00:07 | ggsict (~GRAVESEND@194.80.20.10) joined #ltsp. | |
00:46 | alkisg (~alkisg@ubuntu/member/alkisg) joined #ltsp. | |
01:06 | <alkisg> I'm using dnsmasq as a DHCP server on both of my 2 NICs, on different subnets. I only want to provide "dhcp-boot" (i.e. the boot filename) on one of those two NICs. Anyone knows the syntax for that?
| |
01:07 | <muppis> Just for one nic?
| |
01:07 | <alkisg> Yup.
| |
01:08 | I.e. I want to be a dhcp server for both, but a boot server only for one of them
| |
01:09 | <muppis> And you need same IP pool for both?
| |
01:10 | <alkisg> No
| |
01:10 | E.g. one has 10.160.31.x, the other 192.168.0.x
| |
01:11 | <muppis> Then can be easier make two different settings, one per nic.
| |
01:12 | Ah. Tagging is the key.
| |
01:13 | <alkisg> Yes, but I don't know how to tag based on the incoming nic
| |
01:13 | E.g. dhcp-boot=net:192.168.0.x,pxelinux.0
| |
01:13 | I don't know how to set that net: tag
| |
01:15 | <muppis> This is how I used tagging in dnsmasq. Hope it give a hint. http://pastebin.com/bKLXik3c
| |
01:16 | And works same way for nics.
| |
01:22 | <alkisg> That requires me to declare the macs of all clients, no/
| |
01:22 | ?
| |
01:23 | <muppis> No, just tag for nic in dhcp-range -option.
| |
01:25 | Like dhcp-range=interface:eth1,net:boot, ...
| |
01:26 | dhcp-boot=net:boot,/ltsp..
| |
01:27 | <alkisg> Ah I can tag for interface?!
| |
01:27 | <muppis> Sure.
| |
01:27 | <alkisg> That should solve my problem then: dhcp-boot=interface:eth1,/pxelinux.0
| |
01:28 | Doesn't that send the boot filename only for eth1?
| |
01:29 | <muppis> Yes, but i think you can't use interface in dhcp-boot. Use tag set in dhcp-range, like I put above.
| |
01:29 | net:boot is that tag I used in example.
| |
01:31 | <alkisg> Thanks!
| |
01:31 | <muppis> Works?
| |
01:31 | <alkisg> It's for another teacher, I don't have the setup here
| |
01:31 | <muppis> Ok.
| |
01:31 | <alkisg> Will try it in a while though
| |
01:54 | jkinn23 (~adult1@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) joined #ltsp. | |
01:58 | <jkinn23> hi - can anyone help me with shutdown_time option in lts.conf?
| |
01:58 | <muppis> Just ask. Maybe someone knows something.
| |
02:00 | <jkinn23> I have had no luck making it work
| |
02:07 | is there additional configuring i have to do besides adding a line for it in lts.conf?
| |
02:07 | thanks
| |
02:12 | jkinn23 (~adult1@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) left irc: Ping timeout: 240 seconds | |
02:13 | jkinn23 (~adult1@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) joined #ltsp. | |
02:22 | <alkisg> Do you get an entry for shutdown in /etc/cron.d/ltsp on the client?
| |
02:23 | Trixboxer (~Trixboxer@office.supportdepartment.net) joined #ltsp. | |
02:29 | bobby_C (~bobby@85-124-22-227.teleworker.xdsl-line.inode.at) joined #ltsp. | |
02:30 | Da-Geek (~Da-Geek@nat/redhat/x-djqaljpdbhkqqpnk) joined #ltsp. | |
02:31 | <jkinn23> brb...will find out
| |
02:47 | sorry for the delay...i'm working with a new ltsp install. I added shutdown time option and now client hangs at loading pxelinux.cfg
| |
02:48 | but there is also no entry in the client i386 directory on server
| |
02:52 | jkinn2 (~jason@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) joined #ltsp. | |
02:52 | jkinn23 (~adult1@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) left irc: Quit: Leaving | |
02:54 | <alkisg> jkinn2: modifying lts.conf shouldn't affect client loading pxelinux.cfg at any way
| |
02:55 | So let's get this from the start. You're having dhcp/tftp problems, and you didn't have those problems before?
| |
02:55 | <jkinn2> not a minute ago...i just rebooted the server and it's working again
| |
02:56 | but the client just shutdown before reaching login screen...i must be on the right track :)
| |
02:58 | <alkisg> There are 2 places for shutdown, one is on the login screen and one is the cron job
| |
02:59 | So yup if it shut down just before the login screen, then I01-halt-check worked.
| |
02:59 | <jkinn2> cool...i have not had this work for the last week that i have been trying it :)
| |
03:00 | if i now change the shutdown option to be under a mac address heading instead of default, it should still work right?
| |
03:08 | dobber (~dobber@89.190.197.77) joined #ltsp. | |
03:19 | <jkinn2> ok, so after a week and attempting shutdown on 3 different servers it is now working....you must be magic alkisg...thanks!
| |
03:27 | * alkisg didn't provide any answers, only questions :) | |
03:28 | <jkinn2> hence the magic ;)
| |
03:29 | i must be dealing with funky hardware or something...i haven't done anything different except change server box
| |
03:49 | Nick change: zz_evil_root -> evil_root | |
05:29 | Da-Geek (~Da-Geek@nat/redhat/x-djqaljpdbhkqqpnk) left irc: Ping timeout: 240 seconds | |
05:30 | Da-Geek (~Da-Geek@nat/redhat/x-fnczhrgnlncevynd) joined #ltsp. | |
06:06 | cyberorg (~cyberorg@opensuse/member/Cyberorg) left irc: Remote host closed the connection | |
06:30 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) joined #ltsp. | |
06:36 | msignor (~test@24.115.67.19.res-cmts.sesp.ptd.net) joined #ltsp. | |
06:36 | <msignor> Anyone with experience installing on debian?
