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


Channel log from 5 March 2012   (all times are UTC)

00:09primeministerp has left IRC (primeministerp!~ppouliot@static-71-174-244-28.bstnma.fios.verizon.net, Remote host closed the connection)
00:12primeministerp has joined IRC (primeministerp!~ppouliot@static-71-174-244-28.bstnma.fios.verizon.net)
00:18risca has left IRC (risca!~risca@wnpgmb0903w-ds01-177-34.dynamic.mtsallstream.net, Quit: Lämnar)
02:12risca has joined IRC (risca!~risca@wnpgmb0903w-ds01-177-34.dynamic.mtsallstream.net)
02:27adrianorg__ has joined IRC (adrianorg__!~adrianorg@186.215.16.184)
02:30vagrantc has left IRC (vagrantc!~vagrant@75-150-46-245-Oregon.hfc.comcastbusiness.net, Ping timeout: 255 seconds)
02:31adrianorg_ has left IRC (adrianorg_!~adrianorg@186.215.16.184, Ping timeout: 276 seconds)
02:32vagrantc has joined IRC (vagrantc!~vagrant@75-150-46-245-Oregon.hfc.comcastbusiness.net)
03:13monteslu has left IRC (monteslu!~monteslu@ip68-109-174-213.ph.ph.cox.net, Ping timeout: 255 seconds)
03:19
<vagrantc>
planning on tagging ltsp 5.3.4 tonight or tomorrow...
03:24
<stgraber>
vagrantc: sounds good, I was planning to tag at some point next week anyway, so I'll just use that one then
03:25monteslu has joined IRC (monteslu!~monteslu@ip68-109-174-213.ph.ph.cox.net)
03:26
<vagrantc>
in fact, i think i'll tag it right now...
03:27
<highvoltage>
vagrantc: whohoo! (for epoptes)
03:28
<stgraber>
vagrantc: ok. I had a quick look at trunk and it looks like I'll be able to upload 5.3.4 to Ubuntu with any need for freeze exception (looks like bugfix only to me)
03:41
<vagrantc>
stgraber: ubuntu's currently at 5.3.2?
04:05vagrantc has left IRC (vagrantc!~vagrant@75-150-46-245-Oregon.hfc.comcastbusiness.net, Ping timeout: 260 seconds)
04:07adrianorg__ has left IRC (adrianorg__!~adrianorg@186.215.16.184, Ping timeout: 245 seconds)
04:10Parker955 is now known as Parker955_Away
04:27alkisg has joined IRC (alkisg!59d2e78b@gateway/web/freenode/ip.89.210.231.139)
04:51alkisg is now known as alkisg_web
04:56alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
05:05andyGraybeals has left IRC (andyGraybeals!~andy@h215.212.22.98.dynamic.ip.windstream.net, Read error: Operation timed out)
05:13map7 has joined IRC (map7!~map7@teksup41.lnk.telstra.net)
05:28loather-work has joined IRC (loather-work!~khudson@wsip-98-175-250-115.sd.sd.cox.net)
05:59cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Remote host closed the connection)
05:59cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg)
06:47killermike has joined IRC (killermike!~killermik@2.26.103.83)
06:49risca has left IRC (risca!~risca@wnpgmb0903w-ds01-177-34.dynamic.mtsallstream.net, Quit: Lämnar)
07:18
<alkisg>
What does ssh-hostchecker do? http://bazaar.launchpad.net/~ltsp-upstream/ltsp/ldm-trunk/revision/1401
07:24
Ah, it strips the IP/hostname part of ssh_known_hosts, allowing the server IP to be changed without the clients being unable to log in?
07:28
But isn't the ltsp server always known as "server" by the clients anyway?
07:29
<muppis>
Does it then remove IP from file so ssh doesn't complain about name / IP mismatch?
07:30
<alkisg>
If so, why would we bother putting the IP in that file, only to remove it afterwards?
07:30
<muppis>
ssh does it anyway.
07:31
<alkisg>
But we're using ltsp-update-sshkeys for that, right?
07:31
<muppis>
Yes.
07:32
I don't have possibility check that right now, I'm using d-i ssh.
07:32
<alkisg>
In ltsp-update-sshkeys: IPS=$(/sbin/ip -o -f inet addr show | sed -e 's,.* \(.*\)/.*,\1,' | grep -v '^127')
07:33
I think that we can just remove support for IPs in ltsp-update-sshkeys then...
