| |
12:19 | <lfaraone> alkisg: We get a busybox shell anyway when the error happens, but hold on.
| |
12:19 | <alkisg> Yup, I'm just wondering if that tftp server is used and if pxelinux.cfg/default & lts.conf get transferred...
| |
12:21 | <lfaraone> alkisg: new behavior now, "spawning shell within the initramfs" and then to busybox.
| |
12:21 | <alkisg> OK. In that busybox shell try: nbd-client <server-ip> 2000 /dev/nbd0
| |
12:22 | <lfaraone> alkisg: connection refused.
| |
12:22 | <alkisg> And is the nbd-server running on the server?
| |
12:22 | sudo netstat -nap | grep 2000
| |
12:23 | <lfaraone> alkisg: aha! it's not running, and it "cannot parse config file" when I try to restart it.
| |
12:23 | /etc/nbd-server/ is empty.
| |
12:23 | /etc/nbd-server/ is empty.
| |
12:23 | <alkisg> That's not a problem
| |
12:23 | ...but it runs from inetd, so don't try to restart it with other ways
| |
12:24 | sudo invoke-rc.d openbsd-inetd restart
| |
12:24 | Anyway, try this: grep nbd /etc/inetd.conf
| |
12:24 | What do you get?
| |
12:25 | <lfaraone> alkisg: ports 2005-2007. see http://sprunge.us/JHMG for output
| |
12:25 | <alkisg> Uh, are those for real?
| |
12:26 | If you have multiple images, you need to pass the nbdport as a kernel parameter in pxelinux.cfg/default
| |
12:26 | What ubuntu version is that?
| |
12:26 | <lfaraone> alkisg: er, yeah. Lucid.
| |
12:26 | alkisg: and if possible, I'd like to switch to only one image.
| |
12:26 | <alkisg> What do you have in pxelinux.cfg/default
| |
12:26 | Sure, then just delete all entries and leave only the one you want, at port 2000
| |
12:27 | ...and see that pxelinux.cfg/default also has that port (2000)
| |
12:27 | But before you change it, what do you have now in pxelinux.cfg/default?
| |
12:27 | <lfaraone> alkisg: http://sprunge.us/PGSL
| |
12:28 | <alkisg> 2003 stream tcp nowait nobody /usr/sbin/tcpd /usr/sbin/nbdrootd /opt/newtsp/images/i386_lucid.img
| |
12:28 | so does that file exist?
| |
12:28 | <lfaraone> alkisg: none of them exist, I archived our old images.
| |
12:29 | alkisg: I added a new line 2000 stream tcp nowait nobody /usr/sbin/tcpd /usr/sbin/nbdrootd /opt/ltsp/images/i386.img
| |
12:29 | <alkisg> OK, that's why it wasn't working then
| |
12:29 | Also change the port in pxelinux.cfg/default
| |
12:29 | <Gadi> ltsp-update-image should care for the ports
| |
12:30 | <lfaraone> It should, but I probably broke something somewhere.
| |
12:31 | <alkisg> Gadi, he deleted the nbd images so...
| |
12:31 | <johnny> i had a problem where it would keep choosing the wrong port
| |
12:31 | <alkisg> johnny: in lucid?
| |
12:31 | <johnny> but my ubuntu reinstall solved that problem..
| |
12:31 | <alkisg> (post alpha2?)
| |
12:31 | <johnny> no.. karmic
| |
12:31 | <lfaraone> alkisg: do I need to reboot inetd?
| |
12:31 | <johnny> and jaunty
| |
12:31 | <Gadi> alkisg: yeah, but it takes the info from the chroot itself
| |
12:31 | <alkisg> OK, new code now, Gadi changed a lot of stuff there
| |
12:31 | Gadi, it didn't boot because the nbd image for port 2003 was deleted - so no problem there...
| |
12:32 | <Gadi> right
| |
12:32 | <alkisg> lfaraone: yes
| |
12:32 | sudo invoke-rc.d openbsd-inetd restart
| |
12:32 | <lfaraone> Ah, that expalins why it wasn't working when I just tried it ;)
| |
12:32 | <Gadi> he should most likely nuke all the image lines in inetd.conf and run ltsp-update-image
| |
12:32 | <lfaraone> Aha, we're working now. Thanks alkisg, Gadi.
