IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 9 September 2016 (all times are UTC)
| 00:27 | kjackal has left IRC (kjackal!~quassel@97.93.161.2, Quit: No Ping reply in 180 seconds.) | |
| 00:30 | kjackal has joined IRC (kjackal!~quassel@97.93.161.2) | |
| 01:18 | kjackal has left IRC (kjackal!~quassel@97.93.161.2, Ping timeout: 244 seconds) | |
| 01:28 | pppingme has left IRC (pppingme!~pppingme@unaffiliated/pppingme, Ping timeout: 244 seconds) | |
| 01:35 | pppingme has joined IRC (pppingme!~pppingme@unaffiliated/pppingme) | |
| 02:02 | Freejack has left IRC (Freejack!~Freejack@unaffiliated/freejack, Ping timeout: 240 seconds) | |
| 02:07 | Freejack has joined IRC (Freejack!~Freejack@unaffiliated/freejack) | |
| 03:52 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
| 04:40 | vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Ping timeout: 244 seconds) | |
| 05:08 | ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz) | |
| 05:36 | cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 240 seconds) | |
| 05:41 | cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg) | |
| 06:03 | kjackal has joined IRC (kjackal!~quassel@97.93.161.2) | |
| 06:22 | cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 252 seconds) | |
| 06:27 | mikkel has joined IRC (mikkel!~mikkel@mail.dlvs.dk) | |
| 06:28 | cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg) | |
| 06:34 | cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 264 seconds) | |
| 06:36 | kjackal has left IRC (kjackal!~quassel@97.93.161.2, Quit: No Ping reply in 180 seconds.) | |
| 06:40 | <alkisg> !nbd-client
| |
| 06:40 | <ltsp> nbd-client: To try mounting the NBD image from the client initramfs: nbd-client 192.168.67.1 -N /opt/ltsp/i386 /dev/nbd0
| |
| 06:42 | kjackal has joined IRC (kjackal!~quassel@97.93.161.2) | |
| 06:52 | kjackal has left IRC (kjackal!~quassel@97.93.161.2, Ping timeout: 244 seconds) | |
| 07:29 | robb_nl has joined IRC (robb_nl!~robb_nl@ip-62-235-225-180.dsl.scarlet.be) | |
| 07:35 | robb_nl has left IRC (robb_nl!~robb_nl@ip-62-235-225-180.dsl.scarlet.be, Ping timeout: 250 seconds) | |
| 07:47 | robb_nl has joined IRC (robb_nl!~robb_nl@62.235.102.130) | |
| 09:44 | forum has joined IRC (forum!~Icedove@93-82-98-103.adsl.highway.telekom.at) | |
| 09:51 | vervelak has joined IRC (vervelak!~vervelak@139.91.200.185) | |
| 09:51 | forum has left IRC (forum!~Icedove@93-82-98-103.adsl.highway.telekom.at, Quit: forum) | |
| 11:08 | JuJuBee has left IRC (JuJuBee!~knichel@24.148.115.153, Quit: Leaving) | |
| 11:52 | sebd has left IRC (sebd!~seb@aditu.ldd.fr, Ping timeout: 244 seconds) | |
| 11:52 | sebd has joined IRC (sebd!~seb@85.31.145.12) | |
| 11:55 | mgariepy has joined IRC (mgariepy!~mgariepy@ubuntu/member/mgariepy) | |
| 11:56 | Faith has joined IRC (Faith!~paty_@unaffiliated/faith) | |
| 11:56 | adrianorg has left IRC (adrianorg!~adrianorg@177.204.159.246.dynamic.adsl.gvt.net.br, Ping timeout: 240 seconds) | |
| 11:58 | adrianorg has joined IRC (adrianorg!~adrianorg@201.22.230.89) | |
| 13:03 | Daniel_n has joined IRC (Daniel_n!c325ec18@gateway/web/freenode/ip.195.37.236.24) | |
| 13:04 | <Daniel_n> hi all.. can someone help me with a lisp cluster?
