IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 2 December 2010 (all times are UTC)
| 00:32 | Trixboxer has joined #ltsp | |
| 00:33 | zz_evil_root is now known as evil_root | |
| 00:33 | dsnetworx has quit IRC | |
| 01:21 | pthsWork has joined #ltsp | |
| 01:34 | sweetpi has quit IRC | |
| 01:51 | vvinet has quit IRC | |
| 02:01 | gnunux has joined #ltsp | |
| 02:02 | daya has quit IRC | |
| 02:04 | vvinet has joined #ltsp | |
| 02:08 | daya has joined #ltsp | |
| 02:15 | cyberorg has quit IRC | |
| 02:16 | cyberorg has joined #ltsp | |
| 02:25 | daya has quit IRC | |
| 02:42 | gorkhaan has joined #ltsp | |
| 02:42 | pthsWork has quit IRC | |
| 02:56 | pthsWork has joined #ltsp | |
| 02:56 | evil_root is now known as zz_evil_root | |
| 03:13 | cyberorg has quit IRC | |
| 03:17 | cyberorg has joined #ltsp | |
| 03:33 | dlezcano has quit IRC | |
| 04:08 | cyberorg has quit IRC | |
| 04:09 | cyberorg has joined #ltsp | |
| 04:27 | DSNetworx has joined #ltsp | |
| 05:00 | ogra_ac_ is now known as ogra_ac | |
| 05:00 | alkisg has quit IRC | |
| 05:44 | DSNetworx has quit IRC | |
| 05:51 | yanu_ has quit IRC | |
| 05:54 | johnny has left #ltsp | |
| 05:55 | johnny has joined #ltsp | |
| 05:56 | jimjimovich has joined #ltsp | |
| 06:14 | <jimjimovich> anyone know a good way to send a message to all LTSP users (for example, that you're going to restart the machine in 15 min)?
| |
| 06:21 | <muppis> I don't rememer actual application name, but using same as is used when leaving message to user who has locked the screen.
| |
| 06:21 | And looping thru logged users.
| |
| 06:23 | dlezcano has joined #ltsp | |
| 06:28 | gorkhaan has quit IRC | |
| 06:36 | <jimjimovich> looks like notify-send might do the job
| |
| 06:38 | <muppis> Sounds familiar. Rught now I don't have system in hand to test.
| |
| 06:38 | Right.. My fingers and N900..
| |
| 06:39 | <jimjimovich> :)
| |
| 06:41 | <muppis> Next weekend I'm doing a lot of testing with ltsp, wlan and xbmc.
| |
| 06:48 | Plan to get ltsp working with internal wlan, even if I had to use internal hdd for booting.
| |
| 06:48 | dlezcano has quit IRC | |
| 06:55 | <jimjimovich> muppis: you think that wlan is fast enough for that?
| |
| 06:58 | <muppis> jimjimovich, propably not. Just testing it for demo purposes.
| |
| 07:02 | Da-Geek has joined #ltsp | |
| 07:03 | <muppis> I've seen used dd-wrt based Buffalo AirStations as converter to boot in ltsp. Want to get rid from that even it won't be pure diskless system.
| |
| 07:08 | Axachi has joined #ltsp | |
| 07:08 | mgariepy has joined #ltsp | |
| 07:14 | MorningSon has joined #ltsp | |
| 07:17 | gorkhaan has joined #ltsp | |
| 07:18 | Axachi has quit IRC | |
| 07:22 | bigd has joined #ltsp | |
| 07:22 | <bigd> hey together
| |
| 07:24 | i'm currently facing a weird problem
| |
| 07:24 | maybe someone has a suggestion on a posible solution
| |
| 07:25 | the thinclients all hang/stuck if the loose connection to the ltsp-server, like while server is rebooting
| |
| 07:26 | is there a posibility to let them shutdown if they loose connection to the server instead of hanging?
| |
| 07:26 | <mgariepy> bigd, are you using nbd-proxy ?