| |
12:33 | <alkisg> Gadi, I told him to change inetd.conf & pxelinux.cfg/default manually, to save him some squashfs time :)
| |
12:33 | lfaraone: yw
| |
12:36 | Gadi, I'm running a bash shell script that waits for tcp connections from /etc/xdg/autostart, and it doesn't die at logoff. Any ideas where I should look to make it die with the session?
| |
12:36 | (kill <pid> kills it normally, it doesn't have any weird hooks...)
| |
12:37 | <Gadi> does it dissociate from the controlling terminal?
| |
12:37 | <alkisg> Yes, it uses a socket for stdin/stdout
| |
12:37 | (i.e. fd/0 is a tcp socket)
| |
12:37 | <Gadi> can you paste the script?
| |
12:37 | <alkisg> Sure
| |
12:38 | http://bazaar.launchpad.net/~sch-devs/sch-scripts/trunk/annotate/head%3A/sbin/sch-client
| |
12:40 | <Gadi> why do you connect it to stdin?
| |
12:41 | that's so you can continually read from it?
| |
12:42 | <alkisg> So that `sh` reads the commands from the server
| |
12:42 | <Gadi> you may want to trap SIGHUP and some other signals
| |
12:42 | <alkisg> OK, can you give me a short version of the theory?
| |
12:42 | So gnome doesn't send a kill signal to the processes that they don't have a tty as stdin?
| |
12:43 | <Gadi> maybe the session manager simply sends a sighup and bash says - I can't sighup, I haven't finished reading from the socket
| |
12:43 | <alkisg> Thanks - let me try that...
| |
12:44 | (when I do "kill job", what signal do I send?)
| |
12:45 | Ah, SIGTERM.
| |
12:46 | Nah, it terminates properly with SIGHUP
| |
12:47 | <Gadi> same with SIGKILL and SIGUSR I assume?
| |
12:47 | SIGUSR1, that is
| |
12:49 | <alkisg> SIGKILL kills it, let me try SIGUSR...
| |
12:50 | Yup, it died with kill -s USR1 <pid>
| |
12:50 | mischko has joined #ltsp | |
12:51 | <Gadi> I guess nobody's telling it to die
| |
12:51 | :)
| |
12:51 | <alkisg> (btw evolution-data-server also doesn't die, and it has redirected fd/0 and 1 to /dev/null, so you might be right there)
| |
12:59 | evil_root has quit IRC | |
13:08 | <alkisg> Bah, I tried starting from /etc/xdg/autostart a simple script with a `sleep 12345` on it, and it also doesn't die at logoff.
| |
13:09 | * alkisg wonders if he could run it from an ldm xsession script instead... | |
13:12 | Remaille has quit IRC | |
13:24 | Barbosa has joined #ltsp | |
13:31 | <Barbosa> hi for all, is there any doc about serial mouse bug in ubuntu karmic ?
| |
13:32 | <sbalneav> http://ubuntuforums.org/showthread.php?t=1331499
| |
13:34 | Barbosa: ^^
| |
13:35 | <Barbosa> sbalneav: Hi, this problem is in thin client ltsp5 serial mouse
| |
13:36 | Ghid0rah has quit IRC | |
13:39 | <sbalneav> Barbosa: So, you'd have to follow the above instructions I posted in the ltsp chroot.
| |
13:40 | CAN-o-SPAM has quit IRC | |
13:54 | bobby_C has joined #ltsp | |
14:06 | vagrantc has joined #ltsp | |
14:10 | CAN-o-SPAM has joined #ltsp | |
14:30 | runout has joined #ltsp | |
14:39 | <alkisg> Hmmm it looks like I need to register the client as part of the session with a dbus call: http://people.gnome.org/~mccann/gnome-session/docs/gnome-session.html#org.gnome.SessionManager.RegisterClient
| |
14:41 | <lfaraone> alkisg: btw, have you encontered "cannot allocate memory" issues when accessing files on USB with LTSP and amd64? I haven't had a chance to test this in Lucid, but it wasn't working in Karmic.
| |
14:41 | * alkisg never uses amd64, so has no testing experience... | |
14:41 | <alkisg> I think ltspfsd doesn't work on amd64 though
| |
14:42 | Lns has joined #ltsp | |
14:42 | <lfaraone> alkisg: yeah, from a cursory reading of the build logs it looks like problems with implicit pointer casts.