07:33
<muppis>
You can try it.
07:37
<alkisg>
I can't reproduce the "problem" though... e.g. I have 10.160.31.10 in my ltsp chroot, I then change my server IP to 192.168.0.1, and I'm still able to ssh to the server without getting a prompt
07:38
I only get a warning about "added 192.168.0.1 to known_hosts"
07:39toscalix has joined IRC (toscalix!~toscalix@212.166.248.36)
07:40
<muppis>
I usually get mismatch warning when key has changed and only IP or name removed from known_hosts
07:40
<alkisg>
Aaah I wonder if the actual problem behind all this is that LDM uses the server IP address instead of the name "server".
07:41
<muppis>
Can be.
07:44
<alkisg>
Yup I think that's it. If, in /etc/ssh/ssh_known_hosts, I have "server" and "10.160.31.10",
07:44
and I change my server IP to "192.168.0.1",
07:45
then I can run `ssh alkisg@server` by getting only a warning,
07:45
but `ssh alkisg@192.168.0.1` needs confirmation
07:46
That that warning would also explain why some times the first login try fails from ldm, while the next ones succeed
07:48
<muppis>
SOunds reasonable.
07:49* alkisg tries with LDM_SERVER=server in lts.conf...
07:51
<alkisg>
...works fine
07:53
!ltsp-update-sshkeys
07:53
<ltsp`>
alkisg: ltsp-update-sshkeys: If you changed your server IP, you need to run ltsp-update-sshkeys, and if you're using NBD (Ubuntu) you also need ltsp-update-image afterwards
07:53
<alkisg>
So instead of that, we could just instruct people to put LDM_SERVER=server in their lts.conf...
07:54
...until we change that to be the default
07:54
<muppis>
Instead of ltsp-update-image ?
07:54
<alkisg>
Instead of updating their ssh keys when their IP changes
07:55
<muppis>
Ah. Sure, that can be easier.
07:57
This is really neat.. I'm using my phone as ap and gw to install Ubuntu from netboot.
08:01
<alkisg>
Haha sounds fun :) At some hotel we used a phone to get 3G internet, then a usb cable to send it to a laptop, then the laptop's wifi card to secretly share it with others via the hotel's per-mac-locked wifi network. All that just because we didn't have a switch :D
08:02loather has left IRC (loather!~khudson@wsip-98-175-250-115.sd.sd.cox.net, Quit: This computer has gone to sleep)
08:02
<muppis>
:D
08:03
I'm sure if that phone can share it by self to multiple clients, but it shows up as regular ap, not as ad-hoc.
08:04
<Hyperbyte>
alkisg, Adroid has an app to share the phone's 3G via wireless, with multiple clients (access point mode) :-)
08:04
<alkisg>
Hyperbyte: that came later though :)
08:04
In 2.2 if I remember well
08:05
<Hyperbyte>
I've used it quite a few times now already... most impressive application was an audio broadcast connection, when we were broadcasting some sports match live from a sports park in Capelle, streamed audio to our studio, and then their internet gave our completely
08:05
<muppis>
Hyperbyte, that what I'm now using with my SGS II.
08:05
<Hyperbyte>
So I put my phone in access point mode and we streamed high quality audio to our studio via 3G. :-)
08:05
<muppis>
Got good 3G. :)
08:06
<Hyperbyte>
muppis, are you sure you're not gonna get a huge bill for data traffic? :)
08:06
<alkisg>
OK I think it's time to start my new ltsp-server package variant, which will use dnsmasq in proxydhcp mode by default (and as dns+tftp server), and will try to automatically configuring everything. Any proposed names? ltsp-server-dnsmasq?
08:06
It'll be accompanied by an ltsp-client package variant, which will have LDM_DIRECTX=true by default, and some other minor changes
08:07alexqwesa_ has left IRC (alexqwesa_!~alex@109.172.15.11, Quit: Хана X'ам !!!)
08:07
<muppis>
Hyperbyte, I'm sure. It's from work.
08:07
<Hyperbyte>
:-D
08:07
alkisg, what's it for?
08:08
<alkisg>
Hyperbyte: in many cases (greek schools included), single nic installations are used with routers as the dhcp servers
08:08
So that package will completely automate their setup
08:08
<muppis>
Well, I got it that at home.
