IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 9 March 2010 (all times are UTC)
| 00:28 | Selveste1 has quit IRC | |
| 00:50 | <alkisg> For fat clients, LOCALDEV needs to be handled in a completely different way than for thin clients. I.e. ltspfs should not be started at all, and polkit-1 policy files should be modified instead.
| |
| 00:52 | I'm thinking of 2 different approaches for that - one is to make a lot of changes in the current code, e.g. `if ! boolean_is_true $LOCALDEV` would become `if ! boolean_is_true $LOCALDEV || boolean_is_true $LTSP_FATCLIENT`,
| |
| 00:53 | and the other one would be to do FAT_LOCALDEV=$LOCALDEV; LOCALDEV=False - so that all scripts think that LOCALDEV is disabled, except for a new client script, that sets up the policy files
| |
| 00:53 | I'm not sure of which of the 2 ways is preferred... stgraber, which would you choose?
| |
| 00:56 | X-Raimo has joined #ltsp | |
| 00:58 | <alkisg> Hmm I'll go for the second one.
| |
| 01:05 | Selveste1 has joined #ltsp | |
| 01:05 | <X-Raimo> hello. can anyone help me with pulseaudio issue http://paste.org.ru/?dyd9jg please? This happens at LTSP Server and client
| |
| 01:16 | daya has quit IRC | |
| 01:25 | * alkisg wonders where's a good place to put a "fix-polkit-1-files" script... | |
| 01:36 | try2free has joined #ltsp | |
| 01:40 | try2free has left #ltsp | |
| 02:10 | daya has joined #ltsp | |
| 02:15 | <ogra> alkisg, what does it do ?
| |
| 02:16 | just cp files to /etc/polkit-1/*.d/ ?
| |
| 02:16 | <alkisg> ogra, I was thinking of modifying things in polkit-1, yeah, but I wonder if that's the right approach
| |
| 02:16 | Maybe it would be best if (1) I fixed ltsp so that the logged on users shows as ACTIVE in PK,
| |
| 02:17 | <ogra> if you put them in the .d dirs i think thats the right approach
| |
| 02:17 | <alkisg> and (2) created a shadow entry to enable the users to get authenticated
| |
| 02:17 | <ogra> i would make an upstart job with a script that checks if you are on a fat client and based on that copies the files or not
| |
| 02:17 | <alkisg> of course, *only* if the admin allows it in lts.conf
| |
| 02:18 | That would make fat clients work like normal ubuntu pcs
| |
| 02:18 | <ogra> does ltsp-client-core expose a dm event for upstart ?
| |
| 02:18 | <alkisg> and, when in the future sbalneav implements libpam_sshauth, I'd just remove the shadow entry, as it would be redudant
| |
| 02:19 | <ogra> i wouldnt fiddle with shadow if possible
| |
| 02:19 | <alkisg> How would the users authenticate?
| |
| 02:19 | E.g. I'd want an admin to be able to mount the internal disks, but not a student
| |
| 02:19 | <ogra> how do they do it now ?
| |
| 02:19 | <alkisg> That's not possible with ltsp currently...
| |
| 02:19 | They don't
| |
| 02:19 | <ogra> at all ?
| |
| 02:20 | <alkisg> Yup
| |
| 02:20 | <ogra> oh
| |
| 02:20 | <alkisg> The user cannot authenticate himself *locally*
| |
| 02:20 | <ogra> i thought you already had a solution and were just held up by PK
| |
| 02:20 | <alkisg> So any localapps which need authentication do not work
| |
| 02:20 | E.g. if the screensaver runs locally, and it locks, there's no way to unlock it
| |
| 02:21 | OK, storing a hash in shadow does have some security implications
| |
| 02:21 | <ogra> gss talks through dbus to the session bus ;)
| |
| 02:21 | <alkisg> But I think it should be there as an option
| |
| 02:21 | <ogra> fix dbus ;)
| |
| 02:21 | <alkisg> That wouldn't help
| |
| 02:21 | <ogra> sure
| |
| 02:21 | Selveste1 has quit IRC | |
| 02:21 | <alkisg> I'd need to fix pam
| |
| 02:21 | <ogra> you'd need to merge server and client busses
| |
| 02:21 | Selveste1 has joined #ltsp | |
| 02:22 | <alkisg> How would that help with unlocking the screensaver?
| |
| 02:22 | <ogra> (in your fat client case it would have to be the system busses)
| |
| 02:23 | the system bus looks for an entry in shadow ... if your busses are merged it looks for that on the server
| |
| 02:23 | <alkisg> But for fat clients the system bus is on the client...
| |
| 02:23 | <ogra> right, because you didnt kmerge the two busses yet
| |
| 02:23 | <alkisg> How can one merge 2 system busses?
| |
| 02:24 | <ogra> the system busses need to be tied together at login time
| |
| 02:24 | with a ssh based dbus protocol :)
| |
| 02:24 | <alkisg> I'm not worried about the communication part
| |
| 02:24 | <ogra> in case of thin clients you use the system bus on the client and merge the session busses
| |
| 02:24 | <alkisg> I just don't think it's possible to merge 2 system buses
| |
| 02:24 | <ogra> in case of fat clients you run the session bus locally and merge the system busses
| |
| 02:25 | <alkisg> So which networkmanager would I keep?
| |
| 02:25 | The server one or the client one?
| |
| 02:25 | Which realtimekit? which pk?
| |
| 02:25 | <ogra> you would need some policies ...
| |
| 02:25 | <alkisg> which avahi?
| |
| 02:25 | which upower?
| |
| 02:25 | <ogra> as i said, you need some policies
| |
| 02:25 | <alkisg> When the client wanted to get to standby, would the server also get to standby?
| |
| 02:26 | I think each system bus is designed to manage exactly one machine
| |
| 02:26 | <ogra> all auth goes through the server bus ... all HW access through the client bus
| |
| 02:26 | <alkisg> I don't think it's possible to merge 2 system busses...
| |
| 02:26 | <ogra> no, but make your session bus talk to one or the other
| |
| 02:26 | (thats what i mean with merge :) )
| |
| 02:27 | network transparent dbus is the only proper solution to either thin or fat clients
| |
| 02:27 | and the proper set of policies
| |
| 02:27 | <alkisg> I can't even begin to understand how that would work.
| |
| 02:27 | I think it's not designed to do that.
| |
| 02:28 | <ogra> you connect your session bus to both system buses at log in (dbus needs to learn that, i thinnk sbalneav looked into various options for dbus over ssh)
| |
| 02:28 | and have a policy for the session bus that ll auth requests go to the server
| |
| 02:28 | everything else goes to the client
| |
| 02:29 | for thin clients you want it exactly the other way round
| |
| 02:29 | <alkisg> How would a session bus be connected to 2 system busses?
| |
| 02:29 | <ogra> you have two session buses that both need to connect to the clients system bus
| |
| 02:29 | <alkisg> Would I duplicate all C pointers inside their source code?
| |
| 02:29 | <ogra> i have no idea ...
| |
| 02:29 | i'm not a dbus hacker
| |
| 02:30 | i just say thats the only proper solution :)
| |
| 02:30 | <alkisg> I mean, *all* client apps would need to be changed to support that
| |
| 02:30 | <ogra> no
| |
| 02:30 | <alkisg> We couldn't just change dbus for that
| |
| 02:30 | <ogra> just the bus
| |
| 02:30 | <alkisg> We'd need to change all apps that talk to dbus...
| |
| 02:30 | <ogra> no
| |
| 02:30 | you need to change the dbus policies
| |
| 02:31 | dbus knows if an auth request comes
| |
| 02:31 | so it needs to route it to the right system bus
| |
| 02:31 | the only apps you will get probs with are gksu apps in the case of your fat client approach ...
| |
| 02:31 | but gksu is going away
| |
| 02:32 | <alkisg> ogra, I still don't think dbus has anything to do with authentication
| |
| 02:32 | <ogra> huh ?
| |
| 02:32 | <alkisg> If a user wants to run something from the console, e.g. sudo ls, it won't work
| |
| 02:32 | <ogra> dbus is 'all' about authentication
| |
| 02:33 | no, as i said, sudo apps will get you probs
| |
| 02:33 | <alkisg> If he wants to run getent shadow, it won't work
| |
| 02:33 | So the proper way to solve this is with a pam module
| |
| 02:33 | <ogra> well, there is no safe solution to that prob
| |
| 02:33 | nothing thats not hackish at least ...
| |
| 02:33 | <alkisg> That way it would work no matter what's used at the backend
| |
| 02:34 | Be it ldap or libpam-sshauth
| |
| 02:34 | <ogra> for GUI apps dbus provides a proper solution if you modify it the right way
| |
| 02:34 | <alkisg> dbus doesn't get involved in that. It just uses whatever pam is available...