| |
14:43 | <alkisg> lfaraone: sbalneav is the person to talk about that, I also don't use ltspfsd :)
| |
14:57 | highvolt1ge is now known as highvoltage | |
14:57 | highvoltage has joined #ltsp | |
14:58 | pmatulis has quit IRC | |
15:02 | mischko has quit IRC | |
15:03 | cliebow has quit IRC | |
15:11 | <sbalneav> lfaraone: Are you talking amd64 clients?
| |
15:11 | <lfaraone> sbalneav: Yes. There's a bug somewhere...
| |
15:11 | <sbalneav> Yes there is.
| |
15:11 | <lfaraone> sbalneav: https://bugs.launchpad.net/bugs/415952
| |
15:12 | <sbalneav> I'm aware
| |
15:12 | <lfaraone> sbalneav: and a friend of mine tried to hack out a patch, https://bugs.launchpad.net/bugs/415952
| |
15:12 | <sbalneav> But I'm somewhat hampered by the fact I don't have a 64 bit client.
| |
15:12 | <lfaraone> * http://sascha.silbe.org/tmp/ltspfs-xdr-type-fix.patch
| |
15:12 | <vagrantc> sbalneav: might be able to use qemu to emulate a 64-bit client
| |
15:13 | <lfaraone> I think if we can clear out the build warnings we'll be fine; I suspect that;'s the cause of the problem
| |
15:13 | (what is indicated in the warnings, that is)
| |
15:14 | <sbalneav> Well, just eliminating the build warnings isn't the main problem.
| |
15:15 | Here's the problem.
| |
15:16 | a struct stat uses types like gid_t and uid_t
| |
15:17 | Originally, they were supposed to be known static types. They're typdef'd to things like int, etc.
| |
15:17 | int changes depending on if you're 64 bit or not.
| |
15:17 | but currently, ltspfs doesn't send any information along to indicate int size on the client or server.
| |
15:18 | Since it's TECHNICALLY possible to have an 32 bit server, and a 64 bit client, I can
| |
15:18 | <vagrantc> uh-oh.
| |
15:18 | <sbalneav> can't just force everything 64 bit.
| |
15:18 | <lfaraone> sbalneav: sure, but you'd be hard-pressed to create the chroot on the server, no?
| |
15:19 | <sbalneav> No
| |
15:19 | create it on another server and just copy the .img across.
| |
15:19 | <vagrantc> can already create armel chroots on an i386 without a hitch
| |
15:19 | amd64 isn't far behind.
| |
15:19 | <lfaraone> mkk.
| |
15:20 | <sbalneav> so, I need to add some smarts to ltspfs to detect int size, and send that along
| |
15:20 | either that, or skip the whole silly thing, and just send statbuf information across as text.
| |
15:20 | <lfaraone> sbalneav: is there a downside for going the text route?
| |
15:20 | <vagrantc> both break backwards compatibility?
| |
15:21 | <sbalneav> vagrantc: Well, of course :)
| |
15:21 | <lfaraone> vagrantc: aren't there "version" flags implemented in ltspfs, so you can make protocol changes without breaking compatibility?
| |
15:21 | <vagrantc> don't know of any such thing
| |
15:21 | <sbalneav> No.
| |
15:22 | <vagrantc> thankfully, we haven't had to do any major changes for quite some time, and that's been rather nice.
| |
15:22 | <lfaraone> Well, if backward compatiblity was a concern that would be the way to go.
| |
15:22 | (in fact, could we add a verison number going forward, and assume version-number-less connections are on the old protocol?)
| |
15:23 | <vagrantc> simply sending the version number would likely break the old protocol, though, no?
| |
15:23 | * vagrantc leaves that to clever people to figure | |
15:24 | <sbalneav> lfaraone: Sure. All it takes is time to code it all up.
| |
15:24 | <lfaraone> Unless we put the versioned protocol on a new port, and listen on both.
| |
15:24 | sbalneav: I'd love to help, but C is my weak spot :(
| |
15:25 | <sbalneav> So then what's with all the "we" stuff then? :)
| |
15:25 | <lfaraone> sbalneav: easier to write than "unless the versioned protocol is placed on a new port" in the passive voice.
| |
15:26 | <sbalneav> Until such a time as I get a good solution that a) works b) doesn't break the universe and c) I have time to code up, I'd say the solution is: 32 bit clients.
