IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 25 February 2013 (all times are UTC)
| 00:06 | tol has left IRC (tol!~tol@lvps178-77-99-218.dedicated.hosteurope.de, Ping timeout: 255 seconds) | |
| 00:12 | tol has joined IRC (tol!~tol@lvps178-77-99-218.dedicated.hosteurope.de) | |
| 00:21 | <sbalneav> jammcq: ping
| |
| 00:24 | <jammcq> sbalneav: Scotty !!!!!!!!!!!!!
| |
| 00:25 | * vagrantc waves | |
| 00:25 | <jammcq> hey vagrantc
| |
| 00:27 | <sbalneav> Hey ho.
| |
| 00:28 | Did alkisg try the shutdown stuff I added in?
| |
| 00:28 | Also, so, I've got it going very nicely. Don't have diskless, but I have a vm that has no users in it, with pam_sshauth set up.
| |
| 00:29 | I can log in with lightdm, and start a remote X session. Very sweet.
| |
| 00:30 | <vagrantc> nice.
| |
| 00:52 | vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Quit: leaving) | |
| 01:45 | ltspuser_41 has joined IRC (ltspuser_41!ad084d59@gateway/web/freenode/ip.173.8.77.89) | |
| 01:47 | <ltspuser_41> Hi, I am having a problem with ltsp, when using text editor like liebra office, typing is fine, but if i use kate then it can take as long as a second for each character to appear? any thoughts
| |
| 01:49 | also, running old gateway all in ones p3's as terminals, none of them will start now, as they use intel 82815 video, using suse 12.2 ltsp, have been using 11.1 for the last long while without problems
| |
| 01:51 | used the oneclick install from the open suse site
| |
| 02:01 | ltspuser_41 is now known as briang | |
| 02:05 | <briang> hello
| |
| 02:11 | Has anybody here had same problems?
| |
| 02:28 | kev_n has left IRC (kev_n!~kevin@ip68-0-150-106.tc.ph.cox.net, Remote host closed the connection) | |
| 02:32 | <briang> hello all
| |
| 02:44 | BrianG2 has joined IRC (BrianG2!~brian@173-8-77-89-Sacramento.hfc.comcastbusiness.net) | |
| 02:47 | ltsp has joined IRC (ltsp!bot@ltsp.org) | |
| 03:04 | anunnakiii has joined IRC (anunnakiii!~anunnaki@c-174-55-35-255.hsd1.pa.comcast.net) | |
| 03:04 | <Enslaver> I have no clue where to begin on getting a fat client going, any tips? Any particular packages that i need to remove or disable to get things running smooth?
| |
| 03:05 | <jammcq> why are you removing things?
| |
| 03:05 | anunnakiii has left IRC (anunnakiii!~anunnaki@c-174-55-35-255.hsd1.pa.comcast.net, Read error: Connection reset by peer) | |
| 03:05 | <jammcq> isn't a fat client simply the same as a thin, but with a window manager and one or more apps?
| |
| 03:05 | <Enslaver> It installs a lot of crap by default, trying to slim it down, plus i've found that other display managers are trying to load over top of ldm
| |
| 03:06 | <jammcq> hmm
| |
| 03:06 | <Enslaver> plus NetworkManager caused me some issues, had to get rid of that too
| |
| 03:06 | PackageKit
| |
| 03:06 | gdm
| |
| 03:09 | Parker955_Away is now known as Parker955 | |
| 03:23 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 03:29 | <alkisg> Good morning
| |
| 03:29 | 00:28 <sbalneav> Did alkisg try the shutdown stuff I added in?
| |
| 03:29 | ==> Scotty as I keep saying... what you committed was what we had at release time, i.e. when the problem occured, they don't do anything to resolve it
| |
| 03:30 | There are 2 ways to solve the problem, the good one and the bad one
| |
| 03:30 | <jammcq> alkisg: scotty was first fixing the problem of a normal logout not running the rc scripts
| |
| 03:30 | <alkisg> The good one is to have the ssh connection up after Xorg has died
| |
| 03:31 | The bad one is just to forcibly kill all the user processes
| |
| 03:31 | <jammcq> once you test that and make sure the scripts are called and do the right thing, then he'll add the signal catching code
| |
| 03:31 | <alkisg> We need your help in the first method; not in the second one...
| |
| 03:31 | jammcq: ok, that part is tested, all the scripts are called and do the right thing
| |
| 03:31 | <jammcq> cuz if it doesn't work on a normal logout, what's the point of worrying if it doesn't work when xorg dies?
| |
| 03:31 | ok, have you told scotty ?
| |
| 03:32 | he asked for feedback
| |
| 03:32 | <alkisg> I didn't know he was asking for feedback on if the scripts run
| |
| 03:32 | <jammcq> all you said was "we're back to where we started"
| |
| 03:32 | so from what you can tell, if a user logs out normally, it's perfectly fine?
| |
| 03:32 | <alkisg> No
| |
| 03:32 | <jammcq> and then they can log in again and still fine?
| |
| 03:32 | <alkisg> No
| |
| 03:32 | <jammcq> you said it runs the scripts????
| |
| 03:33 | <alkisg> We're still were we started, i.e. when the problem existed
| |
| 03:33 | Yes, the scripts don't solve the problem
| |
| 03:33 | <jammcq> so are the scripts not doing the right thing?
| |
| 03:33 | <alkisg> Yes
| |
| 03:33 | <jammcq> so fix the scripts
| |
| 03:33 | <alkisg> I can't
| |
| 03:33 | <jammcq> why not?
| |
| 03:33 | <alkisg> Because they don't run in the correct time
| |
| 03:33 | <jammcq> why not?
| |
| 03:34 | <alkisg> Because the user processes are still running
| |
| 03:34 | <jammcq> the last script to run should be the one that brings downt he ssh tunnel
| |
| 03:34 | is the window mgr not exiting?
| |
| 03:34 | <alkisg> Please, could I have an answer on this question?
| |
| 03:34 | There are two ways to solve the problem
| |
| 03:34 | One, to run the scripts after xorg has died
| |
| 03:35 | Two, to have the scripts kill all the user processes
| |
| 03:35 | <jammcq> don't worry about xorg dying yet. let's make it work normally first
| |
| 03:35 | <alkisg> Which method are we trying?
| |
| 03:35 | <jammcq> we're trying to make a normal logout work
| |
| 03:35 | whatever LDM starts on the way in, it should stop on the way out
| |
| 03:35 | <alkisg> Since now, always, user processes died with xorg
| |
| 03:36 | LDM didn't kill the processes; xorg did
| |
| 03:36 | <jammcq> which user processes?
| |
| 03:36 | <alkisg> Let's say dbus, pulseaudio, gconf...
| |
| 03:36 | <jammcq> who starts those processes?
| |
| 03:36 | <alkisg> firefox, gnome-session, whatever runs locally
| |
| 03:36 | gnome-session starts them
| |
| 03:36 | And LDM starts gnome-session
| |
| 03:36 | <jammcq> then why doesn't gnome-session stop them?
| |
| 03:36 | gnome-session is the parent to those processes,r ight?