| |
| 02:34 | <ogra> libpam-sshauth wont solve the dbus prob
| |
| 02:34 | X-Raimo has left #ltsp | |
| 02:34 | <alkisg> Sure, but it will solve the authentication problem
| |
| 02:34 | And, for fat clients, nothing dbus-related is needed
| |
| 02:35 | <ogra> well, then help finishing libpam-sshauth :)
| |
| 02:35 | fiddling with shadow us definately a bad idea
| |
| 02:35 | <alkisg> I don't have the knowledge to do that
| |
| 02:35 | <ogra> *is
| |
| 02:36 | * ogra needs to reboot after an upgrade... | |
| 02:36 | <alkisg> ogra, I just want the fat clients plugin to work good enough for lucid
| |
| 02:36 | I don't see any other options here
| |
| 02:37 | It's either 'modifying polkit files' or 'inserting a shadow entry'
| |
| 02:37 | It's not nearly as bad as allowing LDM_PASSWORD...
| |
| 02:37 | ...anyone can get that with tftp
| |
| 02:38 | And, the polkit approach won't be there for long - I think the shadow entry gives a more consistent user experience
| |
| 02:38 | Anyway, I'd appreciate any hints on how to make CK think that our session is ACTIVE... :)
| |
| 02:38 | ogra has quit IRC | |
| 02:39 | ogra has joined #ltsp | |
| 02:39 | <nubae> alkisg: and LDAP won't do?
| |
| 02:40 | <alkisg> nubae: ldap would do just fine, but it's not integrated to ltsp
| |
| 02:40 | <nubae> how not? its integrated enough
| |
| 02:40 | I use it
| |
| 02:40 | <alkisg> How do you build a client that uses ldap for authentication?
| |
| 02:40 | <ogra> because you picked a default setup
| |
| 02:41 | <nubae> thesame way build any client
| |
| 02:41 | that has nothing to do with the authentication
| |
| 02:41 | <alkisg> ogra, http://alkisg.pastebin.com/s6uzW0bB
| |
| 02:41 | <ogra> the prob is that there is nobody who said "thats the ubuntu default, screw all other setups"
| |
| 02:41 | <alkisg> nubae: are the clients authenticating *WITH LDAP*? or with ssh to the server, then THEN use ldap?
| |
| 02:41 | <ogra> so if someone has an existing ltsp server that uses ldap and you enforce a default on him you will break the existing setup
| |
| 02:42 | <alkisg> nubae: try this: ltsp-localapps xterm => and then login nubae. Does that work? Can you login locally?
| |
| 02:42 | (or, SHELL_02=shell => login as nubae)
| |
| 02:42 | *SCREEN_02
| |
| 02:43 | <nubae> why wouldnt it work?
| |
| 02:43 | ldap authentication works fine across the board
| |
| 02:43 | <alkisg> Because the client doesn't use ldap
| |
| 02:43 | The server uses ldap...
| |
| 02:43 | <ogra> alkisg, fiddling with shadow is error prone and might even introduce security isssues if yousay "its either PK or shadow", go with PK
| |
| 02:43 | <nubae> u set it up to use ldap
| |
| 02:43 | thats a 5 minute setuop
| |
| 02:43 | <alkisg> nubae: so you don't use ldm
| |
| 02:43 | ?
| |
| 02:43 | because ldm uses ssh, not ldap...
| |
| 02:43 | <nubae> I do use ldm, that has nothing to do with it....
| |
| 02:43 | right
| |
| 02:43 | <alkisg> Sure it does
| |
| 02:44 | <nubae> but ldap supports ssh too
| |
| 02:44 | its an all encompassing authentication system
| |
| 02:44 | <alkisg> nubae: ok, are you one a client?
| |
| 02:44 | <ogra> nubae, the problem is that you might break existing ldap server setups
| |
| 02:44 | <nubae> no, as many as I want
| |
| 02:44 | <alkisg> Put SCREEN_02=shell, and run: getent passwd
| |
| 02:44 | <nubae> why=
| |
| 02:44 | <alkisg> You won't see the ldap users there. Try it.
| |
| 02:44 | <nubae> the client setup is identical to a non ltsp client setup
| |
| 02:45 | knipwim_ has joined #ltsp | |
| 02:45 | <ogra> yes, the client setup should be pretty safe to do
| |
| 02:45 | <nubae> I wont see why users where?
| |
| 02:45 | <ogra> the server setup wont
| |
| 02:45 | <nubae> I have tried it and run it... it runs just fine
| |
| 02:45 | The first time I ran fatclients was LDAP only... there were no issues
| |
| 02:46 | I just uses the passwd shadow copy hack method cause it was easier and didnt require ldap installed in the client
| |
| 02:46 | but both methods work
| |
| 02:46 | ogra: I dont understand what u mean, u dont need to change the ldap server setup
| |
| 02:46 | knipwim has quit IRC | |
| 02:46 | <nubae> it is the same as for any other system
| |
| 02:46 | <ogra> ldap is surely cleaner in that case ... but there is no solution to the server side unless someone defines a standard layout for all ubuntu ldap servers
| |
| 02:47 | <nubae> u mena schema wise?
| |
| 02:47 | mean
| |
| 02:47 | <ogra> nubae, the db layout is freely choosable
| |
| 02:47 | right
| |
| 02:47 | <nubae> ok...
| |
| 02:47 | <alkisg> nubae: in any case, the client session isn't considered active. So even if ldap worked, you couldn't use the local disks etc. We'd need to fix LTSP against consolekit first.
| |
| 02:47 | <nubae> now I get u... still u can have mutliple schemas
| |
| 02:47 | and they are very easily transferable
| |
| 02:47 | <ogra> as long as there is no standard you might completely screw up existing setups
| |
| 02:47 | thats the biggest prob with ldap
| |
| 02:48 | if there would be a defined standard you could just hook into it
| |
| 02:48 | <nubae> that I dont get... u're just talking about fitting an existing database into the ldap server setup... admins go through that all the time
| |
| 02:48 | and thats not ltsp specific
| |
| 02:48 | <ogra> no
| |
| 02:48 | <nubae> I'm running LDAP with ltsp right now... with a standard posix schema
| |
| 02:48 | it runs just fine
| |
| 02:49 | <ogra> if an admin defined a schema for all the stuff he wants authenticated you cant guarantee your schema mautches
| |
| 02:49 | *matches
| |
| 02:49 | so you add your schema additionally ... that means two auth db's
| |
| 02:49 | <nubae> but u can modify a schema (ldap can read more than one schema)
| |
| 02:49 | so?
| |
| 02:49 | <ogra> which defeats the purpose of ldap :)
| |
| 02:49 | the admin has to maintin two auth dbs
| |
| 02:50 | <nubae> vs hacking tons of scripts to get everything from local devs to sound to god knows what else working?
| |
| 02:50 | <ogra> ldap will only work properly if someone finally defines a standard
| |
| 02:50 | <nubae> no... u can have the auth all in one db
| |
| 02:50 | u dont need 2 seperate dbs
| |
| 02:50 | I have it in one db
| |
| 02:50 | <ogra> how if the db schemas dont match ?
| |
| 02:50 | <nubae> its just a matter of choosing one from the beginning
| |
| 02:50 | <ogra> right
| |
| 02:50 | thats my point
| |
| 02:51 | <nubae> but that has to be done anyway
| |
| 02:51 | <ogra> the admin of the existing server already did that cice
| |
| 02:51 | *choice
| |
| 02:51 | <nubae> there are hundreds of apps that use ldap
| |
| 02:51 | they must always adapt the db to make them work
| |
| 02:51 | <ogra> and millions of different schemas
| |
| 02:51 | <nubae> this is no different
| |
| 02:51 | the only issue here is that u need to know LDAP
| |
| 02:51 | which is easier than the violent hacks we've been doing till now
| |
| 02:51 | <ogra> well, then you already lost :)
| |
| 02:52 | <nubae> I'm comparing the alternatives
| |
| 02:52 | LDAP is not a bad one
| |
| 02:52 | <ogra> how would your school teacher that probably is able to run ltsp-cild-client *know ldap* ?
| |
| 02:52 | <nubae> give him webmin
| |
| 02:52 | <ogra> *build
| |
| 02:52 | AGRH!
| |
| 02:52 | <nubae> a teacher wouldnt build it
| |
| 02:52 | <ogra> to trash his server ?
| |
| 02:52 | #no thanks
| |
| 02:52 | <nubae> yeah I know u dont like it, but it actually works well with ldap
| |
| 02:52 | trash?
| |
| 02:52 | <ogra> webmin isnt in debian and ubuntu for a reason :)
| |
| 02:53 | <nubae> sigh...