| |
| 07:27 | <bigd> yes
| |
| 07:27 | <mgariepy> bigd, when the server comes back they should reconnect
| |
| 07:28 | <bigd> only if the server is back in less then 1 min
| |
| 07:28 | if it takes longer the freez
| |
| 07:29 | they
| |
| 07:29 | <mgariepy> to shut them down. you can use ctrl+ sysrq + o
| |
| 07:30 | <bigd> sysrq ?
| |
| 07:30 | <mgariepy> print screen.
| |
| 07:30 | on my kb, i have print screen/sysrq
| |
| 07:31 | <bigd> ah, ok ... german kb ;-)
| |
| 07:31 | but the clients have no kb at all
| |
| 07:32 | they have touchscreens
| |
| 07:32 | <mgariepy> hmm, what version of linux are you using ?
| |
| 07:32 | ubuntu lucid ?
| |
| 07:32 | <bigd> Xubuntu 10.04.1
| |
| 07:33 | im currenlty also trying Xu 10.10
| |
| 07:33 | same problem
| |
| 07:34 | with our current production setup with Ubuntu 8.04 the clients at least stay reachable thru ssh
| |
| 07:34 | so i can shut them down manually
| |
| 07:35 | the clients show following errors:
| |
| 07:35 | <mgariepy> did you tried to disable nbd-proxy ?
| |
| 07:36 | <bigd> nbd0: Attempted send on closed socket
| |
| 07:36 | yea, made no difference
| |
| 07:36 | my colleg tryed that
| |
| 07:36 | <mgariepy> ok
| |
| 07:37 | <bigd> so im not 100% sure what he did exactly
| |
| 07:37 | the rest of the errors shown
| |
| 07:37 | end_request: I/O error, dev nbd0, sector
| |
| 07:37 | SQUASHFS error: squashfs_read_data failed to read block
| |
| 07:37 | SQUASHFS error: Unable to read metadata cache entry
| |
| 07:38 | SQUASHFS error: Unable to read directory block
| |
| 07:38 | i have already tried the mentioned solution in Bug #457702
| |
| 07:38 | but with no luck
| |
| 07:42 | <mgariepy> hmm i have no idea right now
| |
| 07:43 | the bug #457702 is not quite related to that, it was about a problem when trying to shutdown the thin client via ldm.
| |
| 07:44 | <bigd> i know, but as the error was neraly the same
| |
| 07:45 | if at least SSH would stay reachable i would be happy
| |
| 07:46 | i guess it has something to do with nbd and that the root-filesystem gets lost
| |
| 07:47 | <mgariepy> yeah, with nbd-proxy, when the filesystem gets disconnected, it puts it in IOWait, at lease it's supposed
| |
| 07:48 | at least**
| |
| 07:48 | <bigd> yeah, and it seems that doesn't wait
| |
| 07:49 | +nbd-proxy
| |
| 07:51 | can some processes "bypass" that IOwait?
| |
| 07:51 | would fit to that behavier
| |
| 07:52 | <mgariepy> hmm maybe if all the binaries are already in cache.
| |
| 07:53 | but i'm not sure.
| |
| 07:53 | loather-work has quit IRC | |
| 07:54 | <bigd> ah, the clients continues throwing that error above ... when i try ssh they chrush completly
| |
| 07:54 | just in case that makes any difference
| |
| 07:55 | sorry ... "when i to connect via ssh"
| |
| 07:57 | + try
| |
| 08:00 | you know how the end of the IOwait is triggerd?
| |
| 08:02 | maybe the real problem is that nbd-proxy doesn't "stops" waiting
| |
| 08:03 | jimjimovich has left #ltsp | |
| 08:20 | <bigd> test
| |
| 08:21 | <mgariepy> do you see the thin client try to reconnect to the ltsp-server when it comes back ?
| |
| 08:22 | Gadi has joined #ltsp | |
| 08:22 | alkisg has joined #ltsp | |
| 08:23 | <Gadi> Mip5: sorry about yesterday - didn't mean to leave you hanging. Got drawn into meetings and before I knew it, it was time to get home for the kids
| |
| 08:24 | <bigd> were do you mean? as they usually hung completly i didn't cared about that
| |
| 08:25 | the user or in the logs?
| |
| 08:26 | <mgariepy> bigd, the nbd connexion, doest the thin client try to reconnect it when the server gets back ?