| |
| 03:37 | <alkisg> It leaves xorg to kill them
| |
| 03:37 | Not for all of them, no
| |
| 03:37 | It's a parent to less than half of them
| |
| 03:37 | The others have 1 as the parent
| |
| 03:37 | (init)
| |
| 03:37 | <jammcq> then perhaps you need a K9X script to kill them, before the ssh stop script runs
| |
| 03:37 | <alkisg> Then we're trying for solution two, the bad one
| |
| 03:37 | <jammcq> why is that bad?
| |
| 03:37 | <alkisg> I don't need Scotty's help there
| |
| 03:38 | Hmmm or maybe...
| |
| 03:38 | <jammcq> either LDM is gonna kill them or a script is gonna kill them, if you want it to happen before SSH goes away
| |
| 03:38 | <alkisg> OK let's talk about that first,
| |
| 03:38 | <jammcq> SSH isn't going to stick around after LDM, it doesn't need to
| |
| 03:38 | sure
| |
| 03:38 | <alkisg> Why not let xorg kill them?
| |
| 03:39 | <jammcq> it's too late
| |
| 03:39 | <alkisg> Is that just because of a bad design?
| |
| 03:39 | <jammcq> no
| |
| 03:39 | <alkisg> Where the ssh channel can't stay up after xorg?
| |
| 03:39 | <jammcq> whatever starts something should stop it
| |
| 03:39 | ssh is started as a child under LDM which is a child under xorg
| |
| 03:39 | so when LDM goes away, ssh should be gone
| |
| 03:39 | <alkisg> We can change that sequence
| |
| 03:40 | We can have xorg be a child of ldm
| |
| 03:40 | <jammcq> you really want to get into that big of a change?
| |
| 03:41 | I'm a firm believer in unwinding things in the reverse order of how they got started
| |
| 03:41 | sure we can change th order, it just seems like a huge change
| |
| 03:42 | <alkisg> OK there's a long talk there... e.g. I prefer LTSP not to use hacks if possible
| |
| 03:42 | <jammcq> plus, I think it puts alot more complexity into LDM. it has to spawn the Xserver and then wait around for it to actually be ready before proceding
| |
| 03:42 | <alkisg> It's just a matter of who the parent of the ssh channel is
| |
| 03:42 | <jammcq> what's the hack, a script that cleans up when LDM is shutting down?
| |
| 03:42 | <alkisg> The greeter or ldm
| |
| 03:43 | The greeter signals LDM when it has the username/pass anyway
| |
| 03:43 | So ldm wouldn't have to wait for xorg
| |
| 03:43 | But for the greeter as it does now
| |
| 03:43 | <jammcq> LDM shouldn't call the greeter until Xorg is up and running
| |
| 03:43 | <alkisg> Having to write cleanup scripts means we're not putting our code on the correct places
| |
| 03:44 | xinit would call the greeter, not ldm
| |
| 03:44 | Anyways,
| |
| 03:44 | As I said that's a big talk to do so now,
| |
| 03:44 | <jammcq> but your solution is to just have the processes respond to SIGTERM
| |
| 03:44 | <alkisg> So, let us get back to the pth for the solution
| |
| 03:44 | <jammcq> which isn't great either
| |
| 03:44 | <alkisg> My solution is to not worry about the processes
| |
| 03:44 | Not a single pkill, pgrep in the code
| |
| 03:45 | Which is minimal code, and guarantees that any leftover processes are not our bugs
| |
| 03:45 | But again anyways, we can review all that again with the lightdm change
| |
| 03:45 | Let's focus on the current problem + solution,
| |
| 03:45 | so, I thought sbalneav was going to help with the first solution,
| |
| 03:46 | <jammcq> he is
| |
| 03:46 | <alkisg> but we'll be doing the second one instead, i.e. kill all user processes
| |
| 03:46 | <jammcq> he's made LDM run the K scripts on the way out
| |
| 03:46 | <alkisg> It already did
| |
| 03:46 | <jammcq> now he has to add the sig handling
| |
| 03:46 | it used to, then someone ripped it out
| |
| 03:46 | <alkisg> Me
| |
| 03:46 | <jammcq> he put it back
| |
| 03:46 | <alkisg> Trying for the correct solution
| |
| 03:47 | jammcq: Sorry, I might be missing something. Do we plan on killing the user processes or not?
| |
| 03:47 | <jammcq> I'd say yes
| |
| 03:47 | <alkisg> OK, so please bare with me
| |
| 03:47 | For me, that's "solution 2, the bad one"
| |
| 03:47 | I was hoping for "solution 1, let xorg kill them, but have ssh up afterwards"
| |
| 03:47 | <jammcq> how about: "solution 2, the less intrusive one"
| |
| 03:48 | <alkisg> OK, now:
| |
| 03:48 | <jammcq> if you want #1, then we have alot of work to do
| |
| 03:48 | <alkisg> OK
| |
| 03:48 | <jammcq> it's not my decision, i'm just telling how I understand it
| |
| 03:48 | <alkisg> That was why I removed the rcfiles("stop") thought, I was hoping for #1
| |
| 03:48 | But lets move on
| |
| 03:48 | <jammcq> k
| |
| 03:48 | <alkisg> So, for #2, I can do the correct things,
| |
| 03:49 | which would be to either have a save/restore environment script,
| |
| 03:49 | or have a shell trap in X94,
| |
| 03:49 | so that the rcfiles("stop") actually uses the same environment
| |
| 03:49 | Because with the current code, SSH_HOME wasn't even available
| |
| 03:49 | And also the cleanup didn't happen at K* time anyway
| |
| 03:49 | The only K* script was some ltsp-cluster logging
| |
| 03:50 | Anyways I can solve all the shell problems there,
| |
| 03:50 | <jammcq> I think you just need to make sure that the K script for shutting down SSH is one of the last things to run
| |
| 03:50 | <alkisg> And then to kill the processes I'll have something like this: http://paste.ubuntu.com/5563762/
| |
| 03:50 | jammcq: there's only 1 K* script
| |
| 03:50 | That tells ltsp-cluster "log the logout event"
| |
| 03:51 | The cleanup didn't happen at K*
| |
| 03:51 | But don't worry about that part, I can do that
| |
| 03:51 | <jammcq> k
| |
| 03:51 | <alkisg> Check the paste please
| |
| 03:51 | So, that script is tested and it works fine if called by ldm at logout
| |
| 03:52 | <jammcq> so that goes in a K script?
| |
| 03:52 | <alkisg> (as an X* script, K* scripts do not have the proper environment)
| |
| 03:52 | For the purpose of our chat here, let's say it's a K* script, ok, don't worry about that
| |
| 03:52 | <jammcq> when you say 'environment', are you talking about env variables?
| |
| 03:52 | <alkisg> Yes
| |
| 03:52 | E.g. SSH_HOME
| |
| 03:52 | So my question is...
| |
| 03:53 | When xorg or ldm crash, I think that script only has a few msec before it's killed by process management
| |
| 03:53 | So it doesn't have enough time to actually kill the processes and wait for them to die
| |
| 03:53 | <jammcq> it will when LDM catches the signal
| |
| 03:54 | <alkisg> Do you think that ldm catching the signals, will help with that?
| |
| 03:54 | OK, can I have that code to test?
| |
| 03:54 | <jammcq> scotty will do that
| |
| 03:54 | but how about you get things working properly when a normal logout happens?