| |
06:40 | litlebuda (~litlebuda@144.178.108.93.rev.vodafone.pt) joined #ltsp. | |
06:47 | roasted_ (~Jason@c-174-54-217-48.hsd1.pa.comcast.net) joined #ltsp. | |
06:47 | <roasted_> what is the *absolute* minimal RAM you guys would recommend for a lab of 30 systems?
| |
06:47 | on the server that is
| |
06:54 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) left irc: Ping timeout: 255 seconds | |
06:55 | Nick change: ogra_ -> ogra | |
06:57 | <sep> absolute mimium is kinda motherboard dependant but i would not even consider less then 4GB ram
| |
07:00 | <alkisg> roasted_: minimum for what usage? thin or fat clients? localapps or not?
| |
07:00 | All users working at once? Many parameters, they make the answer relative...
| |
07:00 | befrenchy (8e2e1f1e@gateway/web/freenode/ip.142.46.31.30) joined #ltsp. | |
07:01 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) joined #ltsp. | |
07:02 | <sep> my absolute minimum is for 30 machines sitting in the login screen without any users, still 4GB , since it's pointless to buy less...
| |
07:03 | <KBme> yeah, ram is dirt cheap
| |
07:04 | <alkisg> For 30 machines on the login screen without any users, the server shouldn't be spending any RAM on them
| |
07:04 | The RAM is used after the users log on
| |
07:05 | <sep> i know. it would be 4 GB idle. but it's still the least i would put in a server today
| |
07:05 | <KBme> his point is that it's hard to find a machine with less than 4Gb of ram
| |
07:05 | <sep> or the absolute minimum as he asked for :P
| |
07:05 | <alkisg> :)
| |
07:16 | Kicer86 (~Kicer86@host-5db0eeee.sileman.net.pl) joined #ltsp. | |
07:16 | <befrenchy> morning everyone, I have another newbie question this morning regarding the client configuration
| |
07:17 | got my setup working on ubuntu/debian, works great but it looks like the client downloads the LTS image and processing is done entirely on the client
| |
07:18 | what do i need to do to have the client behave as a thin client and run off the server instead?
| |
07:19 | <msignor> befrenchy: are you running an ubuntu distro or a debian distro (i know ubuntu is debian based)
| |
07:19 | <befrenchy> msignor: right now i run it on Ubuntu
| |
07:20 | also had a debian setup
| |
07:20 | <alkisg> Open a terminal. Does it say "befrenchy@server" or "befrenchy@ltsp123" ?
| |
07:20 | <befrenchy> msignor: but i found Ubuntu easier to setup
| |
07:20 | <msignor> it is my understanding that the default behavior is to use an image in Ubuntu. however, you should have to specify on what clients you want to be a "fatclient" so processing is done locally instead of the server.. by default it is thin client
| |
07:20 | i am trying on debian now, which is NFS and a chroot.. and its not been so easy :-)
| |
07:21 | <befrenchy> alkisg: checking
| |
07:22 | alkisg: root@ltsp20
| |
07:23 | <alkisg> befrenchy: and that's inside your gnome session?
| |
07:23 | <befrenchy> alkisg: no just a shell
| |
07:23 | <alkisg> Open a gnome-terminal
| |
07:23 | And see what it says
| |
07:24 | <befrenchy> alkisg: same
| |
07:24 | <alkisg> So, on the client, you have a gnome session, you open gnome-terminal, and it says "root@ltsp20"?
| |
07:25 | <befrenchy> alkisg: yes
| |
07:25 | <alkisg> OK, on that terminal, type: ls /etc/ltsp_*
| |
07:25 | What's the output of that command?
| |
07:25 | <befrenchy> alkisg: i should clarify, client starts off in a shell, i type xinit
| |
07:25 | <alkisg> Right
| |
07:26 | <befrenchy> alkisg: opens up a terminal windows
| |
07:26 | <alkisg> Why do you do it this way?
| |
07:26 | And not let ldm log you on the server?
| |
07:27 | <befrenchy> alkisg: never thought about that, would be just a matter of modifying the lts.conf to start ldm, right?
| |
07:27 | <alkisg> Well the default installation gives you what I'm saying
| |
07:27 | So if you get a shell without messing with the default installation, then you have something broken
| |
07:28 | <befrenchy> alkisg: i modified the lts.conf to start the shell , just for troubleshooting my app
| |
07:28 | <alkisg> You have some problem with an app?
| |
07:29 | <befrenchy> alkisg: not anymore, app issue has been resolved
| |
07:30 | <alkisg> OK, then restore lts.conf so that you log on to the server instead of running things locally on the client?
| |
07:30 | <befrenchy> alkisg: modified the lts.conf to start ldm
| |
07:30 | alkisg: reloading the client
| |
07:30 | <alkisg> Just leave it empty there
| |
07:31 | No need to specify "ldm", it's the default
| |
07:31 | <befrenchy> alkisg: ok
| |
07:31 | alkisg: have the logon screen
| |
07:33 | alkisg: one sec, still trying to logon.
| |
07:34 | alkisg: all i get though is a terminal window after login
| |
07:34 | <alkisg> Was it ever working this way?
| |
07:34 | <msignor> ^ when I had that I needed to install my desktop environment in the chroot
| |
07:34 | <alkisg> Nonono don't do that
| |
07:34 | The desktop environment goes on the server
| |
07:34 | Unless you want a fat client
| |
07:34 | * msignor hides | |
07:35 | <befrenchy> alkisg: ok, cleared the lts.conf, trying again
| |
07:35 | <alkisg> befrenchy: you do have some desktop environment on the server, right?
| |
07:36 | E.g. ubuntu-desktop
| |
07:36 | <msignor> alkisg: i too am having some trouble.. in my case I want a fat client... but keep having NFS issues. so i started over.. Question is: is LDM needed for fat client? just by installing ltsp-server got me a login screen, but when i try to login it just kicks me out again
| |
07:36 | <befrenchy> alkisg: no , i didn't load the desktop env. because this was going to be used with vmware-view client connecting to VMWare virtual machine
| |
07:37 | alkisg: i don't have a need for the desktop env.