| |
| 08:26 | komunista has joined #ltsp | |
| 08:27 | <bigd> nope, or at least there is nothing that lets me see that
| |
| 08:27 | they continue with the "SQUASHFS"-Errors
| |
| 08:28 | <mgariepy> you can see it on the nbd server.
| |
| 08:29 | <bigd> in the logs or by a command?
| |
| 08:29 | <mgariepy> in the log
| |
| 08:31 | Dec 2 02:48:23 ltsp-server nbd_server[858]: connect from 128.0.0.7, assigned file is /opt/ltsp/images/i386.img
| |
| 08:31 | something like that in syslog.
| |
| 08:31 | <bigd> wait, ill try it again to have §fresh" logs
| |
| 08:39 | cool ... new strange behavier ... now it shows the login-screen at the clients
| |
| 08:39 | just seconds after the server is gone
| |
| 08:39 | but all text i cryted
| |
| 08:39 | is
| |
| 08:40 | strange signs what ever
| |
| 08:40 | when the server was back i tried to login
| |
| 08:40 | syslog says nothing abot nbd
| |
| 08:41 | just "ldminfod[..]: connect from ... (...)
| |
| 08:42 | but instead of a desktop the clients crushes and throws again thoses errors
| |
| 08:42 | squashfs-erro etc
| |
| 08:44 | rjune has joined #ltsp | |
| 08:44 | <bigd> and now after trying "ctlr+alt+del" the clients hangs completly again, but like said befor, in production the wont have a kb anyway
| |
| 08:44 | they
| |
| 08:44 | [GuS] has joined #ltsp | |
| 08:44 | [GuS] has joined #ltsp | |
| 08:46 | dlezcano has joined #ltsp | |
| 08:51 | <bigd> so ... the server seems to never recieve an nbd reconnect
| |
| 08:52 | <mgariepy> hmm i really have no idea now
| |
| 08:53 | <Gadi> so, impromptu poll - anyone have any sound issues? I found a race condition and just want to confirm if others have been bitten by it
| |
| 08:54 | <mgariepy> Gadi, dgross yesterday had some.
| |
| 08:54 | <Gadi> so, I have some thin clients that start pulseaudio before the soundcard is ready
| |
| 08:54 | <alkisg> bigd: this should work for most of the cases: while nbd-client -c /dev/nbd0; do reboot --help; sleep 10; done; reboot -fp
| |
| 08:54 | Put it in an rc script on the client
| |
| 08:54 | <ogra_ac> Gadi, get slower clients
| |
| 08:55 | <Gadi> mgariepy: so, you see pulseaudio happily running, but apps get stream errors when trying to create a stream
| |
| 08:55 | mgariepy: I can fix it by starting pulseaudio out of udev
| |
| 08:55 | anyone take issue with that?
| |
| 08:56 | <mgariepy> dgross yesterday had an issue with soundcard not detecting headphone.
| |
| 08:56 | <alkisg> Gadi: doesn't pulseaudio start anyway in recent distros? I mean, wouldn't it be enough to patch its configuration file?
| |
| 08:56 | <ogra_ac> alkisg, in the desktop session
| |
| 08:57 | <Gadi> alkisg: well, we run it as a system process
| |
| 08:57 | <ogra_ac> pulse doesnt start on a system level in any distro
| |
| 08:57 | <alkisg> Ah, ok
| |
| 08:57 | <Gadi> not sure if we couldn't run it as a desktop session process?
| |
| 08:57 | <alkisg> Why is it a good thing ... yeah that? ^^^
| |
| 08:57 | <ogra_ac> from ldm ?
| |
| 08:57 | <alkisg> What does rdesktop need? pulse or alsa?
| |
| 08:57 | <ogra_ac> you need to run the daemon on the client
| |
| 08:58 | but indeed there is no reason to not start it from ldm (we did that in the past btw)
| |
| 08:58 | <Gadi> we did?
| |
| 08:58 | when?
| |
| 08:59 | I thought we always ran it from init
| |
| 08:59 | <ogra_ac> nah
| |
| 08:59 | when i implemented sound first with esd it was run from ldm ...