| |
| 03:54 | <alkisg> OK, please tell him I'm waiting for that, and if it works, I can take care of everything else
| |
| 03:54 | <jammcq> maybe we can get scotty to add the sig_handlers tomorrow
| |
| 03:55 | <alkisg> (05:54:32 πμ) jammcq: but how about you get things working properly when a normal logout happens? ==> that script takes care of that
| |
| 03:55 | <jammcq> ok
| |
| 03:55 | <alkisg> As long as we have 2-3 secs to wait for the processes to exit on TERM so that we don't have to send them a SIGKILL
| |
| 03:55 | I've tested it, it works fine on normal logout
| |
| 03:56 | Not with the code sbalneav pushed, but don't worry about that, I only need the signals code from Scotty...
| |
| 03:57 | Actually, I can try the signal() or XSetIOErrorHandler() calls myself...
| |
| 03:57 | <jammcq> I don't think you want XSetIOErrorHanlder
| |
| 03:58 | cuz LDM doesn't have an open connection to the xserver
| |
| 03:58 | but i'm happy to see you becoming a C hacker
| |
| 03:58 | <alkisg> I've been writing C 20 years ago... I just don't want to learn the glib / Xorg api :-/
| |
| 03:59 | Even written C compilers for parallel programming...
| |
| 03:59 | <jammcq> knowing how to program in C is sorta like knowing how to weld with an arc welder
| |
| 03:59 | so much power in your hands
| |
| 03:59 | but watch it, cuz you can get burned
| |
| 03:59 | <alkisg> Hehe :)
| |
| 04:00 | My objective is to make the Linux experience smoother for schools here
| |
| 04:00 | * jammcq did alot of arc welding in a previous life | |
| 04:00 | sep has left IRC (sep!~sep@40.211.jostedal.no, Ping timeout: 245 seconds) | |
| 04:00 | <alkisg> So I aim at minimal code at the correct places... not big apps
| |
| 04:01 | jammcq: what I'm worried about is that xinit will kill ldm after 1 sec or so, when it sees that xorg has crashed
| |
| 04:01 | Because ldm is what we tell xinit to run inside X
| |
| 04:02 | <jammcq> but LDM will catch that with a signal handler
| |
| 04:02 | <alkisg> I'm thinking xinit will send "TERM; sleep 1; KILL"
| |
| 04:02 | So upon KILL we'll no longer be able to propery run cleanup scripts
| |
| 04:03 | <jammcq> I really don't think xinit will do that
| |
| 04:03 | <alkisg> k, let's hope so
| |
| 04:03 | <jammcq> in fact, I don't think xinit will send anything. I think when xinit dies, the kernel will send the same signal to the children as what was sent to xinit
| |
| 04:04 | so if xinit dies due to a bug, it'll likely be a SIGILL or SIGSEGV
| |
| 04:05 | <alkisg> In shell scripts, we usually catch those: HUP INT QUIT KILL SEGV PIPE TERM
| |
| 04:05 | <jammcq> we might see a SIGHUP or SIGINT
| |
| 04:05 | <alkisg> Should I put those for LDM too?
| |
| 04:05 | <jammcq> well... KILL can't be caught
| |
| 04:05 | but the others yes
| |
| 04:05 | I don't think you'd ever see PIPE on this
| |
| 04:05 | LDM isn't being used in any part of a pipeline
| |
| 04:06 | but ILL would be good to catch too
| |
| 04:06 | that's an Illegal instruction
| |
| 04:08 | * alkisg notes down to remove QUIT from our code... | |
| 04:09 | <alkisg> jammcq: how can I send QUIT from the keyboard?
| |
| 04:09 | INT=Ctrl+C, right?
| |
| 04:09 | man 7 signal
| |
| 04:09 | <jammcq> kill 3
| |
| 04:09 | <alkisg> SIGINT 2 Term Interrupt from keyboard
| |
| 04:09 | SIGQUIT 3 Core Quit from keyboard
| |
| 04:09 | <jammcq> or kill -QUIT
| |
| 04:09 | i think
| |
| 04:10 | <alkisg> No not from a kill command, but from the keyboard...
| |
| 04:10 | Ctrl+C is TERM, right?
| |
| 04:10 | <jammcq> oh
| |
| 04:10 | <alkisg> And Ctrl+/ is KILL?
| |
| 04:10 | <jammcq> do this: stty -a
| |
| 04:10 | you'll see that quit = ^\
| |
| 04:10 | so control backslash
| |
| 04:10 | <alkisg> Gotcha, ty
| |
| 04:15 | Hmm, so one cannot send TERM from the keyboard
| |
| 04:16 | <jammcq> is that a problem?
| |
| 04:16 | <alkisg> Nope not at all, I'm still trying to get all the linux process handling in my mind...
| |
| 04:24 | Parker955 is now known as Parker955_Away | |
| 04:27 | <alkisg> jammcq: when LDM gets an e.g. SIGSEGV from xinit/the kernel, what do you think will happen with the X* scripts? At that point, they're in the X95-run-xsession...
| |
| 04:27 | <jammcq> nothing
| |
| 04:27 | <alkisg> So, will the `su - user -c xsession` command stop? Will they continue to run?
| |
| 04:27 | Will they get a signal as well?
| |
| 04:27 | Will the X96+ scripts run?
| |
| 04:27 | <jammcq> no, I don't think they'll receive the signal
| |
| 04:28 | BrianG2 has left IRC (BrianG2!~brian@173-8-77-89-Sacramento.hfc.comcastbusiness.net, Quit: Konversation terminated!) | |
| 04:28 | <jammcq> not until LDM dies or sends a signal itself
| |
| 04:28 | <alkisg> But the xsession will, right?
| |
| 04:28 | <jammcq> who runs xsession?
| |
| 04:28 | ldm ?
| |
| 04:28 | <alkisg> LDM runs the X* scripts, and one of them runs the xsession
| |
| 04:28 | LDM => ldm-script xsession => X* scripts, X95-run-xsession
| |
| 04:29 | <jammcq> then if LDM intercepts the signal, it's up to you the programmer to do something with it
| |
| 04:29 | <alkisg> So we'll still have to kill the user processes ourselves, gotcha
| |
| 04:29 | cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Quit: cyberorg) | |
| 04:29 | <jammcq> so you might need to send the signal to the X95 script
| |
| 04:29 | <alkisg> It'll be much easier if we forward the signal to the X* shell scripts
| |
| 04:30 | <jammcq> or you can call the K script that does the kill for you
| |
| 04:30 | I think
| |
| 04:30 | <alkisg> The problem with the K* scripts is that the X* scripts environment is not available there
| |
| 04:30 | <jammcq> right
| |
| 04:30 | <alkisg> So normally they don't have access to $SSH_HOME
| |
| 04:30 | It'd be better to run the K* scripts at the end of the X* scripts, or with a trap
| |
| 04:31 | For now I used some "env save / env restore" scripts, but I think I'll remove those
| |
| 04:32 | jammcq: thanks for the process management tutorial, I really think it helped :)
| |
| 04:32 | So I should really write your name in the commit for the fix :D
| |
| 04:32 | <jammcq> I hope i'm right
| |
| 04:33 | too bad we're not doing this in Perl, I could really help there
| |
| 04:34 | <alkisg> My favorite languages were Pascal and Assembly, but now I'm stuck with shell... :)
| |
| 04:35 | cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg) | |
| 04:35 | * alkisg even wrote a tetris in assembly in less than 1k size, two decades ago... | |
| 04:52 | * alkisg dedicates http://www.youtube.com/watch?v=bwAw9ThDQmk to all the good LTSP folks here... | |
| 04:52 | <jammcq> watching
| |
| 04:52 | well... after the advertising :)
| |
| 04:53 | ah, old classic
| |
| 04:55 | sha_ has joined IRC (sha_!~sha@e177172035.adsl.alicedsl.de) | |
| 04:56 | <Enslaver> Hmm, it correctly updates fstab yet doesn't mount /home, maybe somehow rc.sysinit isn't getting run?