| |
| 13:04 | *ltsp cluster
| |
| 13:07 | is there a possibility to make a ltsp cluster high available, so that active terminal session (x over ssh) won't disconnect? or just loose their connection for a short period of time and auto reconnect without rebooting the thin client?
| |
| 13:10 | cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg) | |
| 13:25 | <Daniel_n> is there a possibility to make a ltsp cluster high available, so that active terminal session (x over ssh) won't disconnect? or just loose their connection for a short period of time and auto reconnect without rebooting the thin client?
| |
| 13:25 | kjackal has joined IRC (kjackal!~quassel@97.93.161.2) | |
| 13:37 | <||cw> Daniel_n: you'd need to make each individual program you run on said cluster to be cluster-aware so that it can be highly available. so no. your best option for that is virtialization with some type of continuous availability system. I haven't kept up with that, but vmware's FT is the one I'm aware of, and it's expensive, and has limitations that it not worth it for most cases
| |
| 13:39 | you could certainly use some load balancing in front of an ltsp cluster to make client pick a server, but if one goes down all the client sessions will be started over on a new server
| |
| 13:58 | forum has joined IRC (forum!~Icedove@93-82-98-103.adsl.highway.telekom.at) | |
| 13:59 | forum has left IRC (forum!~Icedove@93-82-98-103.adsl.highway.telekom.at, Client Quit) | |
| 13:59 | ben_roose has joined IRC (ben_roose!~roose@roose.cs.wichita.edu) | |
| 14:01 | JuJuBee has joined IRC (JuJuBee!~knichel@24-148-115-153.ip.mhcable.com) | |
| 14:04 | mikkel has left IRC (mikkel!~mikkel@mail.dlvs.dk, Quit: Leaving) | |
| 14:19 | <JuJuBee> SO I biuld an image yesterday and am trying to use nano in the chroot and it is not installed. When I try apt-get install nano, I get t his.... Reading package lists... Done
| |
| 14:19 | Building dependency tree
| |
| 14:19 | Reading state information... Done
| |
| 14:19 | E: Unable to locate package nano
| |
| 14:19 | I copied the sources.list from the server into the chroot
| |
| 14:21 | <Daniel_n> ||cw: isn't it possible to reconnect the sockets to the new cluster node after the initial socket connection gets lost?
| |
| 14:22 | so there is just a little gap while reconnecting
| |
| 14:24 | <alkisg> Daniel_n: your main worries are for issues on the client, on the server, or both? If it's for issues on the client (e.g. reboots, disconnects) then you can also have a look at x2go, which resumes sessions properly
| |
| 14:24 | JuJuBee: run apt update
| |
| 14:27 | <JuJuBee> alkisg, omg, totally forgot to update db... Thanks.
| |
| 14:28 | <Daniel_n> alkisg: e.g. one thin client is connected to cluster node-1 and node-2 is passive... the x over ssh session of client-1 ist started on node-1.. now i perform a failover (reboot node-1 or stop pacemaker/corosync) the resources(DRBD, IPaddr) are switching over to node-2.. while doing this, i watch the netstat -tupn ... there i see, that the socket connection is disconnected from node-1 and of course ist not connected new to node-2... n
| |
| 14:28 | <||cw> Daniel_n: you do realize that reconnecting to a new server means starting up a new clean login session right?
| |
| 14:28 | <Daniel_n> client, i am not able
| |
| 14:28 | yes
| |
| 14:29 | <||cw> ok, and the IP gets moved to node2?
| |
| 14:29 | <Daniel_n> yes
| |
| 14:29 | <alkisg> Daniel_n: what kind of availabily is that, if you lose your running programs?
| |
| 14:30 | <||cw> it's likely confusing the client, you'd need to somehow bootstrap the nbd session on the server
| |
| 14:30 | or switch to nfs
| |
| 14:31 | <alkisg> I'm not aware of any technology that allows someone to have programs running to serverA and then have them to serverB automatically, except for virtual machine beaming, and that of course requires that you know that serverA is going to have a down time
| |
| 14:31 | <JuJuBee> How come tab line completion does not work in the chroot?