| |
| 08:59 | <Gadi> back when ldm was python?
| |
| 08:59 | <ogra_ac> even hardcoded
| |
| 08:59 | <Gadi> yuck
| |
| 09:00 | * Gadi shakes head at ogra_ac | |
| 09:00 | <ogra_ac> then pulse came around and we just added an if condition to fire off pulse
| |
| 09:00 | <Gadi> you were so young then
| |
| 09:00 | <ogra_ac> depending on what binary was there
| |
| 09:00 | yeah
| |
| 09:00 | less wrinkled
| |
| 09:00 | i think its valuable to have it run from pulse
| |
| 09:00 | err
| |
| 09:00 | ldm
| |
| 09:01 | <Gadi> how come?
| |
| 09:01 | <ogra_ac> but you should do it at ldm startup
| |
| 09:01 | not at login time
| |
| 09:01 | * Gadi thinks a sound daemon should run when/if there a soundcard comes up | |
| 09:01 | <ogra_ac> that way you can have a11y support (like the drum sound in ubuntu when gdm comes up)
| |
| 09:01 | <Gadi> wait - that's a plus?
| |
| 09:01 | :)
| |
| 09:01 | <ogra_ac> why would you need the soundcard before any login manager is there ?
| |
| 09:02 | ask visually impaired people ;)
| |
| 09:02 | <alkisg> bigd: this one's better, as it uses the magic sysrq file so it doesn't call `reboot --help` all the time to force keeping it in RAM:
| |
| 09:02 | while nbd-client -c /dev/nbd0; do sleep 60; done; echo b > /proc/sysrq-trigger
| |
| 09:02 | <Gadi> you need a soundcard to play a sound
| |
| 09:02 | <ogra_ac> yes, but when do you play sounds ?
| |
| 09:02 | <Gadi> when you have a soundcard
| |
| 09:02 | <ogra_ac> only if a userspace app does it
| |
| 09:02 | <Gadi> :P
| |
| 09:02 | userspace apps can't play a sound when the soundcard's not there
| |
| 09:03 | <ogra_ac> indeed
| |
| 09:03 | nor when the soundserver isnt there
| |
| 09:03 | <Gadi> exactly
| |
| 09:03 | but the apps shouldnt start the sound server
| |
| 09:03 | the card should
| |
| 09:03 | if the sound server is ready, the app can use it
| |
| 09:03 | <alkisg> Gadi: but pulse isn't needed for shell or rdesktop SCREENs, so it makes sense to start it from LDM, right?
| |
| 09:04 | <bigd> aaaahm ... what does "nbd_server: Disconnect request received." mean? the client is running and seems working without problems ... or is that normal?
| |
| 09:04 | <ogra_ac> write an upstart job then ;)
| |
| 09:04 | <Gadi> why not udev?
| |
| 09:04 | <ogra_ac> becaue its userspace ?
| |
| 09:04 | <Gadi> what is? the daemon?
| |
| 09:04 | <ogra_ac> yep
| |
| 09:05 | i would use udev for devices but upstart/init for other stuff
| |
| 09:05 | or even go up one more level and use ldm for it
| |
| 09:05 | given that you need a custom startup anyway
| |
| 09:06 | and can never use the distro defaults
| |
| 09:06 | <Gadi> so, just move it into an ldm/rc.d script
| |
| 09:06 | <ogra_ac> yeah
| |
| 09:07 | <Gadi> hmm...
| |
| 09:07 | <ogra_ac> that way you also make sure it doesnt get in your way with rdesktop
| |
| 09:07 | you could even run it with user permissions
| |
| 09:07 | since you have uid and gui
| |
| 09:07 | which adds safety
| |
| 09:07 | <Gadi> as opposed to --system
| |
| 09:08 | <ogra_ac> right
| |
| 09:08 | <bigd> ok ... leaving the original problem unsolved ... can i tragger somehow that all clients shutdown when the server shutsdown/reboots?
| |
| 09:08 | <Gadi> which comes with all the skull and crossbones warnings
| |
| 09:08 | <ogra_ac> s/gui/gid
| |
| 09:08 | <bigd> or (ideal way) let the clients shutdown if the loose connection to the server?