| |
07:37 | <alkisg> msignor: fat clients on debian need some manual fiddling. In ubuntu it's more out of the box for now
| |
07:37 | befrenchy: then you don't need to run things on the server
| |
07:37 | So I don't get your first question...
| |
07:37 | <befrenchy> alkisg: in one case i do
| |
07:38 | alkisg: we have older thin clients (wyse) which don't seem to have enough resources to load my client app
| |
07:38 | <alkisg> I think we need to get one step back, and for you to explain the whole situation before we go on solving things that I don't understand :)
| |
07:38 | <befrenchy> alkisg: that's why i would like to run it off the server then start vmware-view from there
| |
07:39 | alkisg: ok
| |
07:39 | alkisg: what I'm doing is running virtual desktops on Vmware
| |
07:39 | <alkisg> So "vmware view" is "your app" that you were talking about?
| |
07:39 | <befrenchy> alkisg: yes
| |
07:39 | <alkisg> And where is the image?
| |
07:40 | On the server? Or are you connecting remotely somehow?
| |
07:40 | <befrenchy> alkisg: i want to re-use old desktops with PXE boot no hard drive to connect to LTS and launch Vmware -view
| |
07:40 | <alkisg> (never seen vmware view)
| |
07:40 | <befrenchy> alkisg: like rdesktop
| |
07:40 | alkisg: so that works on desktops
| |
07:40 | <alkisg> So it doesn't need access to a .vmdk file?
| |
07:41 | <befrenchy> alkisg: nope, connects to vmware view connection server, based on who logs on, its given the appropriate desktop
| |
07:41 | <alkisg> OK
| |
07:41 | <befrenchy> alkisg: i noticed that if i reboot the server once the client is loaded, it doesn't affect the client in any way
| |
07:41 | <alkisg> So you need that (1) locally for good clients and (2) on the server for old clients?
| |
07:42 | <befrenchy> alkisg: that tells me image loads locally
| |
07:42 | alkisg: exactly :)
| |
07:42 | alkisg: make sense?
| |
07:42 | <alkisg> OK, then you'll need to create a custom screen script for vmware-view, and to only specify that SCREEN_xx for the new clients
| |
07:43 | <befrenchy> alkisg: that part is donw
| |
07:43 | <alkisg> I.e. in lts.conf:
| |
07:43 | [mac:address:of:new:client]
| |
07:43 | SCREEN_07=vmware-view
| |
07:43 | Did you do it with a screen script? Or with some other way?
| |
07:43 | <befrenchy> alkisg: yes
| |
07:43 | <alkisg> Because if you did it some other way that breaks ldm, then... :)
| |
07:43 | <befrenchy> alkisg: any way to group the macs?
| |
07:44 | <alkisg> [mac:address*] works
| |
07:44 | And there's also a "LIKE" clause
| |
07:44 | <befrenchy> alkisg: or do i need to specify each device's mac address in the lts.conf
| |
07:44 | <alkisg> See the star above (*)
| |
07:44 | <befrenchy> alkisg: ok
| |
07:44 | <alkisg> You can also make a script that auto-sets the SCREEN based on the client RAM
| |
07:45 | <befrenchy> alkisg: ok , that's good too
| |
07:45 | alkisg: will look into it
| |
07:45 | <alkisg> That's what I do for fat clients, e.g. > 500 => fat, less than 500 => thin
| |
07:45 | <befrenchy> alkisg: ok now, how do you specify whether it is thin or fat?
| |
07:46 | <alkisg> I don't need to, it's done automatically by the script in the chroot
| |
07:46 | But if I want something special, I can use FAT_RAM_THRESHOLD or LTSP_FATCLIENT in lts.conf
| |
07:46 | <befrenchy> alkisg: hmmm
| |
07:47 | alkisg: not sure if i understand, so how do you tell the client to load on the server?
| |
07:47 | <alkisg> Fat clients need a specially-prepared chroot: https://help.ubuntu.com/community/UbuntuLTSP/FatClients
| |
07:48 | And then they automatically boot as thin or fat, based on the client RAM
| |
07:48 | <befrenchy> alkisg: so it's a separate image then?
| |
07:48 | <alkisg> The same image boots both thin and fat clients
| |
07:49 | <befrenchy> alkisg: does it explain how to do it in this link?
| |
07:49 | <alkisg> Yes, but you don't need it
| |
07:49 | You only need a custom screen script + a localapp
| |
07:50 | <befrenchy> alkisg: so are you saying the lts.conf would be the same for both client types
| |
07:50 | <alkisg> If you use an autodetecting script on the chroot, then you won't need any lts.conf
| |
07:50 | If you don't use such a script, then you'll need different entries for the 2 client types
| |
07:52 | <msignor> alkisg: Can you provide me some suggestions or advice? -- Long story short is I have an embedded box (600mhz) doing torrents, fileserver, etc etc and a separate router doing dns/pxe/etc... I picked up some thin clients with DVI really cheap and my intention was to turn them into "cable boxes" so I could watch downloaded content as well as surf the net in the different rooms in my house. point being bookmarks, preferences, favorites etc are
| |
07:52 | kept regardless of where someone logs in -- fat client seemed to be the best fit given my server hardware.. i don't know if its worth it for me to keep a desktop running 24x7 running ubuntu just to have this capability... thats the part im struggling with.. the least amount of bloat.. should I just cave at this point and forget trying to get this working on debian?
| |
07:52 | Roasted (~jason@206.82.22.190) joined #ltsp. | |
07:53 | ggsict (~GRAVESEND@194.80.20.10) left irc: Quit: Leaving | |
07:53 | <alkisg> msignor: thin client RAM?