| |
| 04:58 | sha has left IRC (sha!~sha@e177172153.adsl.alicedsl.de, Ping timeout: 252 seconds) | |
| 05:01 | highvoltage has joined IRC (highvoltage!~highvolta@ubuntu/member/highvoltage) | |
| 05:12 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Ping timeout: 246 seconds) | |
| 05:14 | <Enslaver> looks like thats a know issue in ubuntu as well
| |
| 05:14 | hmm
| |
| 05:16 | I'm not sure how to proceed on this one, code in my own udev based autofs mounting?
| |
| 05:27 | lmds_ has left IRC (lmds_!~lmds@213.41.240.245, Read error: Connection reset by peer) | |
| 05:33 | BrianG2 has joined IRC (BrianG2!~brian@173-8-77-89-Sacramento.hfc.comcastbusiness.net) | |
| 05:35 | <BrianG2> hi, having a problem with intel 82815 terminals, worked with suse 11.1 ltsp but seems to be blank with suse 12.2, any one know where to look?
| |
| 05:36 | I mean intel video drivers
| |
| 05:45 | jammcq has left IRC (jammcq!~jam@c-69-245-75-255.hsd1.mi.comcast.net, Quit: leaving) | |
| 06:40 | vmlintu has joined IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi) | |
| 06:46 | work_alkisg is now known as alkisg | |
| 07:00 | <alkisg> BrianG2: what kernel does that have?
| |
| 07:00 | And what version of xserver-xorg-video-intel?
| |
| 07:19 | alkisg has left IRC (alkisg!~alkisg@plinet.ioa.sch.gr, Quit: Leaving.) | |
| 07:19 | work_alkisg has joined IRC (work_alkisg!~alkisg@plinet.ioa.sch.gr) | |
| 07:41 | work_alkisg has left IRC (work_alkisg!~alkisg@plinet.ioa.sch.gr, Quit: Leaving.) | |
| 07:45 | work_alkisg has joined IRC (work_alkisg!~alkisg@plinet.ioa.sch.gr) | |
| 08:23 | dievel has joined IRC (dievel!~dievel@2-229-104-66.ip196.fastwebnet.it) | |
| 08:27 | dievel has left IRC (dievel!~dievel@2-229-104-66.ip196.fastwebnet.it, Remote host closed the connection) | |
| 08:42 | <work_alkisg> !arcfour
| |
| 08:42 | <ltsp> arcfour: http://en.wikipedia.org/wiki/RC4 is an SSH cipher which is more than 2 times faster than the default 3DES. To enable it, set LDM_SSHOPTIONS="-o Ciphers=arcfour128".
| |
| 08:42 | work_alkisg is now known as alkisg | |
| 08:49 | highvoltage has left IRC (highvoltage!~highvolta@ubuntu/member/highvoltage, Ping timeout: 276 seconds) | |
| 08:49 | highvoltage has joined IRC (highvoltage!~highvolta@ubuntu/member/highvoltage) | |
| 09:35 | <vmlintu> alkisg: have you had dbus problems on ltsp fat clients? Like dbus not starting properly or applications complaining that they cannot connect to it.
| |
| 09:42 | <alkisg> vmlintu: on the second login of the same user?
| |
| 09:42 | If so, that's the LDM problem I've been talking about...
| |
| 09:42 | .gvfs not unmounting correctly
| |
| 09:42 | bbl
| |
| 09:44 | <vmlintu> no, on the first login
| |
| 09:44 | We are not using sshfs either
| |
| 09:52 | <Enslaver> http://www.inode0.com/ <-- anyone like my logo? I know the logo has a copywrite and yada yada, but i just think that looks neat
| |
| 09:53 | garymc has joined IRC (garymc!~chatzilla@host81-148-67-235.in-addr.btopenworld.com) | |
| 09:56 | <elias_a> Enslaver: What is copywrite?
| |
| 09:57 | <Enslaver> its like a copyright but instead of a right, its write. Like a copyleft
| |
| 09:58 | <elias_a> Sounds like bullshit to me :)
| |
| 09:58 | <Enslaver> it's a coder joke
| |
| 09:58 | <elias_a> Well, it still is bs to me.
| |
| 09:59 | dievel has joined IRC (dievel!~dievel@2-229-104-66.ip196.fastwebnet.it) | |
| 09:59 | dobber_ has joined IRC (dobber_!~dobber@89.190.199.210) | |
| 09:59 | <elias_a> On the other hand a coder who does not know IPR is very useful. Cheap labour. :P
| |
| 10:00 | Enslaver: Nice looking login!
| |
| 10:00 | <Enslaver> ty
| |
| 10:01 | well, thats where open source comes in =)
| |
| 10:02 | vmlintu has left IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi, Ping timeout: 264 seconds) | |
| 10:03 | <Enslaver> now if i could only get this damn nx java client to work and my /home to mount i'd be set
| |
| 10:07 | <elias_a> Enslaver: I still do not understand what open source has to do with graphical design. That is the bullshit part.
| |
| 10:08 | <Enslaver> oh the logo? Yeah i'm not a graphics designer so I dunno, i just threw that together.
| |
| 10:08 | <elias_a> Enslaver: The point is - is it your work or not?
| |
| 10:09 | If not, do you have the permission to use it?
| |
| 10:09 | <Enslaver> no, i just took something and made it look better.
| |
| 10:09 | bobby_C has joined IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at) | |
| 10:09 | <Enslaver> not gonna use it
| |
| 10:09 | <elias_a> IMHO FLOSS advocates should be very careful about IPR issues.
| |
| 10:09 | Ok.
| |
| 10:09 | No I will stop nagging.
| |
| 10:09 | now....
| |
| 10:10 | <Enslaver> I try to take breaks in-between code, that was during a break
| |
| 10:25 | alexqwesa_ has left IRC (alexqwesa_!~alex@109.172.12.47, Quit: Хана X'ам !!!) | |
| 10:25 | alexqwesa_ has joined IRC (alexqwesa_!~alex@109.172.12.47) | |
| 10:31 | <alkisg> The ltsp logo would look nicer if it was a clover :)
| |
| 10:56 | vmlintu has joined IRC (vmlintu!~vmlintu@cs181214103.pp.htv.fi) | |
| 10:59 | <alkisg> vmlintu: when that thing with dbus happens, can you try ls ~/.gvfs; mount | grep gvfs ?
