IRC chat logs for #ltsp on irc.freenode.net (webchat)


Channel log from 15 January 2019   (all times are UTC)

00:00
<nikoh77>
ok, many thanks guys
00:00
see you soon ;-)
00:01
just one last thing
00:01
LIMIT_ONE_SESSION directive work the same way on fat/thin?
00:06
<mwalters>
I'm not sure on that one
00:10nikoh77 has left IRC (nikoh77!~nikoh77@host88-180-dynamic.53-79-r.retail.telecomitalia.it, Quit: Leaving)
00:50vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving)
02:24vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
05:20kjackal has joined IRC (kjackal!~quassel@154.119.55.246)
05:38kjackal has left IRC (kjackal!~quassel@154.119.55.246, Ping timeout: 258 seconds)
05:54spaced0ut has left IRC (spaced0ut!~spaced0ut@unaffiliated/spaced0ut, Ping timeout: 244 seconds)
06:01vsuojanen has left IRC (vsuojanen!~vsuojanen@cable-hml-585686-205.dhcp.inet.fi, Ping timeout: 268 seconds)
06:03vsuojanen has joined IRC (vsuojanen!~vsuojanen@cable-hml-585686-205.dhcp.inet.fi)
06:20kjackal has joined IRC (kjackal!~quassel@154.119.55.246)
06:32kjackal has left IRC (kjackal!~quassel@154.119.55.246, Ping timeout: 258 seconds)
07:25ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz)
07:36vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving)
08:35kjackal has joined IRC (kjackal!quassel@conference/canonical/x-jthvxojahwjictrc)
08:48alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
08:55alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
09:32kjackal has left IRC (kjackal!quassel@conference/canonical/x-jthvxojahwjictrc, Ping timeout: 246 seconds)
09:58kjackal has joined IRC (kjackal!quassel@conference/canonical/x-sbuotvzmrbpqbvfy)
10:04kjackal has left IRC (kjackal!quassel@conference/canonical/x-sbuotvzmrbpqbvfy, Ping timeout: 272 seconds)
10:23kjackal has joined IRC (kjackal!quassel@conference/canonical/x-mkdkuihhyqwdexdg)
10:53kjackal has left IRC (kjackal!quassel@conference/canonical/x-mkdkuihhyqwdexdg, Ping timeout: 268 seconds)
11:53kjackal has joined IRC (kjackal!quassel@conference/canonical/x-gawcmrebxonbmpfo)
12:02kjackal has left IRC (kjackal!quassel@conference/canonical/x-gawcmrebxonbmpfo, Ping timeout: 272 seconds)
12:03kjackal has joined IRC (kjackal!quassel@conference/canonical/x-duxrcunecwbnjrsu)
13:02kjackal has left IRC (kjackal!quassel@conference/canonical/x-duxrcunecwbnjrsu, Ping timeout: 272 seconds)
13:42kjackal has joined IRC (kjackal!quassel@conference/canonical/x-ddwadyaehrgmnnst)
13:57nikoh77 has joined IRC (nikoh77!~nikoh77@93-55-122-94.ip263.fastwebnet.it)
13:59
<nikoh77>
Hi guys, to swap from fat to thin client configuration for all clients i have added LTSP_FATCLIENT=FALSE is it right?
14:00
in default sectio
14:02kjackal has left IRC (kjackal!quassel@conference/canonical/x-ddwadyaehrgmnnst, Ping timeout: 258 seconds)
14:04
<alkisg>
Right
14:16
<nikoh77>
Ok, and now client dont start.... at the end of the boot "freez" on "Starting Update UTMP about System Runlevel Changes..."
14:25
i have a complete journalctl log but i dont see errors....i think
14:30spaced0ut has joined IRC (spaced0ut!~spaced0ut@unaffiliated/spaced0ut)
14:32
<nikoh77>
I only added LTSP_FATCLIENT=FALSE in ltsp.conf some idea?
14:32
sorry lts.conf
14:46kjackal has joined IRC (kjackal!quassel@conference/canonical/x-nfzqfocqcxpngciu)
15:06kjackal has left IRC (kjackal!quassel@conference/canonical/x-nfzqfocqcxpngciu, Ping timeout: 250 seconds)
15:25kjackal has joined IRC (kjackal!quassel@conference/canonical/x-cdoxxlvmjzbsitax)
15:29nikoh77 has left IRC (nikoh77!~nikoh77@93-55-122-94.ip263.fastwebnet.it, Remote host closed the connection)
15:44kjackal has left IRC (kjackal!quassel@conference/canonical/x-cdoxxlvmjzbsitax, Ping timeout: 258 seconds)
15:47nikoh77 has joined IRC (nikoh77!~nikoh77@93-55-122-94.ip263.fastwebnet.it)
15:48
<alkisg>
nikoh77: upload your current lts.conf so that we see
16:54
<nikoh77>
alkisg: how can I upload it here
16:54
?