| |
| 14:31 | <||cw> you're basically ripping the disk out of the client and not replacing it, leaving it without only knowing what it has cached
| |
| 14:31 | <Daniel_n> mhh :-D
| |
| 14:31 | <alkisg> nbd is read-only, it's not that hard to make it reconnect to another server
| |
| 14:32 | <||cw> alkisg: but is going to be seamless?
| |
| 14:32 | <alkisg> But running programs with data in RAM... that's not easily accomplished. How can you read RAM after crash?
| |
| 14:32 | ||cw: yes, there are ways to achieve that
| |
| 14:32 | <Daniel_n> yes i think thats the main prob..
| |
| 14:32 | <||cw> I've seen clients have to restarted just from having their network cable unplugged for a bit
| |
| 14:32 | <alkisg> Daniel_n: and that's why it's not really doable, and you should be looking to implement something else :)
| |
| 14:33 | ||cw: yes the stock nbd-client had issues like that. Not anymore, it can work after disconnecting the cable, but it doesn't support switching servers, as other clients do.
| |
| 14:34 | <Daniel_n> but the loss of the connection is not the main problem.. it would be ok, if the screen blacks out or something like that..
| |
| 14:35 | maybe sth like hyper-v failover clustering is easier to implement :-D
| |
| 14:35 | but what HA is it when the session gets aborted ?!
| |
| 14:35 | the servers are HA, but the sessions aren't..
| |
| 14:36 | <Hyperbyt1> mhmmm
| |
| 14:36 | Hyperbyt1 is now known as Hyperbytge | |
| 14:36 | Hyperbytge is now known as Hyperbyte | |
| 14:37 | <Daniel_n> so if people are working at a company with terminal servers.. and one of the cluster nodes goes down.. the employees get disconnected.. i don't think that this is a sort of HA :-D
| |
| 14:37 | <alkisg> Daniel_n: if a program crashes on one pc, you can't resume it on another pc, because there's no technology that can read crashed ram
| |
| 14:37 | <Daniel_n> yes thats of course right ;-)
| |
| 14:37 | <alkisg> There are some programs or protocols that are stateless, you can only do that for such programs
| |
| 14:38 | So you can't resume sessions from a crashed server, anywhere else, either on that server or on another server
| |
| 14:38 | You'll have to start a new session
| |
| 14:39 | <Daniel_n> ok that makes sense... so the only way, to make a terminal server session HA is virtualization e.g hyper-v or kvm
| |
| 14:39 | yes that would be ok also :-D
| |
| 14:39 | <alkisg> No, you cannot do that with hyper-v or kvm either
| |
| 14:39 | <Daniel_n> so how could i achieve it? :-D
| |
| 14:39 | <alkisg> You can't
| |
| 14:40 | You need to look at other things that can be made more available
| |
| 14:40 | <Daniel_n> isnt there no way to make a HA ltsp cluster?!
| |
| 14:40 | <alkisg> Like, quickly starting a new session, or minimizing the client down time etc
| |
| 14:41 | <Daniel_n> yes that'd be a nice solution
| |
| 14:41 | <alkisg> Daniel_n: I don't know of any way at all in computers to make the HA that you are asking, with or without ltsp
| |
| 14:41 | So you need to minimize your expectations from computers
| |
| 14:41 | <Daniel_n> okay... forget rescuing crashed sessions :-D
| |
| 14:41 | <alkisg> Right
| |
| 14:42 | After that, there are many ways to have more availability
| |
| 14:42 | <Daniel_n> okay.. then: how could i achieve quickly restarting new sessions after a failover in a ltsp cluster? ;-)
| |
| 14:42 | <alkisg> What part failed in your example?
| |
| 14:43 | <Daniel_n> power went off :-D
| |
| 14:43 | <alkisg> In the server or in the client?
| |
| 14:43 | <Daniel_n> cluster node-1
| |
| 14:43 | <alkisg> The cluster node is a server or a client?
| |
| 14:43 | <Daniel_n> server
| |
| 14:43 | sorry :-D
| |
| 14:43 | <alkisg> The client is a thin client or a fat client?