| |
| 09:09 | *trigger
| |
| 09:09 | <Gadi> ogra_ac: the only thing that doesn't solve yet is whether the sound card is ready
| |
| 09:09 | but, I suppose we can background it and have it wait for the soundcard to come up
| |
| 09:09 | or something
| |
| 09:09 | <ogra_ac> well, if X comes up the card should really better be ready
| |
| 09:10 | <Gadi> how do you know?
| |
| 09:10 | <ogra_ac> you can check for it in /dev from your rc script
| |
| 09:10 | <Gadi> right
| |
| 09:10 | <ogra_ac> and background a loop until its up
| |
| 09:11 | <Gadi> unless there's no card
| |
| 09:11 | and then we have a looping process
| |
| 09:11 | :P
| |
| 09:11 | I suppose best to not have a loop
| |
| 09:11 | and just a check
| |
| 09:11 | <ogra_ac> k
| |
| 09:11 | <Gadi> or maybe have it start upon login
| |
| 09:11 | that way, you can logout/login
| |
| 09:11 | and fix it
| |
| 09:11 | <ogra_ac> then you lose the opportunity for a11y support
| |
| 09:12 | <Gadi> perchance
| |
| 09:12 | hmm...
| |
| 09:12 | fine
| |
| 09:12 | on init
| |
| 09:12 | <ogra_ac> or do both :)
| |
| 09:13 | if it doesnt come up on init, firs it up on login
| |
| 09:13 | *fire
| |
| 09:13 | <Gadi> hmm... and I suppose this way, pulse will die with X, so it will relaunch on logout/login
| |
| 09:14 | <ogra_ac> yep
| |
| 09:14 | introduce a new LDM_A11Y var ;)
| |
| 09:14 | <alkisg> bigd: I wrote you a solution above, but you ignored it.
| |
| 09:14 | <ogra_ac> that would start it on ldm startup and play a sound
| |
| 09:15 | if its not set, just run it at login time with user credentials
| |
| 09:15 | <Gadi> too complicated
| |
| 09:15 | let's just start it on init
| |
| 09:15 | if you didn't get your sound, you can get it next time you login
| |
| 09:15 | :)
| |
| 09:15 | <bigd> oh, sorry, just saw it
| |
| 09:16 | alkisg, in which rc-folder i put that?
| |
| 09:16 | <alkisg> !lts.conf
| |
| 09:16 | <ltspbot`> alkisg: "lts.conf" :: http://manpages.ubuntu.com/lts.conf
| |
| 09:16 | <alkisg> See RC_FILES there ^^^
| |
| 09:17 | E.g. RCFILE_01
| |
| 09:18 | <bigd> thanks, ill give it a try
| |
| 09:19 | <alkisg> You're welcome. I'm not sure if a different device is needed because of nbd-proxy - I don't use nbd-proxy.
| |
| 09:19 | Try it from a SCREEN_02 first to verify that it works for you
| |
| 09:19 | !SCREEN_02
| |
| 09:19 | <ltspbot`> alkisg: "SCREEN_02" :: to get a root shell on an Ubuntu thin client: https://help.ubuntu.com/community/UbuntuLTSP/ClientTroubleshooting#Using%20a%20shell%20SCREEN
| |
| 09:20 | <alkisg> Leave it running there, and execute `sudo killall nbd-server` on the server
| |
| 09:20 | Works for me.
| |
| 09:32 | abeehc has quit IRC | |
| 09:32 | abeehc has joined #ltsp | |
| 09:32 | alkisg has quit IRC | |
| 09:33 | gorkhaan has quit IRC | |
| 09:36 | otavio has quit IRC | |
| 09:36 | otavio has joined #ltsp | |
| 09:36 | otavio has joined #ltsp | |
| 09:37 | otavio has quit IRC | |
| 09:37 | otavio has joined #ltsp | |
| 09:39 | dobber has quit IRC | |
| 09:47 | gorkhaan has joined #ltsp | |
| 09:52 | <bigd> oh, alkisg is already gone ... maybe someone else can anser whats the reason not to use?