| |
| 11:01 | Phantomas has joined IRC (Phantomas!~Phantomas@ubuntu/member/phantomas) | |
| 11:05 | <vmlintu> alkisg: I'll try that in a bit
| |
| 11:33 | alkisg has left IRC (alkisg!~alkisg@plinet.ioa.sch.gr, Quit: Leaving.) | |
| 11:36 | anunnaki has joined IRC (anunnaki!~anunnaki@c-174-55-35-255.hsd1.pa.comcast.net) | |
| 12:13 | khildin has joined IRC (khildin!~khildin@ip-80-236-227-148.dsl.scarlet.be) | |
| 12:23 | khildin has left IRC (khildin!~khildin@ip-80-236-227-148.dsl.scarlet.be, Remote host closed the connection) | |
| 12:27 | khildin has joined IRC (khildin!~khildin@ip-80-236-227-148.dsl.scarlet.be) | |
| 12:34 | biomorph has joined IRC (biomorph!~biomorph@91.85.204.16) | |
| 12:40 | khildin has left IRC (khildin!~khildin@ip-80-236-227-148.dsl.scarlet.be, Remote host closed the connection) | |
| 12:54 | vnc786 has left IRC (vnc786!~chatzilla@49.248.129.178, Quit: ChatZilla 0.9.90 [Firefox 11.0/20120410121533]) | |
| 13:14 | Enslaver has left IRC (Enslaver!~Enslaver@c-98-196-42-169.hsd1.tx.comcast.net, Read error: Connection reset by peer) | |
| 13:25 | alexqwesa_ has left IRC (alexqwesa_!~alex@109.172.12.47, Ping timeout: 260 seconds) | |
| 13:33 | Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas) | |
| 13:45 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 13:48 | monteslu_ is now known as monteslu | |
| 13:56 | awilliams has joined IRC (awilliams!mistik1@unaffiliated/mistik1) | |
| 13:56 | vmlintu has left IRC (vmlintu!~vmlintu@cs181214103.pp.htv.fi, Ping timeout: 255 seconds) | |
| 14:09 | vmlintu has joined IRC (vmlintu!~vmlintu@cs181214103.pp.htv.fi) | |
| 14:13 | jammcq has joined IRC (jammcq!~jam@c-69-245-75-255.hsd1.mi.comcast.net) | |
| 14:13 | <jammcq> good morning friends
| |
| 14:15 | dead_inside has joined IRC (dead_inside!~dead_insi@76.75.3.174) | |
| 14:19 | <elias_a> Good afternoon, jammcq!
| |
| 14:26 | <jammcq> hey elias_a
| |
| 14:41 | <alkisg> 'morning Jim
| |
| 14:42 | <jammcq> hey alkisg
| |
| 14:42 | <knipwim> hey guys
| |
| 14:42 | <jammcq> hey knipwim
| |
| 14:54 | <sbalneav> Morning all!
| |
| 15:07 | <jammcq> sbalneav: Scotty !!!!!!!!!!!
| |
| 15:07 | <sbalneav> jammcq: hey ho!
| |
| 15:08 | Cripes. Can't figure out an autoconf issue.
| |
| 15:24 | khildin has joined IRC (khildin!~khildin@ip-80-236-227-148.dsl.scarlet.be) | |
| 15:30 | khildin has left IRC (khildin!~khildin@ip-80-236-227-148.dsl.scarlet.be, Quit: I'm gone, bye bye) | |
| 15:35 | khildin has joined IRC (khildin!~khildin@ip-80-236-227-148.dsl.scarlet.be) | |
| 15:39 | mikkel has joined IRC (mikkel!~mikkel@80-71-132-15.u.parknet.dk) | |
| 15:45 | <sbalneav> I'm trying to edit http://wiki.ltsp.org/mediawiki/index.php?title=Dev:LTSPPamNotes&action=edit, but it's saying I can't save.
| |
| 15:45 | Any of the wiki people around?
| |
| 15:46 | <alkisg> sbalneav: put the content into pastebin so that I try
| |
| 15:46 | knipwim: ^
| |
| 15:47 | <sbalneav> http://pastebin.com/q9w0zRCn
| |
| 15:48 | <alkisg> sbalneav: maybe ...you're not an ltsp dev? :P
| |
| 15:48 | I saved it... you're logged in, right?
| |
| 15:48 | knipwim might be able to put you in the wiki devs team...
| |
| 15:48 | <jammcq> he should have full access
| |
| 15:48 | I checked last week
| |
| 15:53 | <sbalneav> Apparently, I can edit it now. Maybe I just couldn't create. Regardless, thanks. I can continue on from here.
| |
| 16:10 | <knipwim> sbalneav: is libnss-extrausers a debian/ubuntu specific package?
| |
| 16:15 | in other words, to package this for gentoo, do i take the [libnss-extrausers_0.6.orig.tar.gz] file from http://packages.debian.org/sid/libnss-extrausers ?
| |
| 16:16 | or do i take the sources from somewhere else
| |
| 16:16 | Enslaver has joined IRC (Enslaver!~Enslaver@c-98-196-42-169.hsd1.tx.comcast.net) | |
| 16:18 | <knipwim> Enslaver: perhaps a noob question, but how do you shut down the overlayfs enabled system?
| |
| 16:19 | like, do you have to pivot back to the readonly system somewhere?
| |
| 16:19 | <Enslaver> knipwim: OVERLAY=N
| |
| 16:20 | <knipwim> ok, but when i have an overlayd system, when i shutdown, the fuse module gets shut down and the shutdown halts
| |
| 16:21 | before it can reach /sbin/halt
| |
| 16:21 | <Enslaver> I see
| |
| 16:22 | hmmm
| |
| 16:22 | <knipwim> in my logic, i have to pivot back to the original fs somewhere, and then shutdown the fuse module
| |
| 16:23 | <Enslaver> is initram doing the pivot?
| |
| 16:23 | <knipwim> no, init-ltsp.d
| |
| 16:23 | <Enslaver> gotcha, i'll have to write that as part of a cleanup or shutdown method
| |
| 16:24 | biomorph has left IRC (biomorph!~biomorph@91.85.204.16, Remote host closed the connection) | |
| 16:24 | <knipwim> because normally the initramfs does this?
| |
| 16:25 | <Enslaver> yeah thats what handles the pivots for me, i load fuse module in initramfs
| |
| 16:26 | <knipwim> that was the plan, but i still can't boot the nfs dracut initrd
| |
| 16:27 | i got the feeling it has something to do with the client not having a dnsdomainname
| |
| 16:27 | <Enslaver> yeah i ran into that
| |
| 16:27 | <knipwim> did you solve it?
| |
| 16:28 | <Enslaver> my solution was to have my /etc/hosts built up and have dnsmasq issuing the in-addr.arpa's
| |
| 16:28 | but that didn't cause the client not to boot, it just gave me issues with init scripts
| |
| 16:29 | i also wrote in a check to say if $HOSTNAME = "(none)" || "localhost"
| |
| 16:29 | none or localhost
| |
| 16:29 | paste bin your dracut.conf
| |
| 16:29 | <knipwim> so, dnsmasq provides you clients with a dnsdomainname?
| |
| 16:30 | <sbalneav> knipwim: No idea.