08:08
<alkisg>
It'll allow for dual nic installations too of course
08:09
<Hyperbyte>
alkisg, then I'd call it something like 'ltsp-singlenic' or whatever. 'dnsmasq' means nothing to me.
08:09
<alkisg>
!proxy-dhcp
08:09
<ltsp`>
alkisg: proxy-dhcp: https://help.ubuntu.com/community/UbuntuLTSP/ProxyDHCP
08:09
<alkisg>
dnsmasq supports proxydhcp, while dhcp3-server doesn't
08:09
<Hyperbyte>
Well, yeah, I know what it is - I'm just saying....
08:09
Question.
08:09
<alkisg>
So the main difference is that this package will be using dnsmasq
08:10
And since it will be using it for proxydhcp (or normal dhcp), it may as well use it for dns + tftp
08:10
<Hyperbyte>
Why do you create a seperate package, rather than including some setup script with LTSP by default which makes all the configuration changes?
08:10
<alkisg>
There are 2 packages now: ltsp-server and ltsp-server-standalone
08:10
The -standalone variant uses dhcp3 and tftpd-hpa
08:10
<muppis>
It's mostly special cases where dnsmasq is needed.
08:10
<alkisg>
That new variant will use dnsmasq
08:11
<Hyperbyte>
Ah... so ltsp-server is for networks where there's already a DHCP server?
08:11
<alkisg>
I think that in the future we should be using dnsmasq in all default ltsp installations, but it's too early to convince people to switch, so for now a separate package will be needed
08:11
Yes, ltsp-server requires an appropriately configured external dhcp server
08:11
<Hyperbyte>
alkisg, why isn't ltsp-server-standalone a script that just sets the configuration options?
08:11
Why a seperate package?
08:11
<alkisg>
Because ltsp-server-standalone recommends dhcp3 and tftpd-hpa
08:12
While that new package will recommend dnsmasq instead
08:12
It's the way debian and ubuntu handle package dependencies
08:12
<Hyperbyte>
Hm
08:12
<alkisg>
We don't have an other option to tell the user to install dnsmasq instead of dhcp3 + tftpd-hpa
08:13
<Hyperbyte>
Wouldn't it make more sense then to create ltsp-config packages? ltsp-config-standalone, which runs a config script that configures ltsp-server the way you want it, and recommends the two packages
08:13
<muppis>
alkisg,are those just recommends or real depencies?
08:13
<Hyperbyte>
And ltsp-config-singlenic?
08:13
Just thinking out loud here. :)
08:14
<alkisg>
Hyperbyte: you're just using a different name, the ltsp-server-dnsmasq package just does what you say
08:14
Also, dnsmasq supports 2 nic installations too
08:14
So it's not _only_ for single nics, it just covers the proxy mode that dhcp3 doesn't
08:14
<Hyperbyte>
alkisg, ah
08:15
<alkisg>
muppis: they're recommendations, yes, but it's difficult for the user to do all the configuration manually
08:16
<muppis>
I know.
08:16
<Hyperbyte>
ltsp-server-dhcpproxy
08:16
Off to work. :P
08:16
<alkisg>
But it's not _only_ for proxy too :D
08:16
bb!
08:27
So, the ltsp-server-dnsmasq package will also have the following different defaults from upstream ltsp:
08:27
LDM_DIRECTX=True, X_COLOR_DEPTH=16 for thin clients, DNS_SERVER="$SERVER 8.8.8.8 208.67.222.222", LDM_SERVER="server" for allowing server IP changes, USE_LOCAL_SWAP=True, X_NUMLOCK=True only for desktops and not for laptops.
08:27
Anything else needed?
08:28
<muppis>
Is 208.67 OpenDNS?
08:28
<alkisg>
Yeah
08:28
Mainly the ltsp server will be used as a dns server, the others are just fallbacks
08:29
Ah, and I think I'll include a graphical pxemenu in the package too, like this: http://alkisg.mysch.gr/steki/index.php?action=dlattach;topic=2828.0;attach=1672;image
08:30loather-work has left IRC (loather-work!~khudson@wsip-98-175-250-115.sd.sd.cox.net, Quit: This computer has gone to sleep)
08:31
<muppis>
http://alkisg.mysch.gr/steki/index.php?action=dlattach;topic=2828.0;attach=1672;image
08:32
My terminal is full of crap.
08:32
<alkisg>
Hmmm and I'd better put the new tools for exporting ltsp images from virtualbox .vdi disks there as well. So I'm not sure that -dnsmasq is an appropriate name anymore...