| |
15:26 | since there's no problem with 64 bit server and 32 bit clients.
| |
15:27 | <lfaraone> sbalneav: fair enough.
| |
15:27 | <Lns> sbalneav: what if everyone here pitches in some cash for you/whoever to make the fix? :) Would that help with your 'time to code up' situation at all? Just talking hypothetically of course ;)
| |
15:27 | <sbalneav> No
| |
15:27 | Cash I have.
| |
15:27 | <Lns> can I have some then? ;)
| |
15:27 | Barbosa has quit IRC | |
15:28 | <sbalneav> And I've never wanted to be paid for this work.
| |
15:28 | It make it stop being "a fun hobby" and turn it into "a job".
| |
15:28 | <Lns> I hear you
| |
15:28 | money usually just messes things up anyway.
| |
15:29 | <sbalneav> LTSP I will keep having fun with after I have retired from "the job" :)
| |
15:29 | <lfaraone> sbalneav: you have Cash? send some my way :P
| |
15:29 | * lfaraone is still looking for said "job" | |
15:31 | <Lns> sbalneav: you rock. Period. =)
| |
15:31 | <lfaraone> sbalneav: thanks for the explanation.
| |
15:37 | <sbalneav> NP
| |
15:39 | Lns has quit IRC | |
15:39 | Barbosa has joined #ltsp | |
15:48 | <sbalneav> One "stupid" idea that comes to mind is to simply have ltspfsd have a flag ("-64"), and we do some hacky test on the client like uname | grep _64, and then start ltspfsd -64 based on that.
| |
15:48 | * CAN-o-SPAM chimes in | |
15:48 | <CAN-o-SPAM> sbalneav: no idea is stupid ;)
| |
15:53 | runout has quit IRC | |
15:54 | runout has joined #ltsp | |
15:57 | CAN-o-SPAM has quit IRC | |
16:02 | Lns has joined #ltsp | |
16:06 | shamino has quit IRC | |
16:06 | <Gadi> sbalneav: wouldn't the 64-bit client simply have a 64-bit package installed?
| |
16:07 | lucascoala has quit IRC | |
16:21 | Gadi has left #ltsp | |
16:33 | mgariepy has quit IRC | |
17:14 | johnny has left #ltsp | |
17:19 | johnny has joined #ltsp | |
17:33 | Barbosa has quit IRC | |
17:45 | vagrantc has quit IRC | |
17:47 | * Lns is finally kinda getting the hang of libvirt/kvm/qemu/virsh/blah | |
17:49 | alkisg has quit IRC | |
18:15 | Lns has quit IRC | |
18:38 | Faithful has quit IRC | |
18:40 | staffencasa has quit IRC | |
18:45 | bobby_C has quit IRC | |
19:17 | runout has left #ltsp | |
19:41 | wsuetholz has quit IRC | |
20:14 | pmatulis has joined #ltsp | |
20:19 | wsuetholz has joined #ltsp | |
20:21 | lucascoala has joined #ltsp | |
21:10 | DeathLx has joined #ltsp | |
21:10 | <DeathLx> hi
| |
21:12 | jammcq has joined #ltsp | |
21:13 | <jammcq> hello all
| |
21:13 | <DeathLx> some help me xD
| |
21:17 | pmatulis has quit IRC | |
21:26 | johnny has left #ltsp | |
21:31 | <DeathLx> how up the service of the ltsp in fedora
| |
21:43 | <sbalneav> DeathLx: Sorry?
| |
21:43 | How do you install LTSP in fedora, you mean?
| |
21:44 | <DeathLx> i install but, i don't know start
| |
21:45 | i use K12 ltsp
| |
21:46 | <sbalneav> https://fedorahosted.org/k12linux/wiki/UserDocumentation
| |
21:46 | <DeathLx> suport in spanish ?
| |
21:46 | <sbalneav> https://fedorahosted.org/k12linux/wiki/InstallGuide
| |
21:47 | I don't know if fedora has any documentation translated to spanish
| |
21:47 | <DeathLx> me think client is not conect :S
| |
21:48 | i installed correctly all
| |
21:49 | <sbalneav> Have you disabled the firewall? The docs I pointed you to seem to indicate that the firewall on fedora should be disabled.