| |
| 09:52 | +w
| |
| 09:52 | ups ... not to use nbd-proxy
| |
| 09:53 | <mgariepy> bigd, with some network card it doesn't work correctly
| |
| 09:53 | <bigd> ah, ok
| |
| 09:55 | hmm ... maybe ill try it tomorrow w/o nbd-proxy under 10.10 ...
| |
| 10:09 | Trixboxer has quit IRC | |
| 10:18 | bigd has quit IRC | |
| 10:21 | staffencasa has joined #ltsp | |
| 10:29 | gnunux has quit IRC | |
| 10:35 | loather-work has joined #ltsp | |
| 10:37 | gent_geen has quit IRC | |
| 10:38 | gentgeen__ has joined #ltsp | |
| 10:39 | alkisg has joined #ltsp | |
| 11:07 | Pweg has joined #ltsp | |
| 11:08 | Da-Geek has quit IRC | |
| 11:26 | alkisg has quit IRC | |
| 11:58 | mistik1_ has joined #ltsp | |
| 12:01 | mistik1 has quit IRC | |
| 12:01 | mistik1_ is now known as mistik1 | |
| 12:09 | david_ has joined #ltsp | |
| 12:10 | david_ has left #ltsp | |
| 12:28 | vagrantc has joined #ltsp | |
| 13:16 | robehend1 has joined #ltsp | |
| 13:20 | <robehend1> how do you guys manage the desktop settings of users on LTSP? the /etc/skele file? I'm talking adding icons to all desktops, locking down the panels, adding/removing menus, the whole she-bang
| |
| 13:22 | <Gadi> robehend1: sabayon can help with that
| |
| 13:23 | <robehend1> Gadi: I've tried Sabayon, and am currently using it. Problem I have is when I want to update a profile, Sabayon crashes more than runs, or simply doesnt apply. I usually have to build a whole new profile and apply it to the users, which takes much to long
| |
| 13:23 | <Gadi> yeah, it crashed on me until I fixed it
| |
| 13:23 | :)
| |
| 13:23 | <robehend1> oh, do tell
| |
| 13:23 | <ogra_ac> did you file bugs so people can fix it ?
| |
| 13:24 | <Gadi> python is not very forgiving
| |
| 13:24 | <robehend1> and sadly beyond my scope
| |
| 13:24 | <Gadi> I fixed it just last week and IMd scotty
| |
| 13:24 | it was a one line fix
| |
| 13:24 | * Gadi looks for it | |
| 13:24 | <Gadi> one sec
| |
| 13:34 | ok, edit /usr/share/pyshared/sabayon/storage.py
| |
| 13:34 | johnny has left #ltsp | |
| 13:34 | <Gadi> and change
| |
| 13:34 | if backup:
| |
| 13:34 | os.remove (backup)
| |
| 13:34 | to .....
| |
| 13:34 | if backup:
| |
| 13:34 | try:
| |
| 13:34 | os.remove (backup)
| |
| 13:34 | except:
| |
| 13:34 | pass
| |
| 13:34 | that fixed it for me
| |
| 13:35 | <robehend1> hmm, alright, let me get into the server and see what i can do
| |
| 13:36 | <Gadi> (it's on line 697
| |
| 13:36 | or so
| |
| 13:36 | )
| |
| 13:36 | <robehend1> is python spacing dependant? Not familiar with it
| |
| 13:37 | <vagrantc> very much so
| |
| 13:37 | <robehend1> ah, alrighty. will make sure to keep formatting then
| |
| 13:39 | <dberkholz> pay attention to spaces vs tabs too
| |
| 13:39 | <robehend1> ack. i used a combination of both.