| |
| 02:53 | yes yes the ancient argument
| |
| 02:53 | as if in the 5 years nothing has changed
| |
| 02:53 | <ogra> it breaks configs and exposes huge security holes
| |
| 02:53 | <nubae> u should actually look at webmin recently
| |
| 02:53 | no it doesnt
| |
| 02:53 | thats total propaganda
| |
| 02:53 | <ogra> i'm not after looking into webmin
| |
| 02:53 | <alkisg> ebox works with ldap as well, and supports windows clients + AD + samba PDC + ... as well
| |
| 02:53 | <ogra> i just pointed out the weakness of ldap in that setup
| |
| 02:54 | <nubae> ebox breaks far far more than webmin ever did
| |
| 02:54 | <alkisg> It'll even support roaming profiles for linux clients
| |
| 02:54 | <nubae> ebox is like letting a hacksaw at your server
| |
| 02:54 | because it DOES take over all functions
| |
| 02:54 | <alkisg> ebox states exactly which files it modifies
| |
| 02:54 | <ogra> anyway ... /me needs to do some real work
| |
| 02:54 | <nubae> without leaving configuration files intact where they normally are
| |
| 02:55 | <alkisg> webmin modifies whatever it wants and hides its security problems
| |
| 02:55 | <nubae> I've never understood this ubuntu hatred towards webmin but such embracement for ebox... which is 100 times worse
| |
| 02:55 | no it doesnt
| |
| 02:55 | <alkisg> Yup
| |
| 02:55 | <nubae> it doesnt modify ANYTHING in the configuration
| |
| 02:55 | ebox does that
| |
| 02:55 | <alkisg> See the thread on why it got out of debian/ubuntu
| |
| 02:55 | <nubae> u can cleanly uninstall webmin and it would do no harm at all
| |
| 02:55 | <ogra> guys, stop the flamewar :)
| |
| 02:55 | <nubae> try that with ebox
| |
| 02:55 | ok ok
| |
| 02:55 | sorry...
| |
| 02:55 | too much coffee and no cigarettes
| |
| 02:55 | <alkisg> Anyway, on to fix LTSP to work with CK...
| |
| 02:55 | <nubae> gr....
| |
| 02:56 | <ogra> webmin isnt there .... its no choice for ubuntu or debian
| |
| 02:56 | (and i think neither for fedora)
| |
| 02:56 | <nubae> well it can be installed from their site
| |
| 02:56 | fedora its in their repo yeah
| |
| 02:56 | as it is in suse
| |
| 02:57 | that doesnt say much i know...
| |
| 02:59 | Selveste1 has quit IRC | |
| 03:00 | Selveste1 has joined #ltsp | |
| 03:03 | nubae has quit IRC | |
| 03:03 | nubae has joined #ltsp | |
| 03:11 | <alkisg> A "normal" session says: x11-display-device = '/dev/tty7', display-device = ''. A localapps session says: x11-display-device = '', display-device = '/dev/tty7'. I wonder how would I change that...
| |
| 03:13 | <ogra> smells like there is a chvt somewhere in the code
| |
| 03:13 | <alkisg> Not anymore (in the ltsp code)...
| |
| 03:14 | * alkisg looks at the X11/Xsession code... | |
| 03:15 | ogra has quit IRC | |
| 03:15 | bobby_C has joined #ltsp | |
| 03:15 | ogra has joined #ltsp | |
| 03:22 | japerry has quit IRC | |
| 03:25 | <alkisg> SCREEN_02=shell => that executes bash -login, but that's not an actual login.
| |
| 03:25 | Changing this to `su` makes it an actual login, i.e. HOME and USER are set, and ck-list-sessions lists that session
| |
| 03:26 | su needs a bit more memory than bash -login, but I think it's more appropriate...
| |
| 03:26 | Any reason not to change `bash -login` to `su` in screen.d/shell?
| |
| 03:28 | Kicer86 has joined #ltsp | |
| 03:36 | MaRX-Mode has quit IRC | |
| 03:48 | Kicer86 has quit IRC | |
| 03:49 | pths has quit IRC | |
| 03:49 | pts has joined #ltsp | |
| 03:50 | pts is now known as pths | |
| 03:56 | mikkel has joined #ltsp | |
| 04:37 | pmatulis has joined #ltsp | |
| 04:57 | ogra has quit IRC | |
| 04:57 | ogra has joined #ltsp | |
| 04:59 | FinnTux has quit IRC | |
| 05:00 | Selveste1 has quit IRC | |
| 05:01 | cyberorg has quit IRC | |
| 05:01 | FinnTux has joined #ltsp | |
| 05:04 | cyberorg has joined #ltsp | |
| 05:13 | alkisg has quit IRC | |
| 05:15 | pths has quit IRC | |
| 05:30 | daya has quit IRC | |
| 05:33 | pts has joined #ltsp | |
| 05:35 | pts is now known as pths | |
| 05:46 | Hurga has joined #ltsp | |
| 05:48 | <Hurga> I there. I'm experiencing random crashes of Thunderbird on Ubuntu 8.04 LTSP thin clients. Is this a known issue perhaps?
| |
| 05:52 | MaRX-Mode has joined #ltsp | |
| 06:11 | bobby_C has quit IRC | |
| 06:27 | thunsucker has joined #ltsp | |
| 06:43 | Selveste1 has joined #ltsp | |
| 06:55 | hersonls has joined #ltsp | |
| 06:58 | bobby_C has joined #ltsp | |
| 06:59 | thunsucker has quit IRC | |
| 07:07 | mikkel has quit IRC | |
| 07:09 | shawnp0wers has joined #ltsp | |
| 07:13 | slidesinger has joined #ltsp | |
| 07:14 | scottmaccal has joined #ltsp | |
| 07:17 | slidesinger has quit IRC | |
| 07:25 | mgariepy has joined #ltsp | |
| 07:26 | <mgariepy> morning all
| |
| 07:40 | <AndyGraybeal> hi mgariepy :) thank you for the help last night.
| |
| 07:41 | still didn't get autologin to work, i will come back to it.
| |
| 07:45 | <mgariepy> no problem AndyGraybeal ;) it was a pleasure
| |
| 07:45 | etyack has joined #ltsp | |
| 07:49 | <AndyGraybeal> is there somewhere tricky i will be setting a umask using LTSP and Ubuntu/Gnome?
| |
| 07:52 | <Hurga> I'm experiencing random crashes of Thunderbird on Ubuntu 8.04 LTSP thin clients. Is this a known issue perhaps?
| |
| 07:53 | highvoltage has quit IRC | |
| 07:55 | highvoltage has joined #ltsp | |
| 07:56 | litlebuda has joined #ltsp | |
| 08:06 | etyack has left #ltsp | |
| 08:09 | alkisg has joined #ltsp | |
| 08:09 | <AndyGraybeal> should i change the umask in ~/.profile or in /etc/X11/Xsessions.d ? i'm using ubuntu 8.10. i'm asking because i've been reading a lot of websites that say gnome doesn't read the ~/.profile
| |
| 08:20 | litlebuda has quit IRC | |
| 08:23 | Gadi has joined #ltsp | |
| 08:27 | tthorb_ has joined #ltsp | |
| 08:30 | tthorb has quit IRC | |
| 08:33 | prpplague_afk has quit IRC | |
| 08:47 | <sbalneav> Morning all
| |
| 08:48 | japerry has joined #ltsp | |
| 08:48 | <sbalneav> AndyGraybeal: You'd usually want to set up a script in the Xsessions.d file
| |
| 08:48 | what are you wanting to change the umask to?
| |
| 08:48 | Maybe there's a better way to accomplish what you want
| |
| 08:49 | <alkisg> There's also UMASK in /etc/login.defs, I don't know if it's used or not.
| |
| 08:51 | <sbalneav> alkisg: I've never been able to tell.
| |
| 08:51 | I ought to spend 15 minutes at some point and find out the "proper" way to do it.
| |
| 08:52 | Usually, I've always done it as my own script in Xsession.d, since that's kind of "guaranteed" not to get overwritten on an update.
| |
| 08:53 | * alkisg is struggling with consolekit... if anyone has any idea on how to tell CK about our x-display-device, I'm all ears :) | |
| 08:54 | <alkisg> Hmmm maybe that's the bug I'm seeing: https://bugs.launchpad.net/ubuntu/+source/consolekit/+bug/483130
| |
| 09:04 | <sbalneav> Sounds like it.
| |
| 09:05 | There was some activity on the mailing list I think that talked about widening the permissions on ck to allow some of this sort of thing to happen
| |
| 09:05 | might that be the solution?
| |
| 09:12 | <alkisg> Our first problem are not the permissions, it's that the current user is not considered "ACTIVE"
| |
| 09:12 | If it was active, then no policies would need to be changed
| |
| 09:13 | <johnny> alkisg, so.. how does enable autologin with fat clients nowadays? editing the gdm.conf file?
| |
| 09:13 | <alkisg> But I can't find out what does CK need to mark a session as active...
| |
| 09:13 | <johnny> alkisg, you should ask a gnome developer
| |
| 09:13 | <alkisg> johnny: the same as thin clients
| |
| 09:13 | johnny: no gdm involved at all
| |
| 09:13 | <johnny> ah..
| |
| 09:13 | that was nubae's method then
| |
| 09:13 | not the new one
| |
| 09:13 | <alkisg> Yeah
| |
| 09:13 | <Gadi> alkisg: doesnt that script in the bug report fix it for you?