| |
| 16:30 | <knipwim> Enslaver: i use this line (kernel version is an example):
| |
| 16:30 | <Enslaver> http://fpaste.org/OQ27/
| |
| 16:30 | <knipwim> dracut --force -m "kernel-modules nbd nfs network base" --filesystems "squashfs" /boot/initramfs-dracut-x86_64-3.3.8-gentoo.img 3.3.8-gentoo
| |
| 16:30 | Phantomas has joined IRC (Phantomas!~Phantomas@ubuntu/member/phantomas) | |
| 16:31 | <Enslaver> thats my dracut.conf
| |
| 16:31 | <sbalneav> knipwim: I ain't a packaging expert.
| |
| 16:31 | <Enslaver> where's the module for the overlay fs?
| |
| 16:32 | <knipwim> Enslaver: not using that yet, the overlayfs gets called after the initramfs
| |
| 16:33 | komunista has joined IRC (komunista!~slavko@87.244.209.121) | |
| 16:34 | <Enslaver> right, didn't you just say you couldn't boot the dracut initrd? I guess you meant due to hfs?
| |
| 16:34 | nfs
| |
| 16:34 | BrianG2 has left IRC (BrianG2!~brian@173-8-77-89-Sacramento.hfc.comcastbusiness.net, Quit: Konversation terminated!) | |
| 16:34 | <knipwim> yeah, i still got an older (014) dracut initramfs that does work for nfs
| |
| 16:35 | and i can't rebuild that one, because the 014 is incompatible with the current gentoo installs
| |
| 16:35 | i know, it's complicated
| |
| 16:35 | i have to fix the nfs nobody id issue
| |
| 16:36 | i have to fix the nfs nobody id issue
| |
| 16:36 | err
| |
| 16:36 | komunista has left IRC (komunista!~slavko@87.244.209.121, Client Quit) | |
| 16:36 | dobber_ has left IRC (dobber_!~dobber@89.190.199.210, Remote host closed the connection) | |
| 16:37 | dievel has left IRC (dievel!~dievel@2-229-104-66.ip196.fastwebnet.it, Remote host closed the connection) | |
| 16:37 | <knipwim> Enslaver: hmm, you have included the nfsv4 module
| |
| 16:37 | is that a different one than nfs.ko?
| |
| 16:38 | <Enslaver> yes
| |
| 16:38 | it turns out just by adding a flag to your /etc/exports NFS switches to conform without telling you.
| |
| 16:38 | <knipwim> hmm, i haven't included nfs like you do, i only call the dracut nfs module, and that includes the necessary nfs kernel stuff
| |
| 16:39 | maybe that would actually work
| |
| 16:39 | * knipwim is hopefull | |
| 16:39 | <Enslaver> well your NFS is a kernel module correct?
| |
| 16:40 | <knipwim> yes
| |
| 16:40 | the nfs.ko is included
| |
| 16:40 | but no nfsv4.ko
| |
| 16:41 | can you do an lsinitrd enslavers-dracut.img | grep nfsv4 ?
| |
| 16:41 | <Enslaver> you shouldn't need the v4
| |
| 16:41 | <knipwim> bbiab
| |
| 16:42 | <Enslaver> -rwxr--r-- 1 root root 27072 Feb 20 04:56 lib/modules/2.6.32-279.22.1.el6.thinclient.i686/kernel/fs/nfs/nfs_layout_nfsv41_files.ko
| |
| 16:43 | BrianG2 has joined IRC (BrianG2!~brian@173-8-77-89-Sacramento.hfc.comcastbusiness.net) | |
| 16:43 | Enslaver has left IRC (Enslaver!~Enslaver@c-98-196-42-169.hsd1.tx.comcast.net) | |
| 16:46 | <BrianG2> Hi got your advice, Ok I actually turned off encryption for this loop as its internal only, Improved the response 100% thanks for the point. I have not found the version of xserver... yet
| |
| 16:48 | vmlintu has left IRC (vmlintu!~vmlintu@cs181214103.pp.htv.fi, Ping timeout: 255 seconds) | |
| 16:48 | <BrianG2> alkisg -- thanks for your response
| |
| 16:49 | still working on adding the nvidia driver for this configuration, not sure which itel drivers to add
| |
| 17:26 | ||cw has left IRC (||cw!~chris@phpgroupware/cw, Remote host closed the connection) | |
| 17:31 | ||cw has joined IRC (||cw!~chris@gateway.wilsonmfg.com) | |
| 17:31 | ||cw has joined IRC (||cw!~chris@phpgroupware/cw) | |
| 17:33 | mikkel has left IRC (mikkel!~mikkel@80-71-132-15.u.parknet.dk, Quit: Leaving) | |
| 17:33 | Gremble has joined IRC (Gremble!~Ben@cpc35-aztw23-2-0-cust207.18-1.cable.virginmedia.com) | |
| 17:35 | mikkel has joined IRC (mikkel!~mikkel@80-71-132-15.u.parknet.dk) | |
| 17:39 | vmlintu has joined IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi) | |
| 17:43 | garymc has left IRC (garymc!~chatzilla@host81-148-67-235.in-addr.btopenworld.com, Ping timeout: 252 seconds) | |
| 17:53 | komunista has joined IRC (komunista!~slavko@87.244.209.121) | |
| 17:58 | <sbalneav> Gaaahhh. Our lives would be much more simple if any pam program defaulted to a certain service name, but you could override the service name on the command line with --service-name= or -s
| |
| 17:59 | Mip5 has joined IRC (Mip5!~chatzilla@208.72.71.69) | |
| 18:01 | <Mip5> G'morning folks. Anyone have any ideas on why "Reading package lists..." would take several minutes (10 minutes!) to complete? I'm running ubuntu 12.04.1, but with the 3.5 kernel. Thanks!
| |
| 18:03 | <sbalneav> Does it complete?
| |
| 18:03 | <Mip5> It's at 92% now - hopefully it will
| |
| 18:04 | ah, finally! It's just completed. I was going to run updates. Is there something else I should try to fix first?
| |
| 18:05 | <sbalneav> There are about 5000 things that could be slowing it down.
| |
| 18:06 | <Mip5> yeah, I figured. I've never experienced this before. I've been looking at logs for disk errors, but haven't seen anything yet. I'm running software raid, and the disks appear ok. I've tried a reboot, but that didn't help.
| |
| 18:07 | <sbalneav> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=576649
| |
| 18:07 | a few suggestions in there.
| |
| 18:07 | <Mip5> Sweet! I'll look at the link - thanks!
| |
| 18:12 | khildin has left IRC (khildin!~khildin@ip-80-236-227-148.dsl.scarlet.be, Quit: I'm gone, bye bye) | |
| 18:31 | Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas) | |
| 18:37 | <alkisg> stgraber, highvoltage, mgariepy: ltsp meeting if you're around
| |
| 18:52 | <Hyperbyte> !learn hackathon as http://wiki.ltsp.org/wiki/Dev:Hackathon_2013.03
| |
| 18:52 | <ltsp> The operation succeeded.
| |
| 18:54 | vmlintu has left IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi, Ping timeout: 260 seconds) | |
| 18:57 | garymc has joined IRC (garymc!~chatzilla@host86-174-143-125.range86-174.btcentralplus.com) | |
| 19:03 | Patches|Work has joined IRC (Patches|Work!~jacob@nol-lo0.fw.nyc2.appnexus.com) | |
| 19:05 | Patches|Work has left IRC (Patches|Work!~jacob@nol-lo0.fw.nyc2.appnexus.com) | |
| 19:10 | <awilliams> sbalneav: Does gadi still come around?
