IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 12 February 2023 (all times are UTC)
| 02:05 | vagrantc has left IRC (vagrantc!~vagrant@2600:3c01:e000:21:7:77:0:20, Quit: leaving) | |
| 02:35 | vagrantc has joined IRC (vagrantc!~vagrant@2600:3c01:e000:21:7:77:0:40) | |
| 03:45 | wyre is now away: Auto away at Sun Feb 12 03:45:13 2023 UTC | |
| 04:38 | vagrantc has left IRC (vagrantc!~vagrant@2600:3c01:e000:21:7:77:0:40, Quit: leaving) | |
| 08:42 | ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz) | |
| 09:18 | wyre is back | |
| 12:29 | wyre is now away: Auto away at Sun Feb 12 12:29:48 2023 UTC | |
| 13:43 | Lantizia has joined IRC (Lantizia!~Lantizia@user/lantizia) | |
| 13:45 | <Lantizia> anyone know what might stop epoptes from running in this fashion?
| |
| 13:45 | https://i.snipboard.io/UAO2Dt.jpg
| |
| 13:46 | basically this is a dell r610 server with trisquel 10 installed on it (so essentially ubuntu 20.04) with the default mate desktop... then the following was done to install ltsp... https://nopaste.net/VUmPX2jpWU
| |
| 13:47 | which does make the machines boot... which is good (although they oddly boot with a 720x480 resolution)... however running epoptes directly from the dell r610 desktop itself (using the ipmi/idrac6... so as though you're physically at the keyb/mouse/monitor) you get that error
| |
| 13:47 | <alkisg> Lantizia: ls -l /etc/epoptes
| |
| 13:47 | <Lantizia> hello alkisg :) i was hoping you'd be about
| |
| 13:48 | i see server.crt and .key
| |
| 13:48 | <alkisg> Size 0?
| |
| 13:48 | <Lantizia> no one is 1939 bytes the other 3272 bytes
| |
| 13:48 | <alkisg> From phone can't type much
| |
| 13:49 | systemctl status epoptes| nc termbin.com 9999
| |
| 13:49 | <Lantizia> termbin.com/55a7
| |
| 13:51 | <alkisg> newgrp epoptes
| |
| 13:51 | And then: epoptes
| |
| 13:51 | <Lantizia> just run them as my normal user (i.e. able to use sudo, but not root) ?
| |
| 13:52 | <alkisg> Right
| |
| 13:52 | <Lantizia> hmm well then newgrp asks for a password
| |
| 13:53 | * Lantizia is thinking FSF Europe shouldn't have promised them Trisquel :P https://twitter.com/fsfe/status/1624068328053415936 | |
| 13:53 | <alkisg> Are you sure you're a member of the epoptes group?
| |
| 13:54 | <Lantizia> the 'admin' user (which is user 1000, the one made when installing the system) is not no
| |
| 13:54 | or it would show with 'groups'
| |
| 13:54 | but I did run... gpasswd -a admin epoptes
| |
| 13:54 | (see the nopaste earlier)
| |
| 13:54 | so that should have done it
| |
| 13:55 | <alkisg> And yet it appears it didn't
| |
| 13:55 | Try logout/login
| |
| 13:55 | <Lantizia> indeed - i've ran it again... logging out and in...
| |
| 13:56 | yay! working epoptes :P
| |
| 13:56 | glad it was something simple
| |
| 13:56 | <alkisg> 👍
| |
| 13:56 | <Lantizia> btw when it says pc's ltsp131 and ltsp232... how is it inventing those numbers?
| |
| 13:57 | ah! ip
| |
| 13:57 | <alkisg> Yeh
| |
| 13:57 | <Lantizia> yeah they're new nic's and I've not static assigned them an ip yet
| |
| 13:57 | thought it looked odd
| |
| 13:58 | now I just need to fix this awful default display resolution
| |
| 13:58 | but I guess that is a trisquel matter really
| |
| 14:01 | thanks alkisg :)
| |
| 14:03 | <alkisg> 😀
| |
| 14:07 | <err404[m]> alkisg: I found the issue \o/
| |
| 14:08 | the issue is: password for remote users should not begin with space character...