| |
| 14:43 | <||cw> what your wanting I think is Continuous Availability, not High Availability. High just means it's down for as little as possible
| |
| 14:43 | <Daniel_n> thin
| |
| 14:44 | <alkisg> OK then you press alt+ctrl+backspace on the client to kill the session, if it hasn't been already killed,
| |
| 14:44 | and the display manager (ldm) selects the next available server to login to
| |
| 14:44 | <||cw> for Continuous you need something to synchronously copy all process, ram, and state info to the standby, which often requires the application to be aware of it
| |
| 14:46 | <Daniel_n> ||cw: yes you are right.. that was in my mind.. but i think it is to hard to make it
| |
| 14:46 | alkisg: this sounds like a nice solution...
| |
| 14:47 | <||cw> it is quite hard. which is why vmware can charge so much for it
| |
| 14:47 | <Daniel_n> just for better understanding... my "thin client" is a raspberry pi 3 with some data on the sd card.. (this is a thin client, isnt it? :-D )
| |
| 14:48 | <||cw> thin means the user session is on the server, fat means the user session is on the client. whether or not there's a local disk isn't really important
| |
| 14:48 | <alkisg> A raspberry can be either a thin client (processes running on the server) or a fat (locally)
| |
| 14:49 | <Daniel_n> okay.. thin client
| |
| 14:49 | <||cw> also pi3 can have a local disk or not. pi2 requires a local disk
| |
| 14:50 | <alkisg> ||cw: ? do pi's have that difference?!
| |
| 14:50 | Don't pi3's also require a local sd card?
| |
| 14:50 | <||cw> yes, only the pi3 supports native network boot
| |
| 14:50 | <Daniel_n> pxe also
| |
| 14:50 | kjackal has left IRC (kjackal!~quassel@97.93.161.2, Ping timeout: 244 seconds) | |
| 14:50 | <Daniel_n> pi2 doest support pxe i think
| |
| 14:50 | <||cw> https://www.raspberrypi.org/blog/pi-3-booting-part-ii-ethernet-all-the-awesome/
| |
| 14:51 | <alkisg> They managed to implement proxy dhcp in rom? coool!
| |
| 14:51 | (client, of course)
| |
| 14:52 | Is that for new Pi3's that are going to be shipped in the f uture?
| |
| 14:52 | Or is there a way to flash existing roms?
| |
| 14:52 | <||cw> I think all pi3s
| |
| 14:52 | <Daniel_n> when i turn off the ltsp server which the client is connected to (ipaddress switches over to node-2), anyway there is a gui on the thin client.. i see my terminals which i opened before... ping is working... but if i want to save e.g. an open file (nano) says "socket is not connected"..
| |
| 14:52 | <||cw> https://www.raspberrypi.org/documentation/hardware/raspberrypi/bootmodes/net_tutorial.md
| |
| 14:53 | <alkisg> Cool. /me waves, later...
| |
| 14:53 | <Daniel_n> if i reboot the pi, it will connect to the "new" ltsp and it works fine
| |
| 14:54 | <||cw> Daniel_n: you mean localapp terminals?
| |
| 14:55 | or? save where?
| |
| 14:55 | <Daniel_n> no saving the file in my home directory (ltsp user chrooted)
| |
| 14:57 | <||cw> more specific....
| |
| 14:57 | Daniel_n has left IRC (Daniel_n!c325ec18@gateway/web/freenode/ip.195.37.236.24, Quit: Page closed) | |
| 14:57 | Daniel_n has joined IRC (Daniel_n!c325ec18@gateway/web/freenode/ip.195.37.236.24) | |
| 14:57 | <||cw> more specific....
| |
| 14:57 | <Daniel_n> ok
| |
| 14:57 | <||cw> like, sshfs mounted local?
| |
| 14:57 | and localapps or not?
| |
| 14:57 | <Daniel_n> thin client connects to the ltsp server (node-1)
| |
| 14:58 | <||cw> shutting off the server means your screen would freeze otherwise
| |
| 14:59 | <Daniel_n> ltsp node-1 has mounted a DRBDevice (on which the /home directory of all ltsp users are)
| |
| 14:59 | when turning off node-1 the IP and the DRBDevice were switched over to node-2 directly..
