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


Channel log from 25 March 2016   (all times are UTC)

00:13vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Ping timeout: 244 seconds)
01:16neo_ has joined IRC (neo_!d91b21d5@gateway/web/freenode/ip.217.27.33.213)
01:17
<neo_>
hello
01:19neo_ has left IRC (neo_!d91b21d5@gateway/web/freenode/ip.217.27.33.213, Client Quit)
01:22dtcrshr has left IRC (dtcrshr!~datacrush@unaffiliated/datacrusher, Quit: Saindo)
06:02F-GT has left IRC (F-GT!~phantom@ppp121-44-24-185.lns20.syd4.internode.on.net, Ping timeout: 260 seconds)
06:15F-GT has joined IRC (F-GT!~phantom@ppp121-44-91-206.lns20.syd4.internode.on.net)
07:19vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
07:29fnurl has left IRC (fnurl!3cf8605f@gateway/web/cgi-irc/kiwiirc.com/ip.60.248.96.95, Ping timeout: 264 seconds)
07:50vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving)
07:59robb_nl has joined IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be)
08:02ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz)
08:06gvy has joined IRC (gvy!~mike@altlinux/developer/mike)
08:53kjackal has joined IRC (kjackal!~kjackal@2a02:587:3101:3500:3c88:f068:592:8232)
09:04kjackal has left IRC (kjackal!~kjackal@2a02:587:3101:3500:3c88:f068:592:8232, Ping timeout: 246 seconds)
09:11alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
09:17alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
09:23Anderson69s has joined IRC (Anderson69s!~Anderson6@81.18.188.59)
09:32kjackal has joined IRC (kjackal!~kjackal@2a02:587:3101:3500:3c88:f068:592:8232)
09:38kjackal has left IRC (kjackal!~kjackal@2a02:587:3101:3500:3c88:f068:592:8232, Ping timeout: 246 seconds)
09:38sutula_ has joined IRC (sutula_!~sutula@207-118-162-223.dyn.centurytel.net)
09:39man has joined IRC (man!~username@unaffiliated/man)
09:39gehidore has left IRC (gehidore!~username@unaffiliated/man, Ping timeout: 240 seconds)
09:39sutula has left IRC (sutula!~sutula@207-118-162-223.dyn.centurytel.net, Ping timeout: 240 seconds)
09:39sutula_ is now known as sutula
09:42izzle121 has left IRC (izzle121!~izzle121@2601:193:c200:1d00:52e5:49ff:fec0:534b, Ping timeout: 240 seconds)
09:42izzle121 has joined IRC (izzle121!~izzle121@2601:193:c200:1d00:52e5:49ff:fec0:534b)
09:47book` has left IRC (book`!~book`@2607:5300:100:200::160d, Ping timeout: 248 seconds)
09:53book` has joined IRC (book`!~book`@167.114.152.105)
10:00Anderson69s has left IRC (Anderson69s!~Anderson6@81.18.188.59, Quit: My Mac has gone to sleep. ZZZzzz…)
10:04book` has joined IRC (book`!~book`@105.ip-167-114-152.net)
10:07robb_nl has left IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be, Ping timeout: 264 seconds)
10:25robb_nl has joined IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be)
10:26adrianorg has left IRC (adrianorg!~adrianorg@189.58.183.198.dynamic.adsl.gvt.net.br, Ping timeout: 260 seconds)
10:28adrianorg has joined IRC (adrianorg!~adrianorg@187.113.250.186)
10:41Anderson69s has joined IRC (Anderson69s!~Anderson6@81.18.188.59)
10:56kjackal has joined IRC (kjackal!~kjackal@2a02:587:3101:3500:3c88:f068:592:8232)
11:05kjackal has left IRC (kjackal!~kjackal@2a02:587:3101:3500:3c88:f068:592:8232, Ping timeout: 246 seconds)
11:10Anderson69s has left IRC (Anderson69s!~Anderson6@81.18.188.59, Quit: My Mac has gone to sleep. ZZZzzz…)
11:18
<alkisg>
OK new highscore: [ 5.546362] block nbd9: queue cleared
11:19
...boot to LDM in 5.5 seconds :D
11:22
<sebd>
alkisg: hi, can you share your specs ? when do you count from ?