| |
| 14:09 | now, all people know I use space char in my password
| |
| 14:10 | <alkisg> Omg 😳
| |
| 14:11 | <err404[m]> I start all my password by space char, to prevent to store it in bash history when accidentaly typed in terminal
| |
| 14:14 | <alkisg> unset HISTFILE
| |
| 14:15 | <err404[m]> I use history every time
| |
| 14:16 | did you have any idea how to allow space char or some special char in begin of a password for remote user? it is a pam issue?
| |
| 14:17 | if it will not fixed soon, I think it will be good to document it 😛
| |
| 14:17 | <alkisg> On phone, can type much, later
| |
| 14:17 | <err404[m]> yes, I know you are on phone, and it is also sunday
| |
| 14:18 | anyway, I am happy to have a full working ltsp
| |
| 14:20 | the start by space char in password is a common issue for many software, I should test it from the beginning 😛
| |
| 14:29 | <Lantizia> alkisg, although I've got this R610 server running trisquel desktop, ltsp, and epoptes... and naturally all the 30 pc's are 'epoptes-clients'... would I be able install epoptes (the server component) on another standalone pc and control clients from that (instead of having to idrac to the R610 all the time to use epoptes)
| |
| 14:53 | yacsa71 has joined IRC (yacsa71!~yacsa71@152.207.241.108) | |
| 14:53 | <yacsa71> Hello, I installed ltsp 20.04 chroorless, what do I do so that the users' home is not deleted?
| |
| 14:54 | yacsa71 has left IRC (yacsa71!~yacsa71@152.207.241.108, Client Quit) | |
| 14:56 | yacsa71 has joined IRC (yacsa71!~yacsa71@152.207.241.108) | |
| 14:57 | yacsa71 has left IRC (yacsa71!~yacsa71@152.207.241.108, Client Quit) | |
| 15:38 | <Lantizia> alkisg, the kernel that is sent to the machines to network boot... is that exactly the same kernel as what the server boots? as it seems it is missing /dev/dri/card0
| |
| 15:38 | so the graphics come up with vesa and 640x480
| |
| 16:33 | <alkisg> yacsa71: by default, it's not deleted. If it is deleted for you, login to a client, and tell us the output of this command: mount | nc termbin.com 9999
| |
| 16:33 | Lantizia: re epoptes, you should be able to run epoptes from any PC as long as it's on the local network
| |
| 16:34 | re kernel, yes it should be the same; what's the output of this command on the client? (uname -a; cat /proc/cmdline; ls -l /boot /lib/modules ) | nc termbin.com 9999
| |
| 18:36 | vagrantc has joined IRC (vagrantc!~vagrant@2600:3c01:e000:21:7:77:0:20) | |
| 20:06 | wyre is back | |
| 20:09 | <Lantizia> alkisg, hey with the graphics thing, i'm being advised to try 'nomodeset' when it comes to the thick clients that network boot.... would this go in /srv/tftp/ltsp/ltsp.ipxe or some other place?