| |
| 09:14 | <alkisg> Gadi, no, it turns out it's a different problem
| |
| 09:14 | <johnny> alkisg, if you ever have obscure questions like that.. you should ask people on #gnome-hackers on irc.gnome.org/irc.gimp.net
| |
| 09:14 | <alkisg> The problem he mentions was fixed in lucid
| |
| 09:14 | <Gadi> alkisg: it sounds similar, as we start our session through xinit
| |
| 09:14 | <alkisg> johnny: thanks, I'll join #gnome-hackers now
| |
| 09:14 | <johnny> friendly people
| |
| 09:14 | mostly anyways :)
| |
| 09:15 | <alkisg> Gadi: let me post you an example from ck-list-sessions to see the difference...
| |
| 09:16 | http://alkisg.pastebin.com/9dV0M2Qi
| |
| 09:16 | The difference is that we do get an x11-display (while he didn't), but we don't get an x11-display-device
| |
| 09:17 | So we need to find out how to tell CK to use `display-device` as the `x11-display-device` instead
| |
| 09:18 | <Gadi> alkisg: what is the value of XDG_SESSION_COOKIE in both cases?
| |
| 09:19 | <alkisg> Gadi, as far as I can see, it's ok - ck-launch-session sets it
| |
| 09:21 | <Gadi> alkisg: what happens if you add a script in the fat client: /etc/X11/Xsession.d/89-stripcookie that does: unset XDG_SESSION_COOKIE
| |
| 09:21 | <nubae> btw... opensuse has a fatclient version for ltsp... u should check how they are doing it
| |
| 09:22 | <alkisg> Gadi: I don't think XDG_SESSION_COOKIE exists at that point
| |
| 09:22 | <Gadi> it should
| |
| 09:22 | the next script calls: if [ -z "$XDG_SESSION_COOKIE" ] && [ -x "$CK_LAUNCH_SESSION" ]; then
| |
| 09:22 | STARTUP="$CK_LAUNCH_SESSION $STARTUP"
| |
| 09:22 | fi
| |
| 09:22 | <alkisg> It gets created later on, when the STARTUP command line is evaluated
| |
| 09:23 | <Gadi> oh, I see
| |
| 09:23 | what you mean
| |
| 09:23 | <alkisg> Btw, we create 2 CK sessions, we shouldn't do that
| |
| 09:23 | ...but that's another story :)
| |
| 09:34 | <AndyGraybeal> sbalneav: hi thank you for responding. i'm thinking about setting the umask to 002
| |
| 09:34 | sbalneav: should i just make a file inside of /etc/X11/Xsessions.d/
| |
| 09:35 | sbalneav: or should i put it into a file that is already there?
| |
| 09:35 | <sbalneav> Create a new file
| |
| 09:35 | that way it won't get overridden if something upgrades
| |
| 09:35 | <AndyGraybeal> sbalneav: is ther a naming scheme for Xsessions.d ? or call it anything i like?
| |
| 09:36 | nubae has quit IRC | |
| 09:37 | nubae has joined #ltsp | |
| 09:43 | <AndyGraybeal> is see the files have numbers in front of them
| |
| 09:47 | i made a file named /etc/X11/Xsessions.d/45umask_002 with umask 002 .. inside fo the file and it works great!
| |
| 09:47 | <sbalneav> That's to set order
| |
| 09:50 | <alkisg> Gadi, are the Xsession.d scripts executed as root? Or as the user logging on?
| |
| 09:50 | <Gadi> user
| |
| 09:51 | <alkisg> I wonder if we need to "su" before executing them, to go through pam...
| |
| 10:15 | vvinet has joined #ltsp | |
| 10:20 | MorpheusDe has joined #ltsp | |
| 10:22 | bobby_C has quit IRC | |
| 10:25 | <AndyGraybeal> sudo /etc/init.d/ntp restart
| |
| 10:25 | hmm :)
| |
| 10:28 | <MorpheusDe> hi there - is there more information/ documentation available regarding the thick client concept introduced in 5.2?
| |
| 10:37 | <johnny> MorpheusDe, do you have a specific question about it?
| |
| 10:38 | <alkisg> Here's a small wiki page I wrote about it: https://help.ubuntu.com/community/UbuntuLTSP/FatClients
| |
| 10:40 | thunsucker has joined #ltsp | |
| 10:41 | <MorpheusDe> right now I'm looking through the upstream documentation and I can't find anything specifically dealing with thick clients. Being fairly new to LTSP I did not want to clutter up the server on the first install. I chrooted into the i386 jail and installed ubuntu desktop which of course did not work. As I already said a) I did not wanted to clutter up the server itself, b) wanted to have the ability to install different versions of desktop env
| |
| 10:44 | <thunsucker> MorpheusDe: I believe you need to install the gnome desktop onto the server then recreate the client
| |
| 10:44 | <alkisg> MorpheusDe: your message was too long and it was cutted at "have the ability to install different versions of desktop env"
| |
| 10:45 | But sure, you can have as many different chroots/desktop environments as you want
| |
| 10:45 | staffencasa has joined #ltsp | |
| 10:46 | <alkisg> But you need to run ltsp-build-client --fatclient to make it. Just installing ubuntu-desktop on the chroot isn't enough
| |
| 10:46 | <MorpheusDe> alkisg: Oops sorry - I'm able to read it ok using pidgin.. here's the rest - b) wanted to have the ability to install different versions of desktop environments - But of course the fat clients do not perform so well on weak clients - but I guess it's the only way to go if you wan to do what I tried to do
| |
| 10:46 | <alkisg> The ltsp-build-client plugin does other things...
| |
| 10:46 | <MorpheusDe> alkisg: Ahm - you mean as thick clients, right?
| |
| 10:46 | <alkisg> For clients with < 500 RAM I think thin chroots are prefered
| |
| 10:46 | For clients with > 500, I'd go for fat
| |
| 10:47 | MorpheusDe: yes, thick == fat
| |
| 10:47 | <MorpheusDe> alkisg: To clarify - without thick clients I can't have different desktops envs on one server right?
| |
| 10:47 | <alkisg> Sure you can
| |
| 10:48 | Both with thin & with fat clients.
| |
| 10:48 | <MorpheusDe> alkisg: Hmm - I think I still don't get some parts of the main concept then...
| |
| 10:48 | <alkisg> MorpheusDe: what do you want to have as the end result?
| |
| 10:48 | What are your needs?
| |
| 10:49 | <MorpheusDe> alkisg: Primarily testing right now - Installed ltsp, installed ubuntu-desktop package on the server - so if you would add for example kde - the thin client can choose the env somehow?
| |
| 10:50 | <alkisg> Yes, from ldm
| |
| 10:50 | But that's not the case for fat clients
| |
| 10:50 | For fat clients, you'd install the desktop environments *on the chroot*, not on the server
| |
| 10:51 | <MorpheusDe> alkisg: Alright - so I could have one client booting up gnome, one kde and so on?
| |
| 10:51 | <alkisg> Yes
| |
| 10:51 | staffencasa has quit IRC | |
| 10:52 | otavio has quit IRC | |
| 10:52 | <MorpheusDe> alkisg: Thanks - Got to do more testing on that sounds interesting....
| |
| 10:52 | <alkisg> You're welcome
| |
| 10:52 | otavio has joined #ltsp | |
| 10:53 | <Gadi> (heh - which brings up a good point regarding the role of ldminfod on fat clients - I suppose you need to run it locally and redirect ldm to get its info from a local ldminfod)
| |
| 10:53 | <MorpheusDe> alkisg: Just one more question regarding the italc topic - Yesterday you said copy the keys from the server to the separate workstation - What about controlling a second ltsp server - I guess the key structure of the second machine would overwrite the one of the first...
| |
| 10:54 | <alkisg> In all the "admin" PCs you need the same keys (both private & public). In all the "clients" you need the same public keys.
| |
| 10:56 | staffencasa has joined #ltsp | |
| 10:56 | <MorpheusDe> alkisg: Yepp - you're right - that was a really stupid question - Simply a usual private public key concept....
| |
| 10:57 | alkisg: One more question and I'm out for more testing - I guess it's simply enough to install the italc client on the ltsp servers if I'm using a separate machine for controlling, right?
| |
| 10:58 | <alkisg> Yup
| |
| 10:58 | <MorpheusDe> THX
| |
| 11:04 | Webuser314 has joined #ltsp | |
| 11:09 | <Webuser314> hi, having difficulty detecting a camera in ubuntu ltsp 9.10. localdev is working in that usb memory sticks are detected correctly, but nothing for a camera. The camera (canon a95) used to work out of the box in standalone ubuntu. gphoto2 reports nothing can't seem to find anything from forums about how to troubleshoot this. any pointers? thanks
| |
| 11:15 | <Gadi> Webuser314: localdev does not support PTP, only mass storage
| |
| 11:15 | if you can set the camera to mass storage mode, it will work
| |
| 11:16 | <Webuser314> I thought the fact that it was detected in ubuntu standalone and automounted as desktop item meant that it was mass storage? I guess I was wrong?