| |
| 19:10 | <alkisg> !seen Gadi
| |
| 19:10 | <ltsp> Gadi was last seen in #ltsp 43 weeks, 0 days, 20 hours, and 34 seconds ago: * Gadi looks around...
| |
| 19:10 | <awilliams> I guess not
| |
| 19:10 | <alkisg> And before that, he hadn't shown up for a year... /me misses Gadi
| |
| 19:11 | <awilliams> yea
| |
| 19:15 | <jammcq> awilliams: Gadi grew up and got a real job
| |
| 19:15 | so we don't get the pleasure of having him around here everyday
| |
| 19:17 | <BrianG2> alkisg thanks for the help
| |
| 19:18 | GodFather has joined IRC (GodFather!~rcc@fw.acurrus.com) | |
| 19:19 | <alkisg> BrianG2: you're welcome
| |
| 19:19 | <BrianG2> I checked the log on the client and it tried to load the intel driver but its not found?
| |
| 19:29 | vmlintu has joined IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi) | |
| 19:32 | <BrianG2> How can I get the driver onto my client
| |
| 19:33 | in opensuse I can not find the xserver...intel rpm
| |
| 19:36 | i check the xorg.conf generated by the older ltsp script /usr/share/ltsp/configure-x.sh on opensuse 11.1
| |
| 19:37 | I tried that on the 12.2 system and logs an error TERM variable not set
| |
| 19:46 | Enslaver has joined IRC (Enslaver!~Enslaver@c-98-196-42-169.hsd1.tx.comcast.net) | |
| 19:51 | anunnaki has left IRC (anunnaki!~anunnaki@c-174-55-35-255.hsd1.pa.comcast.net, Read error: Connection reset by peer) | |
| 19:56 | anunnaki has joined IRC (anunnaki!~anunnaki@c-174-55-35-255.hsd1.pa.comcast.net) | |
| 19:56 | Gremble has left IRC (Gremble!~Ben@cpc35-aztw23-2-0-cust207.18-1.cable.virginmedia.com, Quit: I Leave) | |
| 20:11 | garymc has left IRC (garymc!~chatzilla@host86-174-143-125.range86-174.btcentralplus.com, Ping timeout: 276 seconds) | |
| 20:13 | highvolt1ge has joined IRC (highvolt1ge!~highvolta@8ta-151-104-187.telkomadsl.co.za) | |
| 20:15 | <BrianG2> > did find that xf86..intel should be included in image and is not
| |
| 20:15 | highvoltage has left IRC (highvoltage!~highvolta@ubuntu/member/highvoltage, Ping timeout: 256 seconds) | |
| 20:20 | <Mip5> Hi Folks - I'm still getting "Reading package lists...." taking a long time, but in the main OS, and in the chroot. I don't see any errors in the logs regarding my disks. Any suggestions as to where else I can look? This step used to be really fast.
| |
| 20:32 | <sbalneav> Woot. Just did a test with libpam-sshauth. Password expiry works FLAWLESSLY, *AND*, it logs you in right after you change your password.
| |
| 20:32 | I love it when a plan comes together.
| |
| 20:32 | <Enslaver> +1
| |
| 20:33 | <sbalneav> Mip5: Might I suggest asking this question in your distro of choice channel? Probably get a better response there.
| |
| 20:33 | garymc has joined IRC (garymc!~chatzilla@host86-174-143-125.range86-174.btcentralplus.com) | |
| 20:33 | <jammcq> sbalneav: what means sbalneav rocks?
| |
| 20:34 | <BrianG2> ok, just ran the kiwi-ltsp-setup to add nvidia and intel support now the client keeps cycling with error kernal versions do not match? any thoughts
| |
| 20:34 | <Mip5> okay - I tried to ask in ubuntu, but there were many many others in there, and it seemed like the target audience was more casual, novice desktop users. I'll keep hunting around.
| |
| 20:35 | Have you never seen it yourself? It
| |
| 20:36 | <sbalneav> No, I haven't. Do you have additional sources in your sources.list, or just the standard ones?
| |
| 20:37 | <BrianG2> the kiwi ltsp config has cyberorg name in it, so was hoping he could help
| |
| 20:37 | <sbalneav> cyberorg: ^^
| |
| 20:38 | <BrianG2> ??
| |
| 20:38 | <alkisg> It should be very late at night now in india...
| |
| 20:38 | <Mip5> I do have additional sources, but have had those in there for a while w/out this issue. I have an internal mirror for the standard stuff. I do see md0_raid1 showing up in "top", though not typically with a high load
| |
| 20:39 | <sbalneav> Standard debugging procedure: remove any extras, try without those.
| |
| 20:39 | But again, this isn't really an LTSP issue.
| |
| 20:40 | strace may help you to narrow down where the problem is.
| |
| 20:40 | <Mip5> Thanks - I was wondering whether having it both OS and chroot was "interesting" - I'll also try to remove those package lists, and see what happens. It makes sense to make sure that's not the issue.
| |
| 20:40 | I haven't used "strace" - I'll take a look.
| |
| 20:41 | <sbalneav> http://stackoverflow.com/questions/174942/how-to-use-strace
| |
| 20:41 | <Mip5> ;-) - I just saw that link was heading over there! Thanks
| |
| 20:41 | komunista has left IRC (komunista!~slavko@87.244.209.121, Quit: Leaving.) | |
| 20:46 | Mip5 has left IRC (Mip5!~chatzilla@208.72.71.69) | |
| 20:53 | mikkel has left IRC (mikkel!~mikkel@80-71-132-15.u.parknet.dk, Quit: Leaving) | |
| 20:58 | garymc has left IRC (garymc!~chatzilla@host86-174-143-125.range86-174.btcentralplus.com, Ping timeout: 256 seconds) | |
| 20:58 | vmlintu has left IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi, Ping timeout: 252 seconds) | |
| 20:58 | garymc has joined IRC (garymc!~chatzilla@host86-174-143-125.range86-174.btcentralplus.com) | |
| 20:59 | <alkisg> jammcq: re: to your ltsp-developers question
| |
| 20:59 | What do your dhcpd send for boot filename, when PXEClient isn't there?
| |
| 21:06 | <sbalneav> I may fork libnss-extrausers. One of the things it would be handy to do is if it handled an extra hosts file.
| |
| 21:07 | <alkisg> sbalneav: for what?
| |
| 21:07 | <sbalneav> That way, all of our scripts could use the server name "ltspsrv" for the server name.
| |
| 21:07 | <alkisg> They currently use "server"...
| |
| 21:07 | <sbalneav> Do we update the hosts file?
| |
| 21:07 | <alkisg> Yes, by init-ltsp.d
| |
| 21:07 | <sbalneav> ah, ok, nevermind then.
| |
| 21:09 | <alkisg> sbalneav: note thought that the nbd or nfs server can be different than the "connection" server
| |
| 21:09 | So it might be better to define another name for the connection server name
| |
| 21:10 | appserver or something
| |
| 21:10 | ...and put that on /etc/hosts too
| |
| 21:11 | <Enslaver> So someone help me here, as I'm a bit confused, ltspfs handles the LOCALDEV mounting for server to client mounts, but is it also suppose to handle the mounting of NFS mounts exported by the server?