| |
| 20:09 | i guess I'm looking for the regular linux boot line
| |
| 20:15 | <quinox> https://ltsp.org/man/ltsp.conf/#KERNEL_PARAMETERS
| |
| 20:15 | Then run `ltsp ipxe` to update the ipxe file
| |
| 20:16 | you can hack it yourself into ipxe as well, but then it will get lost next time you run `ltsp ipxe`
| |
| 20:17 | <Lantizia> Thanks quinox :)
| |
| 20:29 | quinox, so I did... install -m 0660 -g sudo /usr/share/ltsp/common/ltsp/ltsp.conf /etc/ltsp/ltsp.conf
| |
| 20:29 | and put KERNEL_PARAMETERS="nomodeset" in the [clients] section
| |
| 20:29 | ran... ltsp ipxe
| |
| 20:29 | booted up a machine... and did ... cat /proc/cmdline
| |
| 20:29 | it doesn't show that nomodeset was used
| |
| 20:31 | <alkisg> Lantizia: it should go under [common] to be evaluated on the server, when it runs ltsp ipxe, but anyway, nomodeset is NOT the correct way to go
| |
| 20:31 | nomodeset will disable the kms graphics driver, falling back to vesa, with no acceleration etc
| |
| 20:31 | <Lantizia> I wouldn't want it evaluated on the server
| |
| 20:32 | all the clients are already stuck on vesa
| |
| 20:32 | <alkisg> While ltsp ipxe runs, you want it, to go to ltsp.ipxe
| |
| 20:32 | But anyway it's not the correct way to go
| |
| 20:32 | <Lantizia> https://i.snipboard.io/E4A2Xo.jpg
| |
| 20:32 | <alkisg> Do you have epoptes access to the clients currently?
| |
| 20:32 | <Lantizia> yes
| |
| 20:32 | <alkisg> Do you want to vnc to me?
| |
| 20:32 | !vnc-edide
| |
| 20:32 | <ltspbot> vnc-edide: To share your screen with me, open Epoptes → Help menu → Remote support → Host: srv1-dide.ioa.sch.gr, and click the Connect button
| |
| 20:32 | <alkisg> ...do that from the server
| |
| 20:33 | <Lantizia> done
| |
| 20:34 | <alkisg> Lantizia: I think you're just missing the firmware bits
| |
| 20:34 | <Lantizia> yeah it is intentional lol - it is trisquel
| |
| 20:34 | it wouldn't be much of an FSFE meeting with them included lol
| |
| 20:34 | but from what I've read the 'radeon' driver should work without them
| |
| 20:35 | just lack hardware acceleration
| |
| 20:35 | <alkisg> Ah wait, so you don't want to fix it properly? You want to keep it without firmware?
| |
| 20:35 | <Lantizia> termbin.com/ppuw3
| |
| 20:35 | that is how xorg boots on the clients
| |
| 20:37 | <alkisg> Lantizia: can I add the greek schools ppa which has fixed socat etc versions?
| |
| 20:37 | On the server
| |
| 20:37 | Because now opening the terminal via epoptes doesn't work correctly
| |
| 20:37 | <Lantizia> fixed what?
| |
| 20:37 | <alkisg> This socat version is buggy
| |
| 20:38 | So the terminal that I open from the client doesn't work properly
| |
| 20:38 | I can't use nano etc
| |
| 20:38 | <Lantizia> ok but I'm trying to document anything changed
| |
| 20:38 | so as long as a) I know what has changed (and how) and b) that it is all free (no non-free components)
| |
| 20:39 | alkisg, someone has just mentioned in the #trisquel channel... https://libreplanet.org/wiki/Group:Hardware/research/gpu/radeon
| |
| 20:40 | alkisg, so it seems like it falling back on vesa is a known thing :S
| |
| 20:41 | alkisg, it seems like parabola might work better as they might have incorporated the fixes... and they're noted here against this exact gpu... https://h-node.org/videocards/view/en/363/760G--Radeon-3000--
| |
| 20:42 | but I picked trisquel as it was closer to the ltsp recommended distro of ubuntu mate
| |
| 20:46 | <alkisg> Anything based on initramfs-tools should do
| |
| 20:46 | <Lantizia> ?