| |
| 11:16 | <Gadi> there are two ways to talk to a camera:
| |
| 11:17 | PTP (which is a protocol language) or USB mass storage (which just treats it as a USB flash drive)
| |
| 11:18 | <sbalneav> Webuser314: Currently, LTSP doesn't support the PTP model, but a full desktop system does.
| |
| 11:18 | <Gadi> right
| |
| 11:18 | <Webuser314> Is there a workaround if I can't set the camera to usb mass storage?
| |
| 11:19 | <sbalneav> Yes.
| |
| 11:19 | Yank the sd card out of the camera and read it in a reader.
| |
| 11:19 | <Gadi> you might be able to install an application and run it as a LocalApp that would talk to the camera
| |
| 11:19 | or do what sbalneav suggests
| |
| 11:19 | <sbalneav> or that, which might take some fiddling :)
| |
| 11:20 | <Webuser314> I'm doing localapps already, so I might try that.....thanks again guys, this channel is always so helpful.
| |
| 11:20 | <sbalneav> Webuser314: But just about any modern camera I've seen can switch back and forth
| |
| 11:20 | thunsucker has quit IRC | |
| 11:23 | thunsucker has joined #ltsp | |
| 11:26 | <Hurga> after a few reboots of the client the swap files of nbdswapd are piling up in my /tmp... how long does it take for them to be removed?
| |
| 11:27 | <Gadi> Hurga: does your /etc/hosts.allow have an nbdswapd entry?
| |
| 11:27 | or just an nbdrootd entry?
| |
| 11:35 | etyack has joined #ltsp | |
| 11:36 | mikkel has joined #ltsp | |
| 11:36 | <alkisg> Ugh, I'm able to use d-feet to set the active property to true: 'org.freedesktop.ConsoleKit.Session', 'active', True
| |
| 11:36 | Then everything works properly, devices are mounted etc etc
| |
| 11:37 | I wonder if we could use that as a workaround (I mean, a startup script that does that) :D
| |
| 11:39 | Faithful has quit IRC | |
| 11:41 | <alkisg> stgraber: would you mind if I changed the "su -" call in X95-run-x-session to "sudo", because su creates a duplicate consolekit session?
| |
| 11:44 | <johnny> alkisg, is this something i'm going to hav to be concerned about if i use fat clients?
| |
| 11:44 | <alkisg> johnny: a little
| |
| 11:44 | I.e. right now with the default installation local devices don't work for fat clients, we get authentication problems
| |
| 11:44 | because the session isn't active
| |
| 11:45 | One way to work around it is to mess with policy files
| |
| 11:45 | <johnny> ah.. well then i can't really deploy it then yet
| |
| 11:45 | <alkisg> That's easy to do
| |
| 11:45 | Tested, works etc
| |
| 11:45 | But I don't think it's the right thing to do
| |
| 11:45 | <johnny> can you put that on the wiki as a temporary fix?
| |
| 11:45 | <alkisg> I'm going to commit something in ltsp trunk in the next few days, so no
| |
| 11:46 | One way or another, an upstream workaround or solution is needed
| |
| 11:46 | <johnny> ah..i was hoping to deploy tonight.
| |
| 11:46 | <mgariepy> hey alkisg just had the chance to test clientid in udhcp, it's not working so great on my side.
| |
| 11:46 | <alkisg> mgariepy: how so?
| |
| 11:46 | <mgariepy> all my thin client get the same ip address
| |
| 11:46 | <alkisg> johnny: well of course you could do that tonight
| |
| 11:46 | <johnny> alkisg, how so?
| |
| 11:46 | <alkisg> mgariepy: you mean when using a clientid?
| |
| 11:47 | <johnny> as in.. what do i do? :)
| |
| 11:47 | <mgariepy> when not setting the client id, somehow, it sends the clientid=''
| |
| 11:47 | <alkisg> johnny: change everything to "yes" there: /usr/share/polkit-1/actions/org.freedesktop.udisks.policy
| |
| 11:47 | <Hurga> Gadi: Sorry, been AFK. Yes, my /etc/hosts.allow does have an nbdswapd entry.
| |
| 11:47 | <johnny> alkisg, thanks
| |
| 11:48 | <alkisg> mgariepy: that's the same as before, isn't it?
| |
| 11:48 | <mgariepy> yeah indeed,
| |
| 11:48 | <alkisg> mgariepy: or it is "null" vs "quotes" ?
| |
| 11:48 | <johnny> alkisg, my server has 2 gb ram.. and is used as a desktop in it's own right.. is that going to cause any problems ya think?
| |
| 11:48 | <mgariepy> alkisg, could we include the "-c ''" in the $clientid ?
| |
| 11:49 | <alkisg> johnny: no
| |
| 11:49 | mgariepy: let me give you a small patch...
| |
| 11:49 | <mgariepy> my lease contain two single quote as as uid
| |
| 11:50 | <alkisg> mgariepy: delete this line:
| |
| 11:50 | clientid=${clientid:-"''"}
| |
| 11:50 | completely
| |
| 11:50 | And change the call to:
| |
| 11:50 | if udhcpc -n -c "$clientid" -s /tmp/dhcp-script.sh -i $i $hostname_param $ip_param >/dev/null 2>&1; then
| |
| 11:50 | And see if that works
| |
| 11:51 | <johnny> i'm so ready to upgrade to f13 :)
| |
| 11:51 | just waiting for the most annoying bug for me to get fixed.. related to gnome-keyring
| |
| 11:54 | <mgariepy> alkisg, much better
| |
| 11:54 | <alkisg> mgariepy: should I commit that? It also looks cleaner that the quoted version... :D
| |
| 11:54 | *then
| |
| 11:55 | <mgariepy> i thinks you should commit it, if you don't i might do it ;) haha
| |
| 11:56 | <alkisg> Heh, no, commiting now..
| |
| 11:57 | done
| |
| 11:57 | Hurga has quit IRC | |
| 12:02 | thunsucker has quit IRC | |
| 12:04 | efra has joined #ltsp | |
| 12:11 | ogra_cmpc has quit IRC | |
| 12:13 | <mgariepy> alkisg, thanks :)
| |
| 12:13 | <alkisg> np - thanks for the VTs :)
| |
| 12:17 | johnny has left #ltsp | |
| 12:19 | johnny has joined #ltsp | |
| 12:24 | ogra_cmpc has joined #ltsp | |
| 12:27 | pimpministerp has joined #ltsp | |
| 12:28 | primeministerp has quit IRC | |
| 12:29 | Webuser314 has quit IRC | |
| 12:37 | pimpministerp has quit IRC | |
| 12:38 | joeasaurus has joined #ltsp | |
| 12:38 | <joeasaurus> hey
| |
| 12:38 | i need some help
| |
| 12:38 | please
| |
| 12:40 | I'm trying to set up a cluster of terminal servers using ltsp-cluster
| |
| 12:40 | and i think ive got it set up right, I cant see anything wrong
| |
| 12:40 | but when I got to boot a client, it can't pull the file from TFTP
| |
| 12:40 | alkisg has quit IRC | |
| 12:43 | japerry_cat has joined #ltsp | |
| 12:44 | shaunak has joined #ltsp | |
| 12:44 | <joeasaurus> Anyone upi for helping out?
| |
| 12:45 | up*
| |
| 12:46 | japerry has quit IRC | |
| 12:46 | japerry_cat is now known as japerry | |
| 12:47 | alkisg has joined #ltsp | |
| 12:47 | <Gadi> joeasaurus: ltsp-cluster does not pull lts.conf from tftp
| |
| 12:47 | it contact the control center over http
| |
| 12:47 | <shaunak> Hey I am trying to change the easy-ltsp GUI code. Can anyone help me how to compile the new code and see the changes ? The code is in C#
| |
| 12:49 | <joeasaurus> I know that Gadi, I mean it wont pull the image
| |
| 12:49 | the DHCP works so it gets an address but then it just hangs
| |
| 12:50 | <Gadi> joeasaurus: ah - by the "file" you meant pxelinux.0
| |
| 12:50 | <joeasaurus> uuh yeah
| |
| 12:50 | sorry, should have said so really
| |
| 12:50 | <Gadi> that has nothing to do with the cluster
| |
| 12:51 | make sure you only have inetd on the boot server
| |
| 12:51 | <joeasaurus> Oh right..what should i be looking at?
| |
| 12:51 | <Gadi> and not xinetd
| |
| 12:51 | <joeasaurus> okay
| |
| 12:51 | hang on a sec
| |
| 12:51 | <Gadi> then, make sure you have tftp listening (netstat -anp|grep :69)
| |
| 12:53 | <joeasaurus> okay so i have inetd and not xinetd
| |
| 12:53 | and i believe tftp is listening
| |
| 12:53 | udp 0 0 0.0.0.0:69 0.0.0.0:*
| |
| 12:55 | <Gadi> joeasaurus: ps -efw|grep tftp
| |
| 12:56 | <joeasaurus> cluster 2618 2575 0 18:57 pts/0 00:00:00 grep --color=auto tftp
| |
| 12:56 | <Gadi> that's good
| |
| 12:57 | what error do you get on the client?