| |
| 13:39 | <dberkholz> python has no idea how many spaces a tab is equivalent to
| |
| 13:40 | <Gadi> vaery readable - not very forgiving
| |
| 13:40 | :)
| |
| 13:40 | <robehend1> haha
| |
| 13:40 | <dberkholz> so it's best if it's all one way or the other, doesn't really matter which
| |
| 13:40 | <robehend1> so for that code, what would be the best way to get it in there
| |
| 13:40 | <Gadi> spaces
| |
| 13:41 | seems the authors use the 4 spaces format
| |
| 13:41 | alkisg has joined #ltsp | |
| 13:41 | <robehend1> ah alright
| |
| 13:42 | johnny has joined #ltsp | |
| 13:45 | <robehend1> there all done with spaces
| |
| 13:45 | <komunista> !lts.conf
| |
| 13:45 | <ltspbot`> komunista: "lts.conf" :: http://manpages.ubuntu.com/lts.conf
| |
| 13:46 | <moldy> 4 spaces is pretty popular for python, see PEP8
| |
| 13:46 | i wouldn't call it a "de facto standard", but it's pretty close to it
| |
| 13:48 | <robehend1> ack, i've failed some how. I'm getting an "IndentationError: expected an indented block"
| |
| 13:49 | <moldy> robehend1: that means the indentation is missing or inconsistent
| |
| 13:50 | <robehend1> moldy: ah, ok. I think i found my error, testing now
| |
| 13:50 | <moldy> (python does not care how you indent, as long as you do it consistently)
| |
| 13:51 | <robehend1> tricksy language, isnt it
| |
| 13:51 | <moldy> not really
| |
| 13:51 | <robehend1> ah, there, got it.
| |
| 13:52 | <moldy> if python complains about your indentation, your indentation would be bad style in any language
| |
| 13:52 | <robehend1> you were right, when i cleared out my tabs, i didnt add enough spaces
| |
| 13:52 | i think i just need to sit down for a while and start breaking into python. I've never taken the time
| |
| 13:52 | [GuS] has quit IRC | |
| 13:53 | <moldy> which editor are you using?
| |
| 13:53 | any decent editor should make this trivial :)
| |
| 13:54 | <robehend1> just was using vi, since i was ssh'd in
| |
| 13:56 | <moldy> vi/vim is great
| |
| 13:57 | if you're using vim, see http://vim.wikia.com/wiki/Super_retab
| |
| 13:57 | <robehend1> i'm not as familiar with it as i'd like to be, since I started with Nano.
| |
| 13:58 | <moldy> and also http://henry.precheur.org/vim/python
| |
| 13:58 | (though many distros should come with a vim pre-configured for python these days)
| |
| 13:59 | <robehend1> hmm, alright
| |
| 14:02 | <Gadi> did it fix the crashing for you?
| |
| 14:02 | <robehend1> haven't been able to check yet, sadly. just had an issue with our Moodle server crop up. Love me some k-12 education days
| |
| 14:10 | komunista has quit IRC | |
| 14:18 | RiXtEr has quit IRC | |
| 14:18 | RiXtEr has joined #ltsp | |
| 14:24 | Kicer86 has joined #ltsp | |
| 14:46 | robehend1 has left #ltsp | |
| 14:47 | Mip5 has left #ltsp | |
| 14:52 | <alkisg> Hi vagrantc, do you have a moment to tell me what parts of the filesystem are writeable (and when) when a client boots with NFS? The initramfs is always writeable, and in the end the bind-mounted directories are writeable, right? So the checks for non-writeable /etc/hostname etc are only before the bindmounts function is called?
| |
| 14:54 | And that function is called from the initramfs/nfs-bottom/ltsp? (the ltsp_bindmounts file is missing on Lucid...)
| |
| 14:56 | bobby_C has joined #ltsp | |
| 15:00 | mikkel has joined #ltsp | |
| 15:03 | <vagrantc> alkisg: basically, yes.
| |
| 15:04 | alkisg: client/initscripts/ltsp-bindmounts lists the various defaults for what gets mounted in debian, at least.
| |
| 15:04 | some of which may not be present (i.e. older debian versions)
| |
| 15:04 | <alkisg> I think there's a bug report about it missing on lucid
| |
| 15:04 | <vagrantc> not at all surprising
| |
| 15:05 | alkisg: i'm not sure what checks of /etc/hostname you're talking about
| |
| 15:06 | the mounts would happen in initramfs/nfs-bottom/ltsp ... so it depends when the writeability checks happen
| |
| 15:07 | <alkisg> E.g. in nfs-bottom/ltsp, we have some checks there
| |
| 15:07 | cat /proc/sys/kernel/hostname > ${rootmnt}/etc/hostname || true
| |
| 15:07 | if echo '# Generated by ltsp' 2>/dev/null > "$rootmnt/etc/resolv.conf"; then
| |
| 15:07 | So those aren't needed?