| |
| 20:53 | <alkisg> Lantizia: as expected, nomodeset didn't help
| |
| 20:53 | <Lantizia> and you did a xorg.conf I noticed too
| |
| 20:53 | <alkisg> I was playing with the driver stanza, and the horizontal/vertical timings, to no avail
| |
| 20:54 | Not sure how to use that without firmware... :)
| |
| 20:54 | You can also try trisquel locally, and when you get it working, we can then use that workaround in ltsp too
| |
| 20:55 | <Lantizia> at the moment I'm not even in that location (physically)
| |
| 20:55 | so I can't usb boot anything until tomorrow morning if that's what you're asking
| |
| 20:55 | plus these machines have no hard disks
| |
| 20:56 | <alkisg> Yeh I was suggesting that you'd install it in a usb disk/stick, and play with it until you get it working
| |
| 20:56 | Once you have the solution, it'll be very easy to apply it to ltsp
| |
| 20:56 | But you can also test with ltsp if you prefer...
| |
| 20:57 | ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Quit: Leaving) | |
| 20:57 | <Lantizia> I think right now I'm prepared to tell anyone that comes to this meeting, that if they want 'free software' .. it sometimes means 640x480 :P
| |
| 20:57 | <alkisg> Deblobbed software needs specialized hardware, it's not easy to go full free with generic hardware
| |
| 20:57 | Haha, rihgt
| |
| 20:58 | * Haha, right
| |
| 20:58 | <Lantizia> it's a shame as it's only this one thing... i'm amazed that the audio and network on these machines work
| |
| 20:58 | they're (all 30 of them) compaq 315eu machines
| |
| 20:59 | i'd be happy if vesa was just forced to 1366x768
| |
| 20:59 | or when 1024x768
| |
| 20:59 | *or even
| |
| 21:02 | <alkisg> Add the firmware and try to hide it from the meeting :P
| |
| 21:02 | <Lantizia> don't... tempt... me
| |
| 21:02 | normally we run lug meetings in this room (staffslug.org.uk) and we're still on ubuntu 18.04 and ltsp5
| |
| 21:03 | runs fine though no issues
| |
| 21:03 | so I know it's bloody trisquel
| |
| 21:03 | <alkisg> But ubuntu 18.04 includes the firmware, right?
| |
| 21:03 | <Lantizia> now the room is also used by FSFE Potteries (i.e. https://wiki.fsfe.org/LocalGroups/Potteries )
| |
| 21:03 | alkisg, well yeah - never had graphics issues
| |
| 21:04 | technically that r610 you're controlling now... first partition is that ubuntu 18.04.. second is trisquel
| |
| 21:04 | i'm just dual booting the server lol
| |
| 21:04 | <alkisg> [ 16.798] (II) [KMS] drm report modesetting isn't supported.
| |
| 21:04 | ...due to lack of firmware, I'd believe
| |
| 21:04 | So radeon is out
| |
| 21:05 | <Lantizia> did you see this? https://libreplanet.org/wiki/Group:Hardware/research/gpu/radeon
| |
| 21:05 | i linked to it earlier
| |
| 21:05 | <alkisg> Ah no I only saw the second link which wasn't useful
| |
| 21:05 | Reading...
| |
| 21:06 | <Lantizia> it's possible trisquel with a different kernel (with these added tweaks) might be better at this
| |
| 21:06 | apparently parabola has these already included
| |
| 21:07 | <alkisg> But it says "how to patch in parabola"
| |
| 21:07 | So to me it looks like you'll need to manually apply the changes in both cases
| |
| 21:07 | <Lantizia> yeah but possibly just using the 'linux-libre' kernel might help? i'm not sure if trisquel doesn/doesn't already use that
| |
| 21:08 | <alkisg> Not sure... anyway time for me to go, good luck with this!
| |
| 21:08 | <Lantizia> oh it does lol
| |
| 21:08 | yeah you've gone above and beyond - thanks
| |
| 21:11 | i'll possibly try the beta of trisquel 11 (i'm guessing a later linux-libre kernel) which is based off 22.04
| |
| 21:11 | but for now bed time :)
| |
| 21:24 | vagrantc has left IRC (vagrantc!~vagrant@2600:3c01:e000:21:7:77:0:20, Quit: leaving) | |