| |
07:54 | (specs)
| |
07:54 | <msignor> core duo, 1GB ram
| |
07:54 | not core 2 duo (they are those aopen mac mini clones)
| |
07:54 | some store in town used a POS system with it, and ripped the hard drives out when they closed
| |
07:55 | <alkisg> The embedded box can serve nfs, and your router can serve dhcp/tftp etc? Then I don't see why you'd need a separate server up and running 24/7
| |
07:55 | <befrenchy> alkisg: any docs available for setting up an autodetecting script?
| |
07:55 | <alkisg> Those 2 should do it
| |
07:55 | Roasted (~jason@206.82.22.190) left irc: Client Quit | |
07:56 | <alkisg> befrenchy: I don't think there are, so best thing would be to look at the existing code.
| |
07:56 | Roasted (~jason@206.82.22.190) joined #ltsp. | |
07:56 | <Roasted> hey guys
| |
07:56 | <befrenchy> alkisg: ok, well i know the vmware view client points to a script
| |
07:56 | <msignor> alkisg: exactly.. the embedded box is booting off a 1GB cf card - running a lean debian squeeze called voyage linux.. i mount a SSD that stores the CHroot.. I definately can't run ubuntu server on it
| |
07:56 | <Roasted> I'm trying to figure out how much RAM I would need in my server to handle a lab of 30. What was the common calculations for figuring this out?
| |
07:56 | <befrenchy> alkisg: set to SCREEN_03=vmview
| |
07:57 | alkisg: and then that points to the script
| |
07:57 | <alkisg> Roasted: we answered above
| |
07:57 | <befrenchy> alkisg: u saying i have to modify that script?
| |
07:57 | <Roasted> alkisg, oh, you did? The answer would reside on my other computer then, miles away :(
| |
07:57 | I just now joined on my laptop at work.
| |
07:57 | <alkisg> See the topic for the logs
| |
07:58 | <befrenchy> alkisg: i can show you the content of the script if you want
| |
07:58 | <Roasted> mucho thanks alkisg
| |
07:58 | So 4gb RAM minimal. Operationally speaking should I double that, triple?
| |
07:59 | cyberorg (~cyberorg@opensuse/member/Cyberorg) joined #ltsp. | |
08:00 | <befrenchy> alkisg: here is the content of the vmview script: http://pastebin.com/r2U3g13v
| |
08:03 | <alkisg> msignor: for fat clients you just need nfs + ssh. Does voyage linux have nfs + ssh?
| |
08:03 | You don't need to have the same distro on the server as the chroot
| |
08:04 | <msignor> yes
| |
08:04 | i am upto the point where I was logging in, xfce4 was booting, but nfs wasnt working.. found out i had to install nfs-common in the chroot, and when I did.. now it hangs at boot where it says "setting up ltsp please wait"
| |
08:05 | prior to me installing nfs-common in the chroot it just errord out.
| |
08:06 | alkish: figuring I borked it i started from scratch last night and only installed ltsp server as of now, built a base chroot with xfce4 in it and nfs-common. and when i get to the login screen, i login and it kicks me out. AND i still have all the same errors about permission denied to NFS even though I can map the shares on other machines
| |
08:07 | <alkisg> msignor: one quick way would be to build an ubuntu fat chroot and export it through your debian server
| |
08:07 | For the other solutions you'd need troubleshooting, as fat clients on debian aren't yet automated
| |
08:11 | <msignor> alkisg: could i build the ubuntu chroot inside debian? I assume its: something like "ltsp-build-client 9.04" or something the like.. I did notice that --fat-client is not a valid option for ltsp-build-client
| |
08:12 | alkisg: would an ubuntu chroot then use the images? or would it be off NFS? that is apoint of confusion
| |
08:13 | <alkisg> The default for ubuntu is nbd (==needs nbd-server and inetd), which makes fat clients about 2 times faster
| |
08:13 | (for network disk performance, due to compression etc)
| |
08:13 | But you can also change that to use nfs
| |
08:13 | <Roasted> so I guess there's no operationally ideal amount of ram to use then?
| |
08:14 | <alkisg> Roasted: there's a formula in the docs, see the topic for the link
| |
08:15 | komunista (~slavko@adsl-195-098-015-248.dynamic.nextra.sk) joined #ltsp. | |
08:21 | <Roasted> alkisg, thanks. I guess I'm just trying to get a target idea of what user opinions are that have been in these shoes, though. Like a formula is great but how it works in production is what I'm after. I hate to follow the formula and find out we bottleneck at 20 when users are multitasking as opposed to 30. Yet due to budgeting going insane I have to make SURE what we get is what we NEED.
| |
08:21 | staffencasa (~staffenca@128-193-151-235.oregonstate.edu) joined #ltsp. | |
08:22 | <befrenchy> alkisg: isn't the default behavior of the ltsp-build-client to create a fatclient image?
| |
08:22 | <alkisg> No
| |
08:23 | <befrenchy> alkisg: hmm, well i didn't specify any options when i ran the command
| |
08:23 | <alkisg> You don't have a fat chroot
| |
08:23 | <befrenchy> alkisg:i don't?
| |
08:23 | <alkisg> You just have one localapp, vmware view
| |
08:23 | No
| |
08:24 | (04:25:27 PM) alkisg: OK, on that terminal, type: ls /etc/ltsp_*
| |
08:24 | (04:25:39 PM) alkisg: What's the output of that command?
| |
08:24 | <befrenchy> alkisg: but if i reboot the server my client is unaffected, so i figured it must be fat
| |
08:24 | <alkisg> befrenchy: if you did that, you'd see that you don't have a fat chroot
| |
08:24 | Fat chroots have an /etc/ltsp_fat_chroot file
| |
08:25 | <befrenchy> alkisg: so how does it work then? shouldn't my session freeze up or get disconnected when i reboot my server then?