16:56
<alkisg>
!paste
16:56
<ltsp>
paste: To avoid channel flooding, please upload text longer than 3 lines to http://paste.debian.net. Don't forget to paste the resulting URL here.
17:07
<mwalters>
!logging
17:07
<ltsp>
Error: "logging" is not a valid command.
17:07
<mwalters>
!log
17:07
<ltsp>
I do not know about 'log', but I do know about these similar topics: 'logs', 'irclogs', 'changelog', 'rsyslog'
17:07
<mwalters>
!logs
17:07
<ltsp>
logs: http://irclogs.ltsp.org/
17:07
<mwalters>
!rsyslog
17:07
<ltsp>
rsyslog: To enable remote client logging in Ubuntu (it possibly also works on Debian), see https://bugs.launchpad.net/ubuntu/+source/ltsp/+bug/697387/comments/9
17:09
<mwalters>
looks like this is outdated for 18.04... uncomment l17 and l18 now I think
17:12
<nikoh77>
http://paste.debian.net/1060637/
17:14
<alkisg>
nikoh77: try commenting out #LDM_THEME
17:14
Aaah
17:14
And, and, try UNcommenting #X_SMART_COLOR_DEPTH=False
17:15
<nikoh77>
if i comment LTSP_FATCLIENT=False all work fine
17:16
<alkisg>
(07:14:58 μμ) alkisg: And, try UNcommenting #X_SMART_COLOR_DEPTH=False
17:16
<nikoh77>
alkisg: in chrootlees environment i dont need to build clients right?
17:17
<alkisg>
You don't need to run ltsp-build-client, if that's what you're asking
17:17
<nikoh77>
yes, ok
17:23
..a question.. whe i modify lts.conf, i need to update image to apply?
17:30
<alkisg>
No
17:36bengoa has left IRC (bengoa!~alberto@194.50.55.200, Ping timeout: 240 seconds)
17:37
<nikoh77>
mwalters: thi is the log https://paste.debian.net/1060666/
17:38
the boot sto one second before start login page (xsession?), about when NDB_DISCONNECT
17:39
and... if i comment LTSP_FATCLIENT=False all work fine...
17:41
<mwalters>
alkisg: sorry, I was commenting on something unrelated
17:41
<alkisg>
mwalters: sure, I know, it's nikoh77 that didn't understand that you weren't talking to him
17:42
nikoh77: please read this line:
17:42
(07:14:58 μμ) alkisg: And, try UNcommenting #X_SMART_COLOR_DEPTH=False
17:42
<mwalters>
ops, meant to tag nikoh77 not you
17:42
<alkisg>
That means, remove the # from that line in your lts.conf, and reboot the client
17:45
<nikoh77>
ok i try
17:56
Hey Guys, work perfectly :-))))))
17:56
Thankkkssss
18:17
<alkisg>
You're welcome
18:21
<nikoh77>
just last question: is possible to limit one session so that second thin client with same user dont gain access?
18:23
<mwalters>
that limit one session line should work for thin clients at least. It will ask them if they would like to close/kill other sessions when they log onto another computer
18:23
I'm not sure if/how it works for fat clients
18:24
<nikoh77>
ok many thanks ;-)
18:29nikoh77 has left IRC (nikoh77!~nikoh77@93-55-122-94.ip263.fastwebnet.it, )
18:29
<mwalters>
:)
19:07bengoa has joined IRC (bengoa!~alberto@194.50.55.200)
19:25vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
20:04nikoh77 has joined IRC (nikoh77!~nikoh77@host88-180-dynamic.53-79-r.retail.telecomitalia.it)
20:15kjackal has joined IRC (kjackal!~quassel@154.119.55.246)
20:39
<mwalters>
alkisg: any idea on this? I have one EFI client that boots to tty1 instead of ldm. If I hit ctrl+alt+f7, it shows ldm just fine.
20:39
<alkisg>
Try SCREEN_DEFAULT=07 in lts.conf
20:40
<mwalters>
thanks, I'll give that a shot
20:40
<alkisg>
The screen code design doesn't match the systemd era, it needs rethinking
20:40
<mwalters>
ah I see
20:40
shouldn't matter if I put that under [default], I'm thinking?
20:40
<alkisg>
Right
20:41
<mwalters>
(I don't have any clients that *shouldn't* default to ldm)
20:41
<alkisg>
I'm not sure that this will work; just a try...
20:43
<mwalters>
defaulted to ldm that boot, thanks!