| |
| 15:08 | <vagrantc> ?
| |
| 15:08 | i don't see what's not needed about them
| |
| 15:08 | <alkisg> If we know that /etc/hostname is writeable, why check if it is?
| |
| 15:09 | (just as an example for me to understand the mechanism - I wasn't thinking of those specific lines...)
| |
| 15:09 | Gadi has left #ltsp | |
| 15:09 | * vagrantc doesn't see that as *checking* the writeability of /etc/hostname | |
| 15:10 | <alkisg> # Check if $rootmnt/etc/resolv.conf is writable
| |
| 15:10 | <vagrantc> ah, that one
| |
| 15:11 | someone could request that it not be writeable
| |
| 15:11 | i.e. change the LTSP_RW_* variables
| |
| 15:11 | <alkisg> Ah, I see
| |
| 15:12 | <vagrantc> no harm in having it there, is there?
| |
| 15:13 | <alkisg> No no I was thinking about writing some configuration files and letting the normal os initscripts process them instead of writing the processing code ourselves...
| |
| 15:13 | Erm I'm not expressing that well, I'll give a specific example when I find one :)
| |
| 15:14 | Thanks for the lesson!
| |
| 15:26 | Kicer86 has quit IRC | |
| 15:32 | mgariepy has quit IRC | |
| 16:02 | Gadi has joined #ltsp | |
| 16:06 | gorkhaan has quit IRC | |
| 16:12 | vagrantc has quit IRC | |
| 16:15 | Mobe_ has quit IRC | |
| 16:34 | Mobe has joined #ltsp | |
| 16:37 | gorkhaan has joined #ltsp | |
| 16:56 | mikkel has quit IRC | |
| 16:56 | bobby_C has quit IRC | |
| 16:58 | alkisg has quit IRC | |
| 17:35 | ogra_ac has quit IRC | |
| 17:35 | Pweg has quit IRC | |
| 17:49 | ogra_ac has joined #ltsp | |
| 18:02 | zz_evil_root is now known as evil_root | |
| 18:05 | evil_root is now known as zz_evil_root | |
| 18:07 | zz_evil_root is now known as evil_root | |
| 18:11 | v4s has joined #ltsp | |
| 18:21 | vagrantc has joined #ltsp | |
| 18:44 | staffencasa has quit IRC | |
| 19:07 | Gadi has left #ltsp | |
| 19:23 | vagrantc has quit IRC | |
| 19:31 | v4s has quit IRC | |
| 20:05 | litlebuda has joined #ltsp | |
| 20:11 | evil_root is now known as zz_evil_root | |
| 20:24 | litlebuda has quit IRC | |
| 20:34 | ogra has quit IRC | |
| 20:35 | ogra has joined #ltsp | |
| 20:48 | leio has quit IRC | |
| 20:49 | leio has joined #ltsp | |
| 20:49 | leio has joined #ltsp | |
| 21:08 | ltspbot has joined #ltsp | |
| 21:08 | -leguin.freenode.net- [freenode-info] if you're at a conference and other people are having trouble connecting, please mention it to staff: http://freenode.net/faq.shtml#gettinghelp | |
| 21:09 | ltspbot` has quit IRC | |
| 21:29 | zz_evil_root is now known as evil_root | |
| 22:11 | cody_g has joined #ltsp | |
| 22:17 | cody_g has left #ltsp | |
| 22:19 | MorningSon has quit IRC | |
| 22:34 | daya has joined #ltsp | |
| 23:01 | evil_root is now known as zz_evil_root | |
| 23:01 | <chupacabra> http://austinteapotparty.com/node/1
| |
| 23:01 | zz_evil_root is now known as evil_root | |
| 23:17 | alkisg has joined #ltsp | |
| 23:30 | mistik1_ has joined #ltsp | |
| 23:31 | mistik1 has quit IRC | |
| 23:31 | mistik1_ is now known as mistik1 | |
| 23:58 | alkisg has quit IRC | |