| |
08:25 | <Roasted> well I guess I'll just get a ton of ram then and see what happens
| |
08:26 | <befrenchy> alkisg: everything seems to run local on the client
| |
08:26 | <alkisg> befrenchy: with your xinit procedure? Sure
| |
08:26 | With the default ldm login? Nope
| |
08:26 | MorningSon (~MorningSo@cpe-70-114-21-95.satx.res.rr.com) joined #ltsp. | |
08:26 | <alkisg> As for the server rebooting, either thin or fat, the client disconnects
| |
08:27 | But nbd-proxy tells it to reconnect when the server comes up again
| |
08:27 | So if you're lucky and if the server didn't stay closed for long, you may continue working
| |
08:27 | If you logged on to ldm though, the ssh connection would break
| |
08:27 | <Roasted> It would re-establish though wouldn't it?/
| |
08:27 | <alkisg> So for thin clients their serever session would break
| |
08:27 | No, it wouldn't
| |
08:27 | <Roasted> I've rebooted my test server with a client at login screen and it seems to mesh back up
| |
08:28 | <befrenchy> alkisg: so is the script the best way to set this up ?
| |
08:28 | <alkisg> Roasted: no ssh connection at this point
| |
08:28 | Ssh connection comes after logon
| |
08:28 | (exactly at logon)
| |
08:28 | <Roasted> ah, I see.
| |
08:28 | I was going to say, if I let the client sit it would re-establish and I could log in fine afterwards.
| |
08:29 | <alkisg> That's nbd.
| |
08:30 | <befrenchy> alkisg: should i then rebuild the client image with the -- fatclient option and use LTSP_FATCLIENT=false in my lts.conf file for older thin clients?
| |
08:31 | alkisg: just wondering what works better
| |
08:36 | <alkisg> befrenchy: no, you don't need fat clients
| |
08:37 | You only need [a custom screen script OR an autodetection script] AND [vmware-view as localapp]
| |
08:37 | <Roasted> alkisg, in this documentation when it talks about processor speed and what's recommended, it never mentions cores. Is this assuming it's a single core?
| |
08:38 | <befrenchy> alkisg: i guess i'm still unsure how to how to tell the client in the script whether is should be thin or fat
| |
08:38 | <alkisg> befrenchy: if ram > 500 then SCREEN_07=vmware
| |
08:39 | And maybe you should stop using the name "fat"
| |
08:39 | Because it'll get you confused
| |
08:39 | You won't have any fat clients in your setup.
| |
08:39 | <befrenchy> alkisg: so this change is made in the lts.conf?
| |
08:39 | <alkisg> No, that was the autodetection script
| |
08:40 | So if you put such a script you don't need an lts.conf at all
| |
08:40 | You just put a file in the ltsp_config.d directory in the chroot
| |
08:40 | <befrenchy> alkisg: ok let me read up on this
| |
08:40 | alkisg: thanks for your help once again
| |
08:40 | :)
| |
08:40 | <alkisg> You're welcome
| |
08:44 | <msignor> alkisg: just wanted to clarify on this - even though i make a chroot that is ubuntu
| |
08:44 | <Roasted> would anybody else have any idea?
| |
08:44 | <msignor> it will NOT be using NFS ex: I mist do ltsp -rebuild - client
| |
08:46 | <alkisg> msignor: The default for ubuntu chroots is to use NBD, not NFS. But you will want to use nfs for /home.
| |
08:48 | artista_frustrad (~artista_f@187.7.151.137) joined #ltsp. | |
08:49 | <Roasted> suppose not
| |
08:49 | shawnp0wers (~Adium@linuxjournal/staff/shawnp0wers) joined #ltsp. | |
08:54 | drdata (~drdata@pool-173-57-189-158.dllstx.fios.verizon.net) left irc: Ping timeout: 255 seconds | |
08:56 | <msignor> alkisg so I can just edit the fstab in the chroot to mount home properly correct? What would be the proper syntax? I found the debian ltsp client didin't support --fat-client.. could you suggest what I should do to build the chroot?
| |
08:57 | <alkisg> msignor: in Ubuntu fat clients you can share home by just specifying NFS_HOME=/home in lts.conf
| |
08:57 | I don't know how you can build an ubuntu chroot from debian. Install an ubuntu server in a vm?
| |
09:01 | <msignor> can i just copy paste the chroot? literally? I have an ubuntu desktop as my laptop atm...
| |
09:02 | <alkisg> Yes
| |
09:02 | <msignor> alkisg: sounds like your solution is a better fit to the problem and me having to add every freaking app to xfce by hand
| |
09:03 | is the client build a standalone package or does it depend on ltsp-server?
| |
09:04 | alkisg: do I even need LTSP-server installed on the server? if I am using the NBD image and fat client? (aside from wanting to rebuild a chroot image)
| |
09:04 | and finally - do i need to worry about ssh keys or anything being generated inproperly?
| |
09:05 | <alkisg> For a home setup you can just tell the client to ignore ssh keys completely
| |
09:05 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) left irc: Ping timeout: 240 seconds | |
09:05 | <alkisg> ltsp-server isn't a heavy package. You can have it installed both on your laptop and on your debian server, it doesn't affect ram/performance
| |
09:21 | <msignor> ok, cool - the ubuntu build client supports the fat-client switch, i assume I should use that over the LTSP_FATCLIENT=True?
| |
09:21 | <alkisg> It's not an option, you do need the --fat-client switch
| |
09:22 | <msignor> hrmn. ok - the squeeze packaga didn't support it. i read that that setting in LTS.conf was the same thing
| |
09:22 | Roasted (~jason@206.82.22.190) left irc: Ping timeout: 255 seconds | |
09:23 | <msignor> good to know it wasn't going to work no matter what I did, lol.
| |
09:24 | <alkisg> It would work if you manually did the same things that the fat client plugin does
| |
09:24 | <msignor> alkisg: ldm-server is required on the server correct?