11:23
5.5 seconds from power on ?
11:23
<alkisg>
Hi sebd, this is a virtualbox thin vm under my core i5 system, booting in bridged gigabit networking, counting the dmesg time at ldm
11:23
So it's after the kernel and initrd are loaded via tftp
11:23
<sebd>
ok
11:23
still good
11:24
<alkisg>
A really ancient pentium 3 that used to boot in 1 minute when I was testing ltsp in 2008, now boots in 30 secs
11:25
<sebd>
let's check my average PCs' dmesg
11:26
<alkisg>
The nbd9 message is when ldm checks for a newer chroot, so it's the last dmesg entry for me when a client boots up
11:26
I'm using that one because it's the easiest way to see it without counting externally
11:27robb_nl has left IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be, Ping timeout: 260 seconds)
11:36
<sebd>
ok so in my real world, switched-on clients score between 10 and 38 seconds. Most of them are not gigabit, that is why.
11:38
a lot of them start in around 15 seconds.
11:39
<alkisg>
I've been doing a lot of optimizations in the ltsp code
11:39
So if you want to test if it improved your boot time, you'd have to use the greek schools proposed ppa...
11:40
Which distro/version?
11:40
<gvy>
alkisg, my ancient PIII-500 hosts back in 2006 booted way faster using ALTSP (and didn't end up booting Edubuntu's LTSP5 due to 64M RAM)... nice timing tho :)
11:41
<sebd>
alkisg: debian 8 jessie
11:42
<alkisg>
sebd, can you upload one of the fast dmesg of your clients?
11:42
in pastebin or so...
11:42
or mail if you don't want it public
11:44
<sebd>
alkisg: ATM the fastest is 9.something, I am looking for a better one (I administer 4 real LTSP networks)
11:44
<alkisg>
No need; I just want to see if it has any big delays there
11:44
Using debian, you were spared the 10 second delay in Ubuntu that got me started with all this
11:44
<sebd>
ok
11:44
<alkisg>
But there were some others that I optimized, and I wanted to see if you have them there
11:45
Any dmesg under 15-20 seconds is fine
12:11
sebd: thanks! You were not affected by any of the 3 major timeouts, so you'll only see a small benefit from the recent changes
12:12
You *would* be affected by 2 of them if you upgraded to stretch, but they've been solved now
12:12
So you're lucky, you will never even see them :)
12:12Anderson69s has joined IRC (Anderson69s!~Anderson6@81.18.188.59)
12:16
<sebd>
:)
12:36
<alkisg>
gvy: ubuntu 12.04 with ltsp-trunk boots in 27 seconds in a pentium 3 with 64mb ram here
12:36
<gvy>
:)
12:37
<alkisg>
16.04 would boot in around 30, if not for some drm issues with its ancient graphics card
12:38
(but not with 64 ram anymore)
12:42Anderson69s has left IRC (Anderson69s!~Anderson6@81.18.188.59, Quit: My Mac has gone to sleep. ZZZzzz…)
12:42robb_nl has joined IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be)
13:03Anderson69s has joined IRC (Anderson69s!~Anderson6@81.18.188.59)
13:18fnurl has joined IRC (fnurl!3cf8605f@gateway/web/cgi-irc/kiwiirc.com/ip.60.248.96.95)
13:23robb_nl has left IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be, Ping timeout: 268 seconds)
13:23Anderson69s has left IRC (Anderson69s!~Anderson6@81.18.188.59, Quit: My Mac has gone to sleep. ZZZzzz…)
13:43klausade has left IRC (klausade!~klaus@cm-84.215.169.187.getinternet.no, Ping timeout: 264 seconds)
13:54kjackal has joined IRC (kjackal!~kjackal@2a02:587:3101:3500:3c88:f068:592:8232)