| |
| 21:12 | lotharn has left IRC (lotharn!~taco@24.154.55.32, Quit: no) | |
| 21:12 | <alkisg> ...no?
| |
| 21:13 | <Enslaver> I didn't think so, so.. what does?
| |
| 21:13 | <alkisg> NFS mounts from the server are mounted by fstab
| |
| 21:13 | Normal init scripts... mountall in ubuntu
| |
| 21:13 | <Enslaver> well, fstab defines the mounts that are needed
| |
| 21:13 | right, so system init files right?
| |
| 21:13 | <alkisg> RIght
| |
| 21:13 | <Enslaver> rc.sysinit on red hat for example
| |
| 21:14 | but we over-ride the init by calling init-ltsp
| |
| 21:14 | <alkisg> Nope
| |
| 21:14 | We just run stuff before it, and then chain to it
| |
| 21:14 | <Enslaver> I see it calls /sbin/init
| |
| 21:14 | <alkisg> Yup.
| |
| 21:15 | <Enslaver> but that doesn't call rc.sysinit
| |
| 21:15 | <alkisg> /sbin/init is what's usually called, when nothing is specified, right?
| |
| 21:16 | So it's supposed to start your initscripts...
| |
| 21:16 | highvolt1ge has left IRC (highvolt1ge!~highvolta@8ta-151-104-187.telkomadsl.co.za, Ping timeout: 256 seconds) | |
| 21:16 | <alkisg> Just an idea, if you don't get things on /etc/fstab mounted... maybe systemd or whatever else you're using, waits for the network to be available, and it doesn't receive an "if up" event, because the network connection is static, by the initramfs?
| |
| 21:16 | Do other services run?
| |
| 21:17 | <Enslaver> yes
| |
| 21:17 | but not all of them
| |
| 21:17 | for instance the prompt doesn't get set, which is in /etc/bashrc
| |
| 21:18 | but sshd starts
| |
| 21:18 | yet it doesn't nfs mount /home
| |
| 21:19 | <alkisg> So some events are missing, maybe the ones caused by the network being up
| |
| 21:19 | In older ubuntu/ltsp releases, there was no if-up event because of the "manual" eth connection in /etc/network/interfaces
| |
| 21:19 | If I remember well, now we "fake" such an event...
| |
| 21:20 | <Enslaver> where?
| |
| 21:20 | lotharn has joined IRC (lotharn!~taco@24.154.55.32) | |
| 21:20 | <sbalneav> http://wiki.ltsp.org/wiki/Dev:LTSPPamNotes updated. Should be some information there for people to nose around with.
| |
| 21:20 | <alkisg> Enslaver: Can't remember off-hand
| |
| 21:21 | <Enslaver> anything you can remember from it that i can grep for?
| |
| 21:21 | <alkisg> Enslaver: what are you using, systemd?
| |
| 21:21 | <Enslaver> init.d
| |
| 21:23 | chkconfig controlled initscripts
| |
| 21:23 | <alkisg> Ah now I remember... in previous releases ltsp-build-client was deleting all the services except for some whitelisted ones
| |
| 21:23 | Then we moved to blacklisting instead, and we kept the network up init script
| |
| 21:24 | <Enslaver> ok i'll look through my plugins
| |
| 21:24 | <alkisg> Or directly check your chroot services with chkconfig
| |
| 21:27 | <Enslaver> nfs - off, nfslock off, network off, rpcbind off, netfs off
| |
| 21:27 | maybe one of them
| |
| 21:28 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.) | |
| 21:28 | GodFather has left IRC (GodFather!~rcc@fw.acurrus.com, Quit: Leaving) | |
| 21:29 | <Enslaver> I think i'll do a check to look for the server:mount point and enable services based on that
| |
| 21:34 | garymc has left IRC (garymc!~chatzilla@host86-174-143-125.range86-174.btcentralplus.com, Quit: ChatZilla 0.9.90 [Firefox 18.0.2/20130201065344]) | |
| 21:37 | <Enslaver> alkisg: know anything about ltspfs? http://snag.gy/yZjdM.jpg
| |
| 21:38 | scratch that
| |
| 21:38 | 0444 perms might be the cause :)
| |
| 21:45 | <sbalneav> By the way, there's a handy little utility I worked up as part of libpam-sshauth: waitfor. Can wait for either a file's appearance, or disappearance. Waits in .2 second increments. Also has a timeout option.
| |
| 21:46 | We might find it userful to avoid some "sleep 5"'s we've had in the past.
| |
| 21:46 | <jammcq> sbalneav: what's it waiting for?
| |
| 21:46 | vmlintu has joined IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi) | |
| 21:49 | <Enslaver> like xatomwait?
| |
| 21:49 | <sbalneav> just waits on a file's appearance or disappearance.
| |
| 21:49 | I used it in the session script to wait until the socket appeared from the ssh connection
| |
| 21:49 | <jammcq> Linux has a very nice api for detecting the file
| |
| 21:50 | dnotify or soemthing like that
| |
| 21:50 | <sbalneav> http://bazaar.launchpad.net/~sbalneav/ltsp/ltsp-pam-examples/view/head:/ltsp-pam/ltsp-session
| |
| 21:51 | yeah, this is a little simpler. But it's lightwight. Compiled, it's only 5k, and doesn't depend on any other libs :)
| |
| 21:52 | http://bazaar.launchpad.net/~ltsp-upstream/ltsp/libpam-sshauth/view/head:/src/waitfor.c
| |
| 22:13 | adrianorg has left IRC (adrianorg!~adrianorg@177.132.221.167, Read error: Connection reset by peer) | |
| 22:13 | adrianorg_ has joined IRC (adrianorg_!~adrianorg@177.156.59.36) | |
| 22:34 | vagrantc has joined IRC (vagrantc!~vagrant@75-150-46-245-Oregon.hfc.comcastbusiness.net) | |
| 22:34 | vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc) | |
| 22:45 | vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Quit: leaving) | |
| 23:00 | designbybeck has left IRC (designbybeck!~quassel@x179y132.angelo.edu, Remote host closed the connection) | |
| 23:03 | hachque has left IRC (hachque!quassel@2600:3c01::f03c:91ff:fe96:5060, Quit: No Ping reply in 180 seconds.) | |
| 23:03 | hachque has joined IRC (hachque!quassel@2600:3c01::f03c:91ff:fe96:5060) | |
| 23:06 | dead_inside has left IRC (dead_inside!~dead_insi@76.75.3.174, Ping timeout: 252 seconds) | |
| 23:06 | vmlintu has left IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi, Ping timeout: 255 seconds) | |
| 23:22 | bobby_C has left IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at, Ping timeout: 245 seconds) | |
| 23:29 | anunnaki has left IRC (anunnaki!~anunnaki@c-174-55-35-255.hsd1.pa.comcast.net, Read error: Connection reset by peer) | |
| 23:30 | anunnaki has joined IRC (anunnaki!~anunnaki@174.55.35.255) | |
| 23:40 | BrianG2 has left IRC (BrianG2!~brian@173-8-77-89-Sacramento.hfc.comcastbusiness.net, Quit: Konversation terminated!) | |