| |
09:24 | <alkisg> I think it's optional
| |
09:24 | <msignor> ok, just checking i saw that on ubuntu it installed it as default, on debian it didn't
| |
09:27 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) joined #ltsp. | |
09:28 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) left irc: Client Quit | |
09:28 | <msignor> alkisg: i assume the manual tweaking the plugin did is what you were referring to about the tweaks needed for a debian system
| |
09:29 | <alkisg> Yes
| |
09:29 | <msignor> ok - that makes me feel much better, i knew this couldn't be _that_ hard.. i had to be missing something
| |
09:30 | im building the chroot now, lets see how it goes afterwards - on my blazing 1.5m connection
| |
09:32 | klausade (~klaus@cm-84.215.171.127.getinternet.no) left irc: Ping timeout: 240 seconds | |
09:33 | <befrenchy> alkisg: i had my server down for the last 20 minutes and my client running vmware view client is still up and running, completely independent from the server
| |
09:33 | <alkisg> So?
| |
09:33 | Does this mean that it has local storage? :)
| |
09:34 | Try ls -R /
| |
09:34 | <befrenchy> alkisg: perhaps not but i don't really need storage
| |
09:34 | <alkisg> This should make it halt :)
| |
09:34 | Thin of fat doesn't make a difference on when the client hangs
| |
09:35 | It's just a matter of when it needs to read its (absent) network disk
| |
09:35 | <befrenchy> alkisg: when would it do that?
| |
09:36 | <alkisg> E.g. when a cron job runs
| |
09:36 | <befrenchy> alkisg: if all i use it for is vmware view, essentially the same as rdesktop, should it ever need to access the disk for anything?
| |
09:36 | <alkisg> Or when, on a fat client, you try to access the menus
| |
09:36 | Or when it needs to swap something
| |
09:37 | <befrenchy> alkisg: i guess my question is how do i run the vmware view client directly from the server so it uses the server resources instead of my client hardware
| |
09:37 | alkisg: do i need to login via ldm first?
| |
09:37 | <alkisg> You just use a default chroot for that
| |
09:37 | Yes
| |
09:37 | Or in other words, you need SCREEN_07=ldm
| |
09:37 | <befrenchy> anyway to autologin and launch that application automatically?
| |
09:37 | <alkisg> either from lts.conf or from the autodetection script
| |
09:38 | Sure, there's LDM_AUTOLOGIN etc
| |
09:38 | See the docs for autologin
| |
09:38 | <befrenchy> alkisg: ah ha, cool
| |
09:38 | <alkisg> You can also have a custom xsession on the server
| |
09:38 | <befrenchy> alkisg: this should help with the old thin clients devices
| |
09:39 | alkisg: k, will keep playing with it and see if i can figure it out
| |
09:40 | alkisg: but i guess i have to login to ldm, i should probably install the desktop components?
| |
09:41 | <alkisg> That'd be easier for you, yes
| |
09:46 | Roasted (~jason@206.82.22.190) joined #ltsp. | |
09:49 | alkisg (~alkisg@ubuntu/member/alkisg) left irc: Quit: Leaving. | |
10:00 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) joined #ltsp. | |
10:01 | klausade (~klaus@cm-84.215.171.127.getinternet.no) joined #ltsp. | |
10:01 | vvinet (~vince@2001:470:8829:1000:227:eff:fe25:ee64) left irc: Read error: Operation timed out | |
10:09 | Da-Geek (~Da-Geek@nat/redhat/x-fnczhrgnlncevynd) left irc: Quit: Leaving | |
10:14 | Lumiere (~jstraw@pool-74-96-91-120.washdc.fios.verizon.net) left irc: Ping timeout: 250 seconds | |
10:16 | Lumiere (~jstraw@pool-74-96-91-120.washdc.fios.verizon.net) joined #ltsp. | |
10:29 | garymc (~chatzilla@host81-148-15-59.in-addr.btopenworld.com) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.6.15/20110303024001] | |
10:30 | dobber (~dobber@89.190.197.77) left irc: Remote host closed the connection | |
10:54 | litlebuda (~litlebuda@144.178.108.93.rev.vodafone.pt) left irc: Remote host closed the connection | |
10:55 | msignor (~test@24.115.67.19.res-cmts.sesp.ptd.net) left irc: Ping timeout: 252 seconds | |
11:00 | drdata (~drdata@tw2.udallas.edu) joined #ltsp. | |
11:31 | msignor (~test@24.115.67.19.res-cmts.sesp.ptd.net) joined #ltsp. | |
11:45 | andygraybeal (~andy.gray@obsidian.casanueva.com) left irc: Read error: Connection reset by peer | |
11:47 | mistik1 (mistik1@unaffiliated/mistik1) left irc: Read error: Connection reset by peer | |
11:47 | mistik1 (mistik1@unaffiliated/mistik1) joined #ltsp. | |
11:54 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) joined #ltsp. | |
12:01 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) left irc: Quit: I Leave | |
12:13 | Trixboxer (~Trixboxer@office.supportdepartment.net) left irc: Quit: "Achievement is not the end, its the beginning of new journey !!!" | |
12:17 | vagrantc (~vagrant@75-150-46-245-Oregon.hfc.comcastbusiness.net) joined #ltsp. | |
12:25 | <msignor> alkisg: you around still?
| |
12:27 | vmlintu (~vmlintu@nblzone-240-143.nblnetworks.fi) joined #ltsp. | |
12:32 | befrenchy (8e2e1f1e@gateway/web/freenode/ip.142.46.31.30) left irc: Quit: Page closed | |
12:38 | bobberty (~martian_l@124-197-42-208.callplus.net.nz) joined #ltsp. | |
12:39 | <bobberty> hey, how do i tell LTSP's DHCP to listen to the network card for requests?