13:56ben_roose has joined IRC (ben_roose!~roose@roose.cs.wichita.edu)
13:57Anderson69s has joined IRC (Anderson69s!~Anderson6@81.18.188.59)
14:08Anderson69s has left IRC (Anderson69s!~Anderson6@81.18.188.59, Quit: My Mac has gone to sleep. ZZZzzz…)
14:37robb_nl has joined IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be)
14:38m3741 has joined IRC (m3741!8c20b7fe@gateway/web/freenode/ip.140.32.183.254)
15:02fenixfunk5 has left IRC (fenixfunk5!~fenix@mail.lbathivel.com, Quit: Quitte)
15:04Anderson69s has joined IRC (Anderson69s!~Anderson6@81.18.188.59)
15:18
<alkisg>
!client-list
15:18
<ltsp>
client-list: to get a list of all nbd-clients (which sometimes is the same as ltsp clients), run: netstat -tn | sed -n 's/.*:10809 *\([0-9.]*\):.*/\1/p' | sort -Vu
15:21Anderson69s has left IRC (Anderson69s!~Anderson6@81.18.188.59, Quit: My Mac has gone to sleep. ZZZzzz…)
15:55robb_nl has left IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be, Ping timeout: 264 seconds)
16:09robb_nl has joined IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be)
16:17robb_nl has left IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be, Ping timeout: 250 seconds)
16:30kjackal has left IRC (kjackal!~kjackal@2a02:587:3101:3500:3c88:f068:592:8232, Ping timeout: 246 seconds)
16:33gvy has left IRC (gvy!~mike@altlinux/developer/mike, Quit: Leaving)
17:28vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
17:51man is now known as gehidore
18:08izzle121 has left IRC (izzle121!~izzle121@2601:193:c200:1d00:52e5:49ff:fec0:534b, Remote host closed the connection)
18:27eu^athedsl-20573 has joined IRC (eu^athedsl-20573!554a7b43@gateway/web/freenode/ip.85.74.123.67)
18:46
<vagrantc>
alkisg: ltsp seems fine for the most part, but the new epoptes failed to start... haven't dug in much yet
18:47
<alkisg>
vagrantc: the server side?
18:47
Like, in service epoptes (re)start?
18:48
Or the GUI? Or the client?
18:48
<vagrantc>
alkisg: server-side
18:49
<alkisg>
sudo service epoptes restart, put syslog+stdout to pastebin?
18:50
<vagrantc>
+stdout ?
18:50
<alkisg>
stderr
18:52
<vagrantc>
from service? there's absoletely no output there.
18:53
should i have to do anything beyond installing epoptes?
18:54
<alkisg>
sudo apt-get install epoptes
18:54
add yourself to the epoptes group
18:54
logout/login _or_ use newgrp epoptes
18:54
and then you should be able to run the UI
18:54
Do you want me to have a quick look via vnc?
18:54
!vnc-dide
18:54
<ltsp>
vnc-dide: To share your screen with me, run this: sudo apt-get --yes install x11vnc; x11vnc -connect srv1-dide.ioa.sch.gr - this is a reverse connection, it doesn't need port forwarding etc.
18:54
<vagrantc>
http://paste.debian.net/418764/
18:54
there should be an epoptes process running, yes?
18:57
<alkisg>
http://paste.debian.net/418765/
18:57
<vagrantc>
no such luck here
18:58
running /etc/init.d/epoptes from an xterminal breaks my session, for good measure
18:59
<alkisg>
Wow
18:59
<vagrantc>
hrm. not anymore
18:59
<alkisg>
http://bazaar.launchpad.net/~epoptes/epoptes/trunk/changes?filter_file_id=epoptes.init-20111017094800-yrt7sme6ru9g1gzu-1
18:59
<vagrantc>
so, it seems to fail due to /etc/lts.conf not being present?
18:59
<alkisg>
Last change in 2012
19:00
...I don't see anything in the code reading lts.conf in the daemon...
19:00
<vagrantc>
hrm.