| |
| 14:59 | and the pi doesn't freeze
| |
| 15:01 | <||cw> are you sure you're running thin?
| |
| 15:01 | <Daniel_n> but if i then try to open a new terminal.. the screen went black (but i have a mouse :-D )
| |
| 15:01 | http://pinet.org.uk
| |
| 15:01 | <||cw> mouse cursor is rendered locally
| |
| 15:02 | <Daniel_n> i think its thin
| |
| 15:04 | i am using PiNet, based on ltsp
| |
| 15:04 | on the sd card are just some boot files and config files
| |
| 15:05 | about 30MB
| |
| 15:05 | <||cw> that doens't mean it's not a fat client
| |
| 15:06 | the difference is that thin clients the apps run on the server, so if the server crashes, so do the apps. fat clients are still diskless, but the apps run on the client, which means anything in ram keeps working
| |
| 15:06 | <Daniel_n> mhh but i am only using the terminal...
| |
| 15:07 | and the terminal window stays open
| |
| 15:07 | <||cw> run a ps on the server, if you don't see the processes that are open on the client, then you're runnign fat
| |
| 15:07 | for thin, the window would stay open, but the terminal would no longer respond
| |
| 15:07 | <Daniel_n> ok i see.. i check that
| |
| 15:08 | <||cw> the server is only sending X drawing commands to the client, if it doens't send a new command, what's on the screen stays there
| |
| 15:10 | <Daniel_n> on the raspian image is minecraft.. so if i start it, it should appear in server ps right?
| |
| 15:11 | <||cw> ok, if minecraft is running well, you are almost certainly running fat mode, or at least minecraft is running as a localapp
| |
| 15:17 | <Daniel_n> mhh :-D i don't see any of my opened apps :-D neither server ps or client ps
| |
| 15:18 | adrianorg has left IRC (adrianorg!~adrianorg@201.22.230.89, Ping timeout: 264 seconds) | |
| 15:18 | kjackal has joined IRC (kjackal!~quassel@97.93.161.2) | |
| 15:19 | adrianorg has joined IRC (adrianorg!~adrianorg@177.204.76.55.dynamic.adsl.gvt.net.br) | |
| 15:20 | <Daniel_n> okay... forgot ps -a ;)
| |
| 15:25 | but if i open minecraft i don't see it in server or client ps -a
| |
| 15:26 | ||cw: at client ps -a there are only some ltsp-localappsd and ltsp-remoteapps ldm sh ssh sh ps kinit xautowait processes
| |
| 15:27 | <||cw> try ps aux
| |
| 15:28 | forum has joined IRC (forum!~Icedove@93-82-98-103.adsl.highway.telekom.at) | |
| 15:31 | <Daniel_n> |cw: ps aux doesn't show anything at the pi
| |
| 15:31 | ||cw: ps aux doesn't show anything at the pi
| |
| 15:32 | at the server i see the sessions of my ltsp user
| |
| 15:34 | <||cw> ok, so then on the client, when the server stops, these apps stop accepting input right?
| |
| 15:34 | <Daniel_n> yes
| |
| 15:34 | <||cw> then I don't understand what issue you described
| |
| 15:35 | <Daniel_n> :-D
| |
| 15:36 | it has to be fat client.. i turned off both servers... minecraft is still running
| |
| 15:37 | <||cw> or localapps...
| |
| 15:37 | maybe ps auxww ?