| |
| 12:57 | <joeasaurus> I get "TFTP timeout"
| |
| 12:57 | or something similar
| |
| 12:59 | <Gadi> joeasaurus: do you have more than one network interface in the server?
| |
| 13:00 | <joeasaurus> Yeah but Im not using them atm, I have all the servers talking to each other and the clients through a switch
| |
| 13:01 | <Gadi> joeasaurus: but are more than one of them configured to an IP on the same subnet?
| |
| 13:01 | <joeasaurus> no
| |
| 13:01 | <Gadi> (ie, your routing table might be screwed up)
| |
| 13:01 | <joeasaurus> they're not configured for anything
| |
| 13:01 | but ill check the routes anyways
| |
| 13:02 | <Gadi> so ifconfig shows only one?
| |
| 13:02 | <joeasaurus> nah it shows all three but only one has an ip address, the other two just display there HW address
| |
| 13:02 | primeministerp has joined #ltsp | |
| 13:03 | <joeasaurus> this is my routing table
| |
| 13:04 | 192.168.10.128 * 255.255.255.192 U 0 0 0 eth1
| |
| 13:04 | link-local * 255.255.0.0 U 1000 0 0 eth1
| |
| 13:04 | default 192.168.10.189 0.0.0.0 UG 100 0 0 eth1
| |
| 13:06 | with 192.168.10.128 as the network and 192.168.10.189 as the gateway
| |
| 13:09 | shawnp0wers has quit IRC | |
| 13:10 | <Gadi> 255.255.255.192?
| |
| 13:10 | did you configure dhcpd.conf that way, as well?
| |
| 13:10 | default is: 255.255.255.0
| |
| 13:11 | plus, with this netmask, you will only be able to boot a handful of clients
| |
| 13:13 | <joeasaurus> yup dhcp is all set up
| |
| 13:13 | Ill paste the conf for you
| |
| 13:13 | I'm only planning on nine clients to start with
| |
| 13:14 | pimpministerp has joined #ltsp | |
| 13:14 | <Gadi> ok - make sure it is /etc/ltsp/dhcpd.conf you are editing
| |
| 13:14 | <joeasaurus> oh?
| |
| 13:14 | i didnt know that file existed
| |
| 13:15 | ive been doing the dhcp3 one
| |
| 13:15 | <Gadi> ah - surprise!
| |
| 13:15 | :)
| |
| 13:15 | <joeasaurus> ive just gone to edit that file
| |
| 13:15 | <Gadi> you may just need to copy your edited one there and restrt dhcp3-server
| |
| 13:15 | <joeasaurus> and it doesn't exist
| |
| 13:15 | yeah alright ill do that
| |
| 13:16 | <Gadi> really?
| |
| 13:16 | primeministerp has quit IRC | |
| 13:16 | <Gadi> did you install ltsp-server on that machine?
| |
| 13:16 | <joeasaurus> yeah
| |
| 13:16 | the directory is there
| |
| 13:16 | <Gadi> because it should put that as default
| |
| 13:16 | <joeasaurus> but that dhcp file isnt
| |
| 13:16 | <Gadi> huh
| |
| 13:16 | well, dhcp3-server will look for /etc/ltsp first and then /etc/dhcp3
| |
| 13:17 | so, maybe that wasnt the issue
| |
| 13:17 | still, please paste the conf
| |
| 13:17 | <joeasaurus> ill copy it over anyways then
| |
| 13:17 | <Gadi> ok
| |
| 13:17 | and restart the service
| |
| 13:19 | <joeasaurus> I'll pastebin it because irc wont let my message span more than one line
| |
| 13:19 | ive copied it and restarted the service though
| |
| 13:20 | http://pastebin.com/6BJEm6Rr
| |
| 13:23 | <Gadi> filename should be: /ltsp/i386/pxelinux.0
| |
| 13:23 | <joeasaurus> it used to be that but i changed it to see if it mattered
| |
| 13:24 | and it doesnt :P
| |
| 13:24 | <Gadi> (unless you set up your tftp server manually)
| |
| 13:24 | <joeasaurus> coz i saw a config that had an if in it
| |
| 13:24 | <Gadi> and your server is 192.168.10.181?
| |
| 13:24 | <joeasaurus> yeah
| |
| 13:25 | <Gadi> try setting option subnet-mask
| |
| 13:25 | pimpministerp has quit IRC | |
| 13:25 | primeministerp has joined #ltsp | |
| 13:25 | <Gadi> ie: option subnet-mask 255.255.255.192;
| |
| 13:26 | <joeasaurus> oh, btw, i remember, I changed a config to go all the way to ltsp so that i only need /i386/pxelinux.0
| |
| 13:27 | coz that makes more sense to me to have the arch folder as the differentiator and not the ltsp folder
| |
| 13:27 | if that makes sense?
| |
| 13:27 | shaunak has quit IRC | |
| 13:28 | <Gadi> it makes sense if you dont use tftp for anything else - of course, by changing it, it makes getting support a heck of a lot harder
| |
| 13:30 | <alkisg> ...and it also breaks ltsp-update-kernels :)
| |
| 13:30 | <joeasaurus> yeah, I only use tftp for this
| |
| 13:30 | I'll change it back if its better that way?
| |
| 13:31 | I only changed the one line so it's no hassle
| |
| 13:33 | okay, I changed it
| |
| 13:33 | /etc/default/tftpd-hpa now reads:
| |
| 13:33 | RUN_DAEMON="no"
| |
| 13:33 | OPTIONS="-l -s /var/lib/tftpboot"
| |
| 13:34 | /etc/ltsp/dhcpd.conf has the options subnet-mask
| |
| 13:34 | and ive changed filename to /letsp/i386/pxelinux.0
| |
| 13:38 | I cant test the clients atm coz they're at my college
| |
| 13:40 | <Gadi> joeasaurus: is there any particular reason you are using such a restricted and non-standard subnet mask? Are you limited by the college?
| |
| 13:40 | <joeasaurus> Well, the whole thing is a project me and my networking teacher have going
| |
| 13:40 | <Gadi> ah, ok
| |
| 13:40 | <joeasaurus> and he manages 3/4 different labs
| |
| 13:41 | so he had to slot the cluster inside one of the ranges he had
| |
| 13:41 | and 192 was the biggest
| |
| 13:42 | it gives us 62 hosts though so it's enough, I wouldnt really call it restricted. It's not a /29 or anything :P
| |
| 13:42 | <Gadi> well, restricted relative to the default
| |
| 13:43 | <joeasaurus> yeah, i got ya. I'm pretty sure the dhcp is set up correctly though. The clients get an IP address like i said, the boot just fails at the TFTP is all
| |
| 13:45 | <Gadi> which is why I suggest the option subnet-mask
| |
| 13:45 | <joeasaurus> yeah
| |
| 13:45 | <Gadi> you would get the same thing if, for example, you set a weird MTU option in dhcp
| |
| 13:45 | <joeasaurus> oh right
| |
| 13:46 | <Gadi> the client may not have its interface configured to access the network properly
| |
| 13:46 | <joeasaurus> nah, it does
| |
| 13:46 | <Gadi> you dont know that
| |
| 13:46 | <joeasaurus> coz they are windows computers during the day
| |
| 13:46 | so i know they can use the network xD
| |
| 13:46 | <Gadi> which means they are physically connected properly
| |
| 13:46 | <joeasaurus> yeah
| |
| 13:47 | <Gadi> that doesnt mean that when they pxe boot, that dhcp is configuring the nic properly
| |
| 13:47 | <joeasaurus> aah i see what you mean
| |
| 13:47 | well when the pxe is up
| |
| 13:47 | it shows all the info
| |
| 13:47 | it has IP Address, subnet
| |
| 13:47 | dhcp server address
| |
| 13:47 | and gateway
| |
| 13:48 | and dns address too i believe
| |
| 13:48 | but i guess it could show that and not actually apply the settings
| |
| 13:48 | <Gadi> hmm...
| |
| 13:48 | well, you should be able to sniff the traffic and see where it dies
| |
| 13:49 | <joeasaurus> okay
| |
| 13:50 | so imma try and boot them tomorrow
| |
| 13:50 | and sniff the traffic as it happens
| |
| 13:50 | and then I'll get back to you in the evening?
| |
| 13:51 | <Gadi> ah, Norway
| |
| 13:51 | yes, evening
| |
| 13:51 | :)
| |
| 13:51 | <joeasaurus> well, about this time again?