20:46
<alkisg>
Nice
20:54
<mwalters>
It's strange, I was pulling my hair out w/ this intel NUC7... it absolutely refuses to boot using the EFI image. iPXE grabs it fine... I have other efi clients grabbing it fine... (not asking for help, just commenting/venting)
20:54
There's no option for legacy boot in the bios =/
20:55
I can see the image being offered, the silly thing just refuses to grab it
20:55
I'm using it as my personal desktop right now with iPXE on a usb stick... otherwise it's a solid little computer for a fat client
20:59
<quinox>
I still have to figure out how to make this iMac do network boot
20:59kjackal has left IRC (kjackal!~quassel@154.119.55.246, Ping timeout: 268 seconds)
20:59
<vagrantc>
SCREEN_DEFAULT should be set to the highest defined SCREEN_NN
21:00
really odd that that would break; it's pretty straightforward code
21:04
<mwalters>
it was on this nuc, so who knows ;)
21:20
<alkisg>
vagrantc: I'm starting to design the new tftp layout/code for ltsp6... I'm thinking we could default to ipxe, and only document syslinux/grub as manual options for those rare cases where ipxe doesn't fit
21:20
I got both uefi and non uefi working
21:20
ipxe can't load a 32bit kernel from an uefi booted 64bit client
21:20nikoh77 has left IRC (nikoh77!~nikoh77@host88-180-dynamic.53-79-r.retail.telecomitalia.it, Ping timeout: 268 seconds)
21:20
<alkisg>
But I think we can ignore that combination...
21:21
A question about architectures... now we say "i386/amd64", I'm thinking of switching to `uname -m`
21:22
This is i686 for 32bit, x86_64 for 64bit, and armv7l for raspberrys, in all distros that I tested with
21:22
ipxe will allow us to only put a couple of static files in tftp, and boot the kernel/initrd via http
21:25
Scenarios I've tested: (1) BIOS > TFTP undionly.kpxe > boot.ipxe script, (2) BIOS > grub > ipxe.lkrn > boot.ipxe script, (3) UEFI > TFTP snponly.efi > boot.ipxe script, (4) UEFI > grub > ipxe.efi > boot.ipxe script
21:26
So the server only has undionly.kpxe, snponly.efi, and our static boot.ipxe script, which supports "if" and variables, so it can be a static example that the user will manually modify if the defaults don't work for him
21:27
No more smart logic with 3 layers, too difficult for users to figure it out :D
21:33
<vagrantc>
using 'uname -m' requires a running system and may return unexpected output in a number of circumstances
21:37
<alkisg>
But https://wiki.debian.org/SupportedArchitectures is debian specific, while uname -m is kernel specific, right?
21:37
So uname -m should be "cross distro"...
21:38
The client requesting the chroot will have running code (initramfs), yeah
21:41
<vagrantc>
uname -m may return i586 or i686 or a number of other things ... it's generally distro-specific, actually
21:42
armv7l is the kernel, not the userspace
21:42
and depending on what kernel you're running, you may get a different result
21:42
<alkisg>
And that would be a good thing because we would want to differentiate between those, right?
21:42
<vagrantc>
really, you want to query the distro's idea of architecture, otherwise you'll get something wrong
21:43
<alkisg>
Another example... dnsmasq/ipxe can see if a client is bios/uefi 32bit/uefi 64bit
21:44
We can then direct it to load a kernel based on that information only,
21:44
now, if we keep using distro-defined names like now, the code is a mess,
21:44
<vagrantc>
but what userspace do you give it, if there are more than one userspace supported by that kernel?
21:45
<alkisg>
while, if we require specific symlinks, it makes things easier, e.g. "vmlinuz.i686" or "vmlinuz.x86_64" etc
21:45
The chroot is configured in dnsmasq
21:45
The same chroot can be booted with different kernels
21:45
<vagrantc>
not always
21:46
<alkisg>
Sure. So, if the symlink exists, then it can be used, if not, it's not available
21:46
But to make things easier in the tftp stage, we'd want all the distros to use the same symlink names
21:47
Imagine writing the "ltsp-update-kernels" code as an ipxe script, chaos
21:47adrianor1 has joined IRC (adrianor1!~adrianorg@189.114.158.19)
21:49
<alkisg>
I.e. in the ltsp-client package, instead of the kernel postinst generating /boot/pxelinux.cfg, it would just create the appropriate kernels/initrd symlinks, with cross-distro names
21:49
Anyways it got a bit late... g'night for now, let's continue some other time
21:51adrianorg has left IRC (adrianorg!~adrianorg@177.204.151.149.dynamic.adsl.gvt.net.br, Ping timeout: 268 seconds)
22:16nikoh77 has joined IRC (nikoh77!~nikoh77@host88-180-dynamic.53-79-r.retail.telecomitalia.it)
22:48spaced0ut has left IRC (spaced0ut!~spaced0ut@unaffiliated/spaced0ut, Quit: Leaving)
22:49
<nikoh77>
hi, is possible to use raspberry pi 3 as a thin client?
23:05ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Remote host closed the connection)
23:17
<vagrantc>
nikoh77: yes, but don't have high expectations
23:17
you might be better off using them as fat clients, and even then, they're quite slow
23:17
it's a bit more complicated to do right
23:18
but technically possible
23:30
<nikoh77>
ok thanks
23:30nikoh77 has left IRC (nikoh77!~nikoh77@host88-180-dynamic.53-79-r.retail.telecomitalia.it, Quit: Leaving)