| |
| 15:37 | which should show the full command line, possibly needed to see localapps
| |
| 15:38 | I don't have any localapp configs here so I'm nto sure
| |
| 15:39 | <Daniel_n> ps aux ont shows warning: bad ps syntax
| |
| 15:40 | only
| |
| 15:44 | kjackal has left IRC (kjackal!~quassel@97.93.161.2, Ping timeout: 244 seconds) | |
| 15:48 | kjackal has joined IRC (kjackal!~quassel@97.93.161.13) | |
| 15:48 | Royale_AJS has joined IRC (Royale_AJS!~Royale_AJ@24-236-193-237.dhcp.aldl.mi.charter.com) | |
| 15:53 | kjackal has left IRC (kjackal!~quassel@97.93.161.13, Ping timeout: 240 seconds) | |
| 16:02 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
| 16:10 | kjackal has joined IRC (kjackal!~quassel@97.93.161.13) | |
| 16:14 | robb_nl has left IRC (robb_nl!~robb_nl@62.235.102.130, Ping timeout: 244 seconds) | |
| 16:30 | Daniel_n has left IRC (Daniel_n!c325ec18@gateway/web/freenode/ip.195.37.236.24, Quit: Page closed) | |
| 17:09 | kjackal has left IRC (kjackal!~quassel@97.93.161.13, Ping timeout: 255 seconds) | |
| 17:12 | kjackal has joined IRC (kjackal!~quassel@97.93.161.2) | |
| 17:25 | kjackal has left IRC (kjackal!~quassel@97.93.161.2, Quit: No Ping reply in 210 seconds.) | |
| 17:30 | kjackal has joined IRC (kjackal!~quassel@97.93.161.2) | |
| 17:43 | kjackal has left IRC (kjackal!~quassel@97.93.161.2, Quit: No Ping reply in 180 seconds.) | |
| 17:45 | kjackal has joined IRC (kjackal!~quassel@97.93.161.13) | |
| 17:46 | JuJuBee has left IRC (JuJuBee!~knichel@24-148-115-153.ip.mhcable.com, Quit: Leaving) | |
| 18:22 | Claudio_ has joined IRC (Claudio_!02e0a8e6@gateway/web/freenode/ip.2.224.168.230) | |
| 18:23 | <Claudio_> I'm experience of very slow client on edubuntu.
| |
| 18:24 | Somebody can help me ?
| |
| 18:26 | <vagrantc> Claudio_: if your hardware is reasonable, fat clients may work.
| |
| 18:26 | Claudio_: other options are to use LDM_DIRECTX=true
| |
| 18:27 | !directx | echo Claudio_
| |
| 18:27 | <ltsp> Claudio_ directx: If you want to turn off encryption to the workstation, which will speed up performance at the cost of security, you need to specify LDM_DIRECTX=True in your lts.conf file. Type !docs in IRC to get a link to the current documentation
| |
| 18:30 | <Claudio_> I can see on server that after any click on client the network of server work at full load for many seconds... and the client show the new windows in5-10 second.
| |
| 18:31 | My server is a core i5 with 4cores and 4GB of Ram.
| |
| 18:32 | and clients are dual core with almost 1GB of ram.
| |
| 18:33 | Now I'm upgrading the server to version 16.04
| |
| 18:41 | <vagrantc> sounds like you should be able to do fat clients
| |
| 18:41 | !fat
| |
| 18:41 | <ltsp> I do not know about 'fat', but I do know about these similar topics: 'fatclients', 'fatclient-printers', 'fat-printers', 'epoptes-fat-clients', 'fat-internal-disks', 'fat-passwd', 'fat-sudo'
| |
| 18:41 | <vagrantc> !fatclients | echo Claudio_
| |
| 18:41 | <ltsp> Claudio_ fatclients: You may find some info about the Ubuntu/LTSP implementation of fat clients at https://help.ubuntu.com/community/UbuntuLTSP/FatClients
| |
| 18:41 | <vagrantc> i don't know ifk you are already
| |
| 18:42 | Claudio_: are all your clients slow, or just a particular client?
| |
| 18:45 | <Claudio_> For me is the first experience.... Trying with two different client but same result...