| |
| 13:51 | <Gadi> sounds good
| |
| 13:51 | <joeasaurus> alright, cool
| |
| 13:51 | thanks alot :)
| |
| 13:52 | <Gadi> cheers
| |
| 13:52 | joeasaurus has quit IRC | |
| 13:57 | johnny has left #ltsp | |
| 13:58 | johnny has joined #ltsp | |
| 14:01 | vandys has joined #ltsp | |
| 14:04 | <vandys> I have an eeeBox B202, it hangs after loading the LTSP client image. From scanning reports, it sounds like I need the jme.ko driver in the client boot image?
| |
| 14:04 | If so, how do I go about rebuilding the client boot image with this module configured in? Pointers to a procedure would be most welcome! Thanks.
| |
| 14:06 | <Gadi> vandys: add a line that reads: jme to /opt/ltsp/i386/etc/initramfs-tools/modules
| |
| 14:06 | vandys: for good measure, I would also add it to: /opt/ltsp/i386/etc/modules
| |
| 14:06 | then:
| |
| 14:07 | chroot /opt/ltsp/i386
| |
| 14:07 | mount -t proc proc /proc
| |
| 14:07 | update-initramfs -u
| |
| 14:07 | /usr/share/ltsp/update-kernels
| |
| 14:07 | umount /proc
| |
| 14:07 | exit
| |
| 14:07 | sudo ltsp-update-kernels
| |
| 14:07 | done
| |
| 14:10 | <vandys> update-kernels was in /usr/lib/ltsp, otherwise all commands worked... lemme go spin up the client and see if it's happier (thanks!)
| |
| 14:11 | vagrantc has joined #ltsp | |
| 14:13 | <vandys> Darn, still hangs, last thing is IP-Config talking about a DHCP RARP. :-(
| |
| 14:15 | <Gadi> vandys: sounds like you are an old distro
| |
| 14:15 | you may not have the jme module
| |
| 14:15 | find /opt/ltsp/i386/lib/modules|grep jme
| |
| 14:15 | <vandys> I'm on karmic:
| |
| 14:16 | <Gadi> you cant be if you had to do /usr/lib/ltsp
| |
| 14:16 | <vandys> Modules in /lib/modules/2.6.31-20-generic/kernel/drivers/net/jme.ko, and the obvious earlier version paths?
| |
| 14:17 | <Gadi> is your chroot a karmic chroot?
| |
| 14:17 | <vandys> Although if I chroot it's feisty? I certainly didn't mean to hold back my LTSP version!
| |
| 14:17 | <Gadi> hehe
| |
| 14:17 | you need to build a karmic chroot
| |
| 14:17 | THEN, do the above commands
| |
| 14:18 | <vandys> Can you point me at a page to tell me how to make this happen? :->
| |
| 14:18 | <Gadi> well, is it okay to delete your current chroot?
| |
| 14:18 | <vandys> Sure, I think I have a couple tweaks which would be easy to save aside. Nothing major.
| |
| 14:19 | <Gadi> then, simply sudo rm -rf /opt/ltsp
| |
| 14:19 | er
| |
| 14:19 | sorry
| |
| 14:19 | sudo rm -rf /opt/ltsp/i386
| |
| 14:19 | (you can back up the /opt/ltsp/images/i386.img to i386-feisty.img to be safe
| |
| 14:19 | and then, simply: sudo ltsp-build-client --arch i386
| |
| 14:20 | if your server is karmic, you should wind up with a karmic chroot
| |
| 14:20 | <vandys> Ah, I'd wondered about that command. I have lots of disk space, so I'll just mv it aside and give that command a spin. Thanks MUCH!
| |
| 14:35 | efra_ has joined #ltsp | |
| 14:35 | efra has quit IRC | |
| 14:39 | MorpheusDe has quit IRC | |
| 14:43 | efra_ has quit IRC | |
| 14:45 | efra_ has joined #ltsp | |
| 14:56 | <alkisg> $%(*#(%*#$% FINALLY a whole day to find out that ck-launch-session is broken, and su doesn't work for display sessions
| |
| 14:57 | I got a python script that properly sets up a CK session.... now to adapt it for ldm/fat clients etc...
| |
| 15:02 | efra__ has joined #ltsp | |
| 15:05 | efra_ has quit IRC | |
| 15:14 | etyack has left #ltsp | |
| 15:24 | vandys has quit IRC | |
| 15:30 | Gadi has left #ltsp | |
| 15:32 | efra_ has joined #ltsp | |
| 15:33 | efra__ has quit IRC | |
| 15:41 | scottmaccal has quit IRC | |
| 15:44 | efra_ is now known as efra | |
| 15:47 | ninou has joined #ltsp | |
| 15:54 | ninou has quit IRC | |
| 15:58 | mgariepy has quit IRC | |
| 15:59 | japerry has quit IRC | |
| 16:07 | bobby_C has joined #ltsp | |
| 16:11 | etyack has joined #ltsp | |
| 16:11 | etyack has left #ltsp | |
| 16:29 | Ahmuck has joined #ltsp | |
| 16:29 | <Ahmuck> hi. i did an update 4-5 days ago, and now have no login on gdm
| |
| 16:29 | just a blank box
| |
| 16:29 | any ideas?
| |
| 16:31 | cascade has joined #ltsp | |
| 16:32 | <alkisg> gdm on the clients?
| |
| 16:34 | miguel has joined #ltsp | |
| 16:34 | miguel is now known as Guest78295 | |
| 16:35 | cascade has quit IRC | |
| 16:40 | <Ahmuck> server
| |
| 16:40 | hrm, not tried the clients yet
| |
| 16:40 | mgariepy has joined #ltsp | |
| 16:41 | <alkisg> So it isn't an ltsp-related problem, right?
| |
| 16:43 | <Ahmuck> i'm guessing so, but not sure. i did "sudo aptitude update" and then "sudo aptitude safe-upgrade" and upon reboot had no way to login. background, accesability icon, and a black box where the login normally is
| |
| 16:44 | <alkisg> Goodnight all
| |
| 16:45 | alkisg has quit IRC | |
| 16:48 | laron has joined #ltsp | |
| 16:48 | <laron> Having trouble making apps run locally on clients
| |
| 16:49 | When I add the l add the lts.conf file
| |
| 16:49 | and rebuild the image, everything works fine
| |
| 16:49 | except, it wont allow any users to login
| |
| 16:52 | <Ahmuck> started when?
| |
| 16:53 | <laron> im sorry?
| |
| 16:53 | <Ahmuck> when did it start? new install or just recently?
| |
| 16:53 | <laron> new install
| |
| 16:54 | johnny has left #ltsp | |
| 16:58 | Guest78295 has quit IRC | |
| 17:00 | pmatulis has quit IRC | |
| 17:02 | laron has quit IRC | |
| 17:05 | Selveste1 has quit IRC | |
| 17:16 | mikkel has quit IRC | |
| 17:24 | efra has quit IRC | |
| 17:27 | efra has joined #ltsp | |
| 17:37 | Lns has joined #ltsp | |
| 17:41 | waldo323 has joined #ltsp | |
| 17:44 | vagrantc has quit IRC | |
| 17:47 | johnny has joined #ltsp | |
| 18:05 | gavmac has joined #ltsp | |
| 18:09 | gavmac has quit IRC | |
| 18:16 | <Lns> just making sure, is there a bug anyone knows about about the gnome panel logout/reboot/shutdown/etc menu just going back to LDM after hitting shutdown?
| |
| 18:18 | efra_ has joined #ltsp | |
| 18:18 | efra has quit IRC | |
| 18:19 | efra_ has quit IRC | |
| 18:29 | pmatulis has joined #ltsp | |
| 18:30 | bobby_C has quit IRC | |
| 19:13 | rjune has joined #ltsp | |
| 19:18 | <johnny> Lns, isn't that what you want to happen?
| |
| 19:19 | <Lns> johnny: not if I hit shutdown....
| |
| 19:19 | wouldn't you think anybody would expect the client to ...umm...shut down when you hit...umm...shutdown? ;)
| |
| 19:19 | <johnny> Lns, that's not possible
| |
| 19:19 | ogra_cmpc has quit IRC | |
| 19:20 | <Lns> johnny: how about if its a fat client?
| |
| 19:20 | ogra_cmpc has joined #ltsp | |
| 19:21 | <johnny> Lns, .. more possible..
| |
| 19:21 | but who knows what happens with ldm in that regard..