| |
21:50 | <DeathLx> yes my firewall is diseable
| |
21:50 | <sbalneav> What happens when you boot? Does the kernel download?
| |
21:51 | To the thin client, that is.
| |
21:51 | <DeathLx> the thin client boot from mac address
| |
21:52 | <sbalneav> So, do you get a login prompt?
| |
21:54 | Does the thin client get a graphical login prompt?
| |
21:54 | Egyptian[Home] has quit IRC | |
21:54 | <DeathLx> not enter to graphical login
| |
21:55 | not boot
| |
21:56 | <sbalneav> So, it ISN'T downloading the kernel, then?
| |
21:57 | <DeathLx> kernel of the ltsp is install
| |
21:57 | <sbalneav> When you turn on the thin client, does the kernel load on to the client? Yes or no?
| |
21:57 | <DeathLx> not load the kernel
| |
21:58 | johnny has joined #ltsp | |
21:58 | <sbalneav> ok
| |
21:58 | So, then you have some problem with your networking.
| |
21:58 | Is dhcp running?
| |
21:58 | <DeathLx> dhcp is running
| |
21:59 | <sbalneav> Does the thin client get an IP address?
| |
22:00 | <DeathLx> no the thin client get an ip address because boot in mac
| |
22:01 | <jammcq> sbalneav: Scotty !!!!!!!!!!!!!!!!!!!!!!!!!
| |
22:01 | <sbalneav> So, the thin client gets an ip addrss, but the kernel doesn't download? If so, then there's something wrong with yout TFTP server.
| |
22:01 | jammcq!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
22:02 | <DeathLx> how donwload the kernel
| |
22:02 | <jammcq> sbalneav: have you seen this: http://www.ibm.com/developerworks/linux/library/l-solarnetone/
| |
22:03 | <sbalneav> Have you done an ltsp-update-kernels?
| |
22:03 | jammcq: heh, cool!
| |
22:04 | <jammcq> sbalneav: just in case you were wondering if this stuff was all worth it
| |
22:04 | <DeathLx> the kenel is 5.1.95
| |
22:04 | <sbalneav> DeathLx: Have you run the command: ltsp-update-kernels?
| |
22:05 | <DeathLx> mmm no
| |
22:07 | <sbalneav> OK, well, I'd start by doing that, then trying to boot the thin client again.
| |
22:08 | <DeathLx> ok, but what is the command :S
| |
22:08 | <jammcq> !docs
| |
22:08 | <ltspbot> jammcq: "docs" :: For the most current documentation, see https://sourceforge.net/apps/mediawiki/ltsp/index.php?title=Ltsp_LtspDocumentationUpstream
| |
22:09 | Egyptian[Home] has joined #ltsp | |
22:10 | <sbalneav> the command is:
| |
22:11 | ltsp-update-kernels
| |
22:11 | then power off the thin client
| |
22:11 | and turn it on again.
| |
22:14 | ogra_cmpc has quit IRC | |
22:15 | <DeathLx> PXE-E53: no boot file name received
| |
22:15 | <jammcq> sounds like another dhcp server on the network
| |
22:17 | <johnny> /me snuck it in there
| |
22:18 | /me pokes jammcq
| |
22:19 | <sbalneav> DeathLx: Do you have another dhcp server on the network?
| |
22:20 | <DeathLx> no
| |
22:20 | <jammcq> hey johnny
| |
22:24 | <sbalneav> DeathLx: Did you do step 8 on this page? https://fedorahosted.org/k12linux/wiki/InstallGuide
| |
22:27 | <DeathLx> all are active
| |
22:27 | ogra_cmpc has joined #ltsp | |
22:28 | <sbalneav> DeathLx: Well, all I can offer is to send an email to the K12Linux mailing list.
| |
22:29 | <DeathLx> :s ok
| |
22:31 | how active the ltsp-dhcp services?
| |
22:36 | slidesinger has joined #ltsp | |
22:40 | <sbalneav> There's also our docs, which have a section on fedora
| |
22:40 | !docs
| |
22:40 | <ltspbot> sbalneav: "docs" :: For the most current documentation, see https://sourceforge.net/apps/mediawiki/ltsp/index.php?title=Ltsp_LtspDocumentationUpstream
| |
22:53 | vagrantc has joined #ltsp | |
23:53 | shamino has joined #ltsp | |