19:01
<alkisg>
# strace -e trace=file service epoptes start 2>&1 | grep lts
19:01
stat64("/opt/ltsp/images", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
19:01
<vagrantc>
you know, this was a system i experimented with libpam-sshauth+lightdm+lightdm-webkit-greeter ...
19:01
<alkisg>
I didn't read lts.conf for me
19:01
*it
19:02
<vagrantc>
maybe this is just a borked system
19:02
<alkisg>
could be, I don't think the epoptes code reads lts.conf
19:02
<vagrantc>
maybe the initramfs-tools hooks without the prereqs ran and mangled some things
19:28
<alkisg>
egrep -qs 'ltsp|nfs|nbd' /proc/cmdline ==> this isn't matching anything, is it?
19:28alkisg is now known as alkisg_away
19:28kjackal has joined IRC (kjackal!~kjackal@2a02:587:3101:3500:3c88:f068:592:8232)
19:29
<vagrantc>
oh yes.
19:29
root on lvm, with ltsp in the volume name
19:30
BOOT_IMAGE=/vmlinuz-4.4.0-1-amd64 root=/dev/mapper/sltsp-root ro quiet
19:31
that's ... not great.
19:31alkisg_away is now known as alkisg
19:31
<alkisg>
vagrantc: sorry, the check for ltsp clients isn't very good then
19:32
The logic is "if it's an ltsp client, allow the EPOPTES lts.conf directive to specify if we want to run epoptes on the client or not"
19:33
vagrantc: can you change that so that it greps for init=/sbin/init-ltsp?
19:33
<vagrantc>
yeah, that egrep never really seemed to be a problem to me ... but i guess my last build i did with lvm ...
19:33
nfs and nbd also seem a bit overly zealous, thinking about it now
19:33
<alkisg>
grep -qs init=/sbin/init-ltsp /proc/cmdline
19:34
We don't care if it's nbd, nfs or aoe
19:34
or even local loopback etc
19:34
Want me to push it?
19:35
if grep -qs init=/sbin/init-ltsp /proc/cmdline && [ -x /usr/bin/getltscfg ]; then
19:35
<vagrantc>
alkisg: that seems much safer, sure.
19:35
alkisg: though i think not all distros have exactly that in the boot commadline ...
19:35
<alkisg>
See, your weird setups do have their uses :D
19:35
Uncovering bugs
19:36
<vagrantc>
seem to recall gentoo having a slightly different commandline
19:36
<alkisg>
We're already using it in epoptes, in ./debian/epoptes-client.init
19:36
So if Gentoo ever gets epoptes, we'll patch both of them then
19:36* vagrantc never expected lvm to be classified as "weird setup"
19:37
<vagrantc>
yup
19:37
<alkisg>
./server/Redhat/configs/pxe-default.conf: APPEND rw initrd=initrd.ltsp init=/sbin/init-ltsp selinux=0 root=dhcp rhgb video=LVDS-1:d rdshell
19:38
Redhat is ok, can't find gentoo
19:38
later!
19:38alkisg is now known as alkisg_away
19:39* vagrantc will push if alkisg_away didn't already
19:39
<vagrantc>
appears to be code for that all over the place
19:40
well, three files, four lines...
19:53
<lee>
MilkDrop & Geiss Kreygasm
19:53
erp
19:53
well, that was the wrong window :D
19:54
<vagrantc>
just for that, you've been sentenced to fix at least one LTSP bug :)
19:55
alkisg_away: seems to be working better now
19:55
<lee>
I'm far more likely to be responsible for creating (or maybe finding) than fixing
19:55
we're on a, ubuntu 12.04 network at the moment, I'm waiting for the 16.04 release before playing with a new LTSP version
19:56
<vagrantc>
you might want to find those bugs *before* it releases :)
19:56
given that it's only a month or so away
19:56
much easier to fix at this point than later
20:03
<lee>
I just don't use it enough tbh
20:03
I set up some fat clients and that's pretty much it
20:03
or at least, that's it for now
20:04
<vagrantc>
alkisg_away: working much better with that patch ... should the similar client-side code be patched as well?