| |
12:40 | <Appiah> it's not a special DHCP server for LTSP
| |
12:40 | if you use dhcpd check the man
| |
12:42 | <bobberty> yeah, DHCPD, it says 'not configured to listen on any interface, if this is not what you want write a declaration in your DHCPD.conf file' this is not what i want
| |
12:44 | <abeehc> man dhcpd.conf
| |
12:47 | <msignor> anyone know the best way to copy a chroot & its symlinks to an archive so it can be put on a USB stick and moved to another machine?
| |
12:49 | <bobberty> i read the MAN, but it doesn't help me tell it to listen for requests
| |
12:54 | Nick change: evil_root -> zz_evil_root | |
12:54 | <abeehc> what dist are you using
| |
12:54 | bobberty (martian_l@124-197-42-208.callplus.net.nz) left #ltsp. | |
12:54 | <abeehc> have you checked the distrib
| |
12:54 | 's docs?
| |
12:55 | are you using binary packages from whatever dist?
| |
12:59 | alkisg (~alkisg@ubuntu/member/alkisg) joined #ltsp. | |
13:02 | <alkisg> msignor: if you'll be using nbd (faster, needs nbd-server) you only need to copy /opt/ltsp/images/i386.img
| |
13:08 | andygraybeal (~andy.gray@obsidian.casanueva.com) joined #ltsp. | |
13:08 | <msignor> alkisg: i still have the old files from the debian build in tftpboot - do those matter now that this is an ubuntu chroot
| |
13:08 | alkisg: also, do i need to make any special provisions to other files to tell it - it is NBD not NFS anymore?
| |
13:09 | <alkisg> You need the new files in tftp i.e. the new kernel
| |
13:09 | <msignor> ok, cool. thought so
| |
13:09 | <alkisg> If you're using nbd, you need to export the image from /etc/inetd.conf
| |
13:11 | <msignor> so i install the nbd-server package.. then edit the inetd.conf to allow the NBD server to export (like i did with NFS)?
| |
13:14 | <alkisg> Hmm if you only need inetd for nbd-server, then don't install it just for this one
| |
13:14 | Make a config file for nbd-server instead
| |
13:15 | <msignor> ok - well i believe i have inetd for running tftpd-hpa
| |
13:16 | i am honestly not sure
| |
13:16 | <alkisg> In ubuntu now tftp runs standalone
| |
13:16 | Not sure about debian
| |
13:17 | <msignor> yeah, debian has a separate init.d for tftpd-hpa
| |
13:17 | im copying the file to a key so I can move it faster than the wifi im on... i need to google for a sample config
| |
13:21 | bobberty (~martian_l@124-197-42-208.callplus.net.nz) joined #ltsp. | |
13:27 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) joined #ltsp. | |
13:27 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) left irc: Remote host closed the connection | |
13:30 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) joined #ltsp. | |
13:35 | Gremble (~Ben@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com) left irc: Client Quit | |
13:44 | Roasted (~jason@206.82.22.190) left irc: Ping timeout: 255 seconds | |
13:46 | shawnp0wers (~Adium@linuxjournal/staff/shawnp0wers) left irc: Quit: Leaving. | |
14:00 | msignor (~test@24.115.67.19.res-cmts.sesp.ptd.net) left irc: Quit: Leaving | |
14:06 | <muppis> alkisg, got dnsmasq working?
| |
14:17 | Kicer86 (~Kicer86@host-5db0eeee.sileman.net.pl) left irc: Quit: KVIrc Insomnia 4.0.2, revision: 4740, sources date: 20100627, built on: 2010-08-08 18:29:00 UTC http://www.kvirc.net/ | |
14:20 | <vagrantc> bobberty: in order to configure your DHCP server, you'll need to provide your networking information and dhcpd.conf file ...
| |
14:20 | bobberty: the example file shipped with LTSP will work as long as your network settings match.
| |
14:20 | bobberty: you'll also need to make sure there isn't another DHCP server on the same network, or it will cause problems.
| |
14:21 | bobberty: or need to switch to using dnsmasq's DHCP proxy configuration
| |
14:21 | <bobberty> so, er, can u tell me how to get it going?
| |
14:21 | <vagrantc> bobberty: what's your server's IP address?
| |
14:22 | <bobberty> 192.168.0.100
| |
14:22 | no, 1.100
| |
14:22 | that could be the problem
| |
14:23 | <vagrantc> that's certainly the problem.
| |
14:23 | bobberty: so change all occurrances of 192.168.0 to 192.168.1 ...
| |
14:23 | bobberty: also, how did you get your IP address?
| |
14:23 | it makes me suspect there's another DHCP server on the network
| |
14:24 | <bobberty> i only have 'netmask 192.168.0.0'
| |
14:24 | <vagrantc> so change it to 192.168.1.0
| |
14:24 | wait, what?
| |
14:24 | netmask should probably be 255.255.255.0
| |
14:24 | <bobberty> subnet is that
| |
14:25 | * vagrantc has to head back to work soon | |
14:25 | <vagrantc> bobberty: how did you get your ip address?
| |
14:25 | 192.168.1.100
| |
14:25 | ??
| |
14:25 | <bobberty> system>preferences>network connections then eth0, edit connections
| |
14:26 | ipv4
| |
14:26 | <vagrantc> why did you pick 100 ?
| |
14:26 | or did it automatically configure itself?
| |
14:26 | <bobberty> because it sounded simple to remember
| |
14:26 | so where do i change to that?
| |
14:26 | <vagrantc> it will be less painful to use something below 20 ... then you won't have to mess with the range
| |
14:26 | <bobberty> OK, i'll change it
| |
14:27 | <vagrantc> i'd recommend 192.168.1.1
| |
14:27 | given that you're using 192.168.1.*
| |
14:28 | <bobberty> my router's that, and if i change that adress, the internet doesn't work
| |
14:28 | <vagrantc> and then change all the 192.168.0 to 192.168.1 in your dhcpd.conf
| |
14:28 | does your router run DHCP?
| |
14:28 | if so, you're going to have problems.
| |
14:28 | <bobberty> so the range, the subnet?