| |
| 19:21 | Lns, you'd have to talk to alkisg about that
| |
| 19:21 | <Lns> I will real soon hopefully =)
| |
| 19:21 | just built up a saucy atom 330 tc
| |
| 19:22 | building a fat client chroot as i type
| |
| 19:24 | japerry has joined #ltsp | |
| 19:26 | <Lns> cheers all
| |
| 19:26 | Lns has quit IRC | |
| 19:28 | try2free has joined #ltsp | |
| 19:30 | tthorb has joined #ltsp | |
| 19:33 | tthorb_ has quit IRC | |
| 19:39 | waldo323 has quit IRC | |
| 19:42 | try2free has left #ltsp | |
| 20:00 | pmatulis has quit IRC | |
| 20:07 | <Ahmuck> johnny: actually, i expected the thin client to shutdown as well when selecting shutdown
| |
| 20:10 | Egyptian[Home] has quit IRC | |
| 20:13 | Egyptian[Home] has joined #ltsp | |
| 20:23 | slidesinger has joined #ltsp | |
| 20:40 | ltspbot has joined #ltsp | |
| 20:44 | shawnp0wers has joined #ltsp | |
| 20:44 | map7 has joined #ltsp | |
| 20:49 | ltspbot has joined #ltsp | |
| 21:02 | mgariepy has quit IRC | |
| 21:25 | map7 has quit IRC | |
| 21:26 | map7 has joined #ltsp | |
| 21:37 | daya has joined #ltsp | |
| 21:40 | jammcq has joined #ltsp | |
| 21:40 | <jammcq> hellooooooo
| |
| 21:41 | johnny has left #ltsp | |
| 21:57 | <rjune> !j
| |
| 21:57 | <ltspbot> rjune: "j" :: jammcq!!!!!!!!!!!!!!!!!!!!!!!
| |
| 21:57 | <rjune> How you doing mate?
| |
| 21:57 | ogra_cmpc has quit IRC | |
| 21:59 | ogra_cmpc has joined #ltsp | |
| 22:04 | johnny has joined #ltsp | |
| 22:12 | <jammcq> rjune: hey
| |
| 22:12 | i'm doing well. how's things in Indiana?
| |
| 22:18 | otavio has quit IRC | |
| 22:29 | <rjune> been better been worse
| |
| 22:45 | otavio has joined #ltsp | |
| 23:01 | alkisg has joined #ltsp | |
| 23:07 | shawnp0wers has quit IRC | |
| 23:21 | FinnTux has quit IRC | |
| 23:21 | FinnTux has joined #ltsp | |
| 23:30 | <johnny> hmm.. so i just put a new terminal in place with 512mb ram alkisg ..
| |
| 23:30 | now all 3 have 512mb
| |
| 23:30 | one problem.. it seems stuck at 800x600 by default
| |
| 23:31 | it's got some nforce chipset in it
| |
| 23:31 | the equivalent of a geforce4 (nv18)
| |
| 23:32 | <alkisg> Lucid? Or Karmic?
| |
| 23:32 | <johnny> karmic
| |
| 23:33 | i'll be upgrading to lucid on that machine once it is actually released
| |
| 23:33 | wanna let the karmic fall out continue
| |
| 23:33 | before causing more cascading problems :)
| |
| 23:33 | <alkisg> That would be a driver problem, i.e. not detecting monitor EDID
| |
| 23:34 | <johnny> yes.. that seems to be the case
| |
| 23:34 | <alkisg> You could either install the nvidia proprietary drivers on the chroot, *IF* you only have nvidia clients, or you could force X_VERTREFRESH and X_HORZSYNC
| |
| 23:35 | alexqwesa__ has joined #ltsp | |
| 23:35 | alexqwesa_ has quit IRC | |
| 23:37 | <johnny> well.. lucid uses nouveau
| |
| 23:37 | so.. it should work fine when i upgrade
| |
| 23:37 | <alkisg> Right, that's why I asked
| |
| 23:37 | <johnny> yeah.. i just can't do that yet
| |
| 23:38 | <alkisg> But it's easy to set X_HORZSYNC, so you could try that until you do
| |
| 23:38 | <johnny> what should i have in total
| |
| 23:38 | it's an lcd screen
| |
| 23:38 | <alkisg> What's the correct resolution?
| |
| 23:38 | <johnny> hmm. good question..
| |
| 23:39 | i'm not sure what is correct :)
| |
| 23:39 | <alkisg> The following should give you 1024x768
| |
| 23:39 | X_HORZSYNC=30.0-61.0
| |
| 23:39 | X_VERTREFRESH=50.0-88.0
| |
| 23:39 | <johnny> yeah.. that sounds good enough for now
| |
| 23:39 | do i have to set CONFIGURE_X too?
| |
| 23:39 | <alkisg> No
| |
| 23:40 | <johnny> that sounds good.. thanks alkisg
| |
| 23:40 | set
| |
| 23:40 | i'm not there..
| |
| 23:40 | so i hope it comes up tomorrow :)
| |
| 23:40 | <alkisg> How's the overall performace?
| |
| 23:40 | *n
| |
| 23:41 | <johnny> i didn't get a chance to set the fat client stuff up yet
| |
| 23:41 | <alkisg> k
| |
| 23:42 | <johnny> i'm going to do that tomorrow
| |
| 23:42 | i thought i'd have time tonight
| |
| 23:42 | but.. no dice
| |
| 23:42 | in fact.. i might have to wait til thursday
| |
| 23:42 | since i'm working there tomorrow
| |
| 23:42 | but only have to do tech for an event on thursda..
| |
| 23:43 | <alkisg> I've pin-pointed down the CK problem, I hope I'll commit a workaround for it tomorrow: https://bugs.launchpad.net/ubuntu/+source/ltsp/+bug/535521
| |
| 23:43 | <johnny> swell
| |
| 23:43 | just don't bet scared to talk to upstream directly :)
| |
| 23:43 | the gnome folks are pretty cool
| |
| 23:44 | i think i'm going try to upgrade my own laptop to fedora 13 alpha in the next few days
| |
| 23:44 | <alkisg> Is there a separate "freedesktop" team for such stuff, or the half of them are gnome team members as well?
| |
| 23:45 | <johnny> i probably wouldn't have upgraded to karmic if i didn't also have to upgrade to karmic before lucid..
| |
| 23:45 | <alkisg> I mean, maybe I should be talking to #freedesktop instead of #gnome-hackers...
| |
| 23:45 | <johnny> well for some of of yes
| |
| 23:45 | just depends on who hangs out in the freedesktop channel..
| |
| 23:45 | not sure if davidz for example is often in freedesktop or not
| |
| 23:46 | <alkisg> Well yesterday I only got a responce from #gnome-hackers, so... :D
| |
| 23:46 | <johnny> exactly..
| |
| 23:46 | :)
| |
| 23:46 | the gnome people are pretty awesome
| |
| 23:46 | hmm.. if mythbuntu had a powerpoint viewer..
| |
| 23:47 | ah.. also need wireless.. can't forget that
| |
| 23:47 | my next project is to setup a computer hooked up to the project
| |
| 23:47 | projector*
| |
| 23:47 | the hard part is the wiring
| |
| 23:47 | since the projector is mounted on the ceiling
| |
| 23:48 | ah..
| |
| 23:49 | alkisg, wanna see?
| |
| 23:49 | <alkisg> Sure
| |
| 23:49 | You could have a small thin client up there ;)
| |
| 23:50 | <johnny> http://www.flickr.com/photos/radarredux/4402142663/
| |
| 23:50 | the screen goes in front of the bookcases
| |
| 23:50 | i have considered a small thin client..
| |
| 23:50 | but that means..
| |
| 23:51 | i hae to get up on a stool to put in a flash drive
| |
| 23:51 | <alkisg> Nice place you got there :)
| |
| 23:51 | <johnny> and it means it's impossible to plug up a wired keyboard
| |
| 23:52 | http://www.flickr.com/photos/43517501@N00/4094307906/
| |
| 23:53 | that's a pic from the outside..
| |
| 23:53 | <alkisg> Well you could put a vga cable... I think they go up to 5m
| |
| 23:53 | <johnny> sure..
| |
| 23:53 | but then i have to be careful of the window
| |
| 23:54 | i want it on the side of the window, not the bar
| |
| 23:54 | <alkisg> Ah
| |
| 23:54 | <johnny> because the bathroom is over on the bar side
| |
| 23:54 | don't want people interrupting the speaker/movie/presentation
| |
| 23:54 | <alkisg> There are also wireless vga broadcasting devices, but they'd cost some money
| |
| 23:54 | mordocai has joined #ltsp | |
| 23:54 | <johnny> yeah.. probably more than i wanna spend
| |
| 23:54 | there are some big posts in the windows
| |
| 23:55 | but that just means the cable must be a bit longer
| |
| 23:55 | i wish vga cables were a bit thinner..
| |
| 23:55 | that'd help
| |
| 23:55 | daya has quit IRC | |
| 23:56 | daya has joined #ltsp | |
| 23:57 | <johnny> alkisg, .. yeah.. we have 3 spaces now..
| |
| 23:57 | but we don't have any tech stuff at the free school
| |
| 23:57 | it has 3 big windows
| |
| 23:57 | and they aren't barred in any way..
| |
| 23:57 | and the place isn't occupied at all times
| |
| 23:57 | also.. every instructor has a set of keys
| |
| 23:57 | so..
| |
| 23:58 | would be easy for stuff to disappear in a number of ways
| |
| 23:58 | <alkisg> Yeah, things would need to be bolted down
| |
| 23:59 | <johnny> yeah.. but.. that kinda makes it hard when only 1/20 classes need any technology
| |
| 23:59 | would take up too much space
| |