20:10
looks like it was already updated in at least one place
20:34Wellholt has joined IRC (Wellholt!51e135b4@gateway/web/freenode/ip.81.225.53.180)
20:40
<Wellholt>
Hope to find answers to how to set up a 9 Minecraft clients through ltsp server.
20:41
<maldridge>
Wellholt: greetings, what you're trying to do shouldn't be any real issues
20:41
you'll want to run your clients as "fat" clients if you can so that you get better performance since Minecraft is highly graphical
20:42
<Wellholt>
Ltsp looks simple but how about the 3D acceleration
20:43
<maldridge>
that is entirely dependant on what your hardware can do. I should point out that unless you have some extra software installed, minecraft expects the renderer to be capable of running on CPU only without GPU assistance
20:43
Wellholt: do you have hardware to test with?
20:43
<Wellholt>
Ok I have to stop thinking about thin clients then.
20:44
Yes I will set up my test rig next week
20:44
<vagrantc>
Wellholt: you can run LTSP fat clients if the machines are capable of running whatever application you want
20:44
Wellholt: but graphics over the network are going to be a limitation if you want to do genuinely thin clients.
20:46
<Wellholt>
Ok and Eave iI use 10Gb nic IT want change
20:49
<vagrantc>
3D graphics over the network is generally very poorly supported, no matter how much bandwidth you have. not that it's impossible, but it would probably involve significant software development.
20:50
hence, LTSP fat clients. you still get the advantage of centralized administration, but you can take advantage of local hardware capabilities, such as 3D graphics.
20:51
<Wellholt>
I like to set up an environment at my kids school for educatinal use and IF They could play Minecraft on the same thin client IT would be a pluss
20:52
<vagrantc>
in general, i wouldn't recommend thin clients except in very specific circumstances. most LTSP deployments i know of are fat clients these days.
20:52
<Wellholt>
Ok I will use a mixed client environment
20:52
<vagrantc>
it's simpler, more efficient, and performs better
20:53
<Wellholt>
Do u know Userful
20:55
<vagrantc>
not off the top of my head
20:55
https://en.wikipedia.org/wiki/Userful ?
20:55
<Wellholt>
I have seen an Implementation of Userful using one graphic Cars for each terminal from a singel server
20:56
Cool thank a lot
20:56
<maldridge>
Wellholt: if you have a mixed environment, look into pnp
20:56
<Wellholt>
Yes
20:58
<vagrantc>
i'm guessing userful is not using off-the shelf free software to accomplish that
20:58
<Wellholt>
Nope
20:58
No
20:59
Really expesive
20:59
<maldridge>
definitly not, my site has one of those and it is a magic black box that we hope never breaks
21:00
<Wellholt>
Huh
21:00
I look
21:02
In that case you could Go for nvidia cuda viritulized GPU infrysta but so teribly expensive
21:03
what IF you like to split a CPU and GPU into 2 clients
21:08
thank for your help ;-)
21:09robb_nl has joined IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be)
21:11ogra_ has left IRC (ogra_!~ogra_@p5098ed03.dip0.t-ipconnect.de, Ping timeout: 276 seconds)
21:12ogra_ has joined IRC (ogra_!~ogra_@p5098ed03.dip0.t-ipconnect.de)
21:12Wellholt has left IRC (Wellholt!51e135b4@gateway/web/freenode/ip.81.225.53.180, Ping timeout: 250 seconds)
21:24robb_nl has left IRC (robb_nl!~robb_nl@ip-83-134-22-118.dsl.scarlet.be, Remote host closed the connection)
21:51ben_roose has left IRC (ben_roose!~roose@roose.cs.wichita.edu, Remote host closed the connection)
22:08
<vagrantc>
alkisg_away: well, uploaded ltsp 5.5.7-1 to debian, might see it in the archives in ~6 hours.
22:08
alkisg_away: if you need/want/can sync it to ubuntu yet.
22:41ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Quit: Leaving)