| |
14:28 | no, i turned it off
| |
14:28 | <vagrantc> ok.
| |
14:29 | so then pick 192.168.1.2 for your server
| |
14:29 | and change your dhcpd.conf to use 192.168.0 to 192.168.1 ...
| |
14:30 | <bobberty> change netmask to 192.168.1.18?
| |
14:30 | <vagrantc> no.
| |
14:30 | netmask should be 255.255.255.0
| |
14:30 | well, good luck, have to get back to work.
| |
14:30 | <bobberty> change what then ,excuse my newbeism
| |
14:30 | ok
| |
14:31 | <vagrantc> change all occurances of 192.168.0. to 192.168.1. ... ALL OF THEM.
| |
14:31 | then set next-server 192.168.1.2
| |
14:31 | in dhcpd.conf
| |
14:31 | <bobberty> range, netmask, subnet, and so on?
| |
14:31 | <vagrantc> your subnet-mask should be 255.255.255.0
| |
14:32 | good luck.
| |
14:32 | <bobberty> yes it works!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
14:32 | :):):):):):)):):):
| |
14:32 | )
| |
14:33 | thank you so much!!!
| |
14:34 | finally
| |
14:38 | komunista (~slavko@adsl-195-098-015-248.dynamic.nextra.sk) left irc: Quit: Leaving. | |
14:53 | alkisg (~alkisg@ubuntu/member/alkisg) left irc: Quit: Leaving. | |
15:12 | msaul_ (~administr@CPE001cf0f564ef-CM0011ae00c1f8.cpe.net.cable.rogers.com) joined #ltsp. | |
15:13 | <msaul_> I haven't been able to ssh into my client from my server - any suggestions?
| |
15:13 | I get "ssh: connect to host 192.168.1.20 port 22: Connection refused"
| |
15:14 | Any help would be greatly appreciated...
| |
15:16 | <bobberty> have you tried changing the port perhaps?
| |
15:16 | <msaul_> so you mean by the ssh command as an option?
| |
15:16 | drdata (~drdata@tw2.udallas.edu) left irc: Ping timeout: 240 seconds | |
15:29 | <bobberty> it was just a suggestion, i don't know about secure shell, but maybe that would fix it
| |
15:30 | <msaul_> I can ping the IP address, but can't connect via ssh... I thought I set the correct iptable policies...
| |
15:30 | thanks anyways...
| |
15:30 | msaul_ (~administr@CPE001cf0f564ef-CM0011ae00c1f8.cpe.net.cable.rogers.com) left irc: Quit: Ex-Chat | |
15:36 | drdata (~drdata@pool-173-57-189-158.dllstx.fios.verizon.net) joined #ltsp. | |
15:38 | cyberorg (~cyberorg@opensuse/member/Cyberorg) left irc: Ping timeout: 255 seconds | |
15:42 | cyberorg (~cyberorg@opensuse/member/Cyberorg) joined #ltsp. | |
15:54 | kelbo (~ops@110.139.58.157) joined #ltsp. | |
16:09 | <vagrantc> msaul_ ssh isn't normally installed on the thin clients
| |
16:09 | apparently gone now... anyways...
| |
16:37 | staffencasa (~staffenca@128-193-151-235.oregonstate.edu) left irc: Quit: Leaving | |
16:51 | bobby_C (~bobby@85-124-22-227.teleworker.xdsl-line.inode.at) left irc: Ping timeout: 260 seconds | |
16:54 | msignor (~msignor@24.115.67.19.res-cmts.sesp.ptd.net) joined #ltsp. | |
17:02 | bobberty (martian_l@124-197-42-208.callplus.net.nz) left #ltsp. | |
17:06 | jkinn2 (~jason@CPE54d46fa9ecb2-CM54d46fa9ecb0.cpe.net.cable.rogers.com) left irc: Ping timeout: 240 seconds | |
17:09 | msignor (~msignor@24.115.67.19.res-cmts.sesp.ptd.net) left irc: Read error: No route to host | |
17:22 | zamba (marius@flage.org) left irc: Read error: Operation timed out | |
17:22 | zamba (marius@flage.org) joined #ltsp. | |
17:29 | msignor (~msignor@24.115.67.19.res-cmts.sesp.ptd.net) joined #ltsp. | |
17:32 | vagrantc (~vagrant@75-150-46-245-Oregon.hfc.comcastbusiness.net) left irc: Ping timeout: 246 seconds | |
18:25 | vagrantc (~vagrant@75-150-46-245-Oregon.hfc.comcastbusiness.net) joined #ltsp. | |
19:32 | vagrantc (~vagrant@75-150-46-245-Oregon.hfc.comcastbusiness.net) left irc: Quit: leaving | |
20:26 | artista_frustrad (~artista_f@187.7.151.137) left irc: Ping timeout: 252 seconds | |
20:38 | artista_frustrad (~artista_f@187.7.151.137) joined #ltsp. | |
20:48 | artista_frustrad (~artista_f@187.7.151.137) left irc: Ping timeout: 240 seconds | |
21:02 | artista_frustrad (~artista_f@187.7.151.137) joined #ltsp. | |
22:14 | alkisg (~alkisg@ubuntu/member/alkisg) joined #ltsp. | |
22:18 | patrickmw (~pwright@ip68-231-4-153.ph.ph.cox.net) left irc: Quit: Ex-Chat | |
22:44 | alkisg (~alkisg@ubuntu/member/alkisg) left irc: Ping timeout: 250 seconds | |
22:48 | cyberorg (~cyberorg@opensuse/member/Cyberorg) left irc: Remote host closed the connection | |
22:52 | MorningSon (~MorningSo@cpe-70-114-21-95.satx.res.rr.com) left irc: Quit: WeeChat 0.3.0 | |
23:46 | Kicer86 (~Kicer86@host-5db0eeee.sileman.net.pl) joined #ltsp. | |
00:00 | --- Thu Mar 17 2011 | |