| |
| 18:50 | Royale_AJS has left IRC (Royale_AJS!~Royale_AJ@24-236-193-237.dhcp.aldl.mi.charter.com, Quit: My Mac has gone to sleep. ZZZzzz…) | |
| 18:51 | cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 252 seconds) | |
| 18:56 | cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg) | |
| 19:09 | K0HAX has left IRC (K0HAX!~michael@shellhost.home.englehorn.com, Quit: ZNC 1.7.x-git-645-c28a290 - http://znc.in) | |
| 19:52 | K0HAX has joined IRC (K0HAX!~michael@shellhost.home.englehorn.com) | |
| 20:28 | <alkisg> !ltsp-pnp | echo Claudio_:
| |
| 20:28 | <ltsp> Claudio_: ltsp-pnp: ltsp-pnp is an alternative (upstream) method to maintain LTSP installations for thin and fat clients that doesn't involve chroots: https://help.ubuntu.com/community/UbuntuLTSP/ltsp-pnp
| |
| 20:28 | <alkisg> This is a good tutorial for a first experience with ltsp ^
| |
| 20:28 | Easy to set up and takes advantage of the client hardware so that it goes fast
| |
| 20:32 | vagrantc: ||cw told me that raspberry pi 3 supports pxe booting without an sd card! https://www.raspberrypi.org/documentation/hardware/raspberrypi/bootmodes/net.md
| |
| 20:32 | It even supports proxydhcp...
| |
| 20:35 | <vagrantc> wow!
| |
| 20:36 | i hadn't bothered with them yet, but if it supports network boot out of the box, that's something worth looking at
| |
| 20:40 | <alkisg> If only they were x86 and fast and supported flash and wine... :)
| |
| 20:41 | <||cw> sorta out of the box... you do have to boot off an SD once and set the network boot option
| |
| 20:42 | haven't you heard? flash is dead :P
| |
| 20:43 | <alkisg> Sure, except we have 10+ gb of flash-based educational software and no millions of $$ to hire programmers to rewrite them in html 5...
| |
| 20:43 | ||cw: set the network boot option where? does it have something like a cmos?
| |
| 20:44 | <||cw> the OTP
| |
| 20:44 | https://www.raspberrypi.org/documentation/hardware/raspberrypi/bootmodes/net_tutorial.md
| |
| 20:46 | TLDR; you add program_usb_boot_mode=1 to the config.txt and reboot, then vcgencmd otp_dump will show the updated setting, then you can power off and take the card out.
| |
| 20:46 | * alkisg can't find what the otp initials mean | |
| 20:46 | <||cw> that link isn' thte best written howto I've ever found, but I'm told it works
| |
| 20:47 | One Time Programmable
| |
| 20:47 | I don't know if that means you can't turn it back off
| |
| 20:49 | <vagrantc> hah, the tutorial explains how to configure for NFS root ...
| |
| 20:50 | <alkisg> I think it does mean that the bits can be written to only once, like the warranty bit
| |
| 20:50 | if so, once you go netbooting, you can't go back.... hehe nice way to push ltsp :P
| |
| 20:52 | <vagrantc> "sudo echo | sudo tee /etc/dnsmasq.conf" awesome.
| |
| 20:53 | * alkisg prefers to start the instructions with "sudo -i" :) | |
| 20:58 | Faith has left IRC (Faith!~paty_@unaffiliated/faith, Quit: Leaving) | |
| 21:16 | <||cw> is that first sudo even needed?
| |
| 21:16 | * vagrantc gives ||cw a prize! | |
| 21:17 | <||cw> yeah... like i said... not the best written
| |
| 21:17 | * vagrantc nods | |
| 21:18 | <||cw> looks fixed int he github version <alkisg> if so, once you go netbooting, you can't go back.... hehe nice way to push ltsp :P
| |
| 21:18 | er
| |
| 21:18 | https://github.com/raspberrypi/documentation/blob/master/hardware/raspberrypi/bootmodes/net_tutorial.md
| |
| 21:28 | * vagrantc wonders if the network booting supports initrd | |
| 21:56 | Royale_AJS has joined IRC (Royale_AJS!~Royale_AJ@24-236-193-237.dhcp.aldl.mi.charter.com) | |
| 22:14 | Royale_AJS has left IRC (Royale_AJS!~Royale_AJ@24-236-193-237.dhcp.aldl.mi.charter.com, Quit: Royale is leaving) | |
| 22:20 | ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Quit: Leaving) | |
| 23:31 | ben_roose has left IRC (ben_roose!~roose@roose.cs.wichita.edu, Remote host closed the connection) | |