08:34
<muppis>
-extended?
08:34
<alkisg>
Maybe... or that -ng (new generation) which I think is frequently used...
08:35
<muppis>
Or next generation.
08:35
-ng is better.
08:36Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com)
08:36
<muppis>
But I need to hit the road. Things to do and I'm still here.
08:36dobber has joined IRC (dobber!~dobber@213.169.45.222)
08:37
<alkisg>
bb :)
08:43khildin has joined IRC (khildin!~khildin@ip-80-236-227-45.dsl.scarlet.be)
08:52toscalix has left IRC (toscalix!~toscalix@212.166.248.36, Read error: Connection reset by peer)
09:03toscalix has joined IRC (toscalix!~toscalix@212.73.54.48)
09:17ogra_ has joined IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de)
09:18
<Hyperbyte>
alkisg, is X_COLOR_DEPTH=16 a big improvement?
09:18
<alkisg>
Hyperbyte: it needs half the network bandwidth
09:19
So depending on your network speed, number of clients etc, yes, it should be
09:19
<Hyperbyte>
Does it save CPU as well?
09:19
<alkisg>
If encrypted, probably. With LDM_DIRECTX=True, I don't think it'll save much cpu
09:27alexqwesa has joined IRC (alexqwesa!~alex@alexo-veto.broker.freenet6.net)
09:32
<Hyperbyte>
alkisg, it does actually
09:32
Just teeeested
09:32
YouTube fullscreen performs with -much- higher FPS, with X_COLOR_DEPTH="16"
09:32
<alkisg>
With ldm_directx=what?
09:32
<Hyperbyte>
True.
09:34
<alkisg>
That's expected... what I don't know is if the colordepth plays any role in cpu usage when ldm_directx=false
09:34
<Hyperbyte>
With X_COLOR_DEPTH="16" fullscreen flash video at 1920x1080 seems to work quite okay
09:34
If I leave that out, it eats so insanely much CPU that even the audio starts stuttering
09:34
<alkisg>
It's half the bandwidth, so from 1.6 gbps now you "only" need 800 mbps
09:34
So for 1 client a gigabit network is enough
09:35
<Hyperbyte>
Yeah, but bandwith is never the issue for me, as it's running at only 2 FPS or so... it's the CPU that can't handle it.
09:36
With X_COLOR_DEPTH="16" the CPU is much more relaxed, and FPS seems more like 15
10:10torbjorn has joined IRC (torbjorn!~torbjorn@2001:840:4051::56)
10:10
<torbjorn>
how can I add the tg3 kernle module found in kmod-tg3 to the initramfs?
10:10
usin gdracut?
10:12
oh --add-drivers apparently
10:27rickogden has joined IRC (rickogden!~rickogden@146.87.65.31)
10:30alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
10:35alexqwesa_ has joined IRC (alexqwesa_!~alex@109.172.15.11)
10:36adrianorg__ has joined IRC (adrianorg__!~adrianorg@186.215.16.184)
10:37alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
10:37alexqwesa has left IRC (alexqwesa!~alex@alexo-veto.broker.freenet6.net, Ping timeout: 260 seconds)
10:53alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
10:55rickogden has left IRC (rickogden!~rickogden@146.87.65.31, Quit: Leaving.)
11:41alexqwesa_ has left IRC (alexqwesa_!~alex@109.172.15.11, Quit: Хана X'ам !!!)
11:42wim_ has joined IRC (wim_!~chatzilla@WEGC203035.KFUNIGRAZ.AC.AT)
11:58ry has left IRC (ry!~ry@static-71-183-64-28.nycmny.fios.verizon.net, Read error: Operation timed out)
12:09rickogden has joined IRC (rickogden!~rickogden@146.87.86.186)
12:16alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
12:17artista-frustrad has joined IRC (artista-frustrad!~fernando@200.247.43.2)
12:22Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Remote host closed the connection)
12:25[GuS] has joined IRC ([GuS]!~MysT@213-117-16-190.fibertel.com.ar)
12:25[GuS] has joined IRC ([GuS]!~MysT@unaffiliated/gus/x-663402)
12:30bauerski has joined IRC (bauerski!~witekb@frodo.psp.opole.pl)
12:32khildin has left IRC (khildin!~khildin@ip-80-236-227-45.dsl.scarlet.be, Ping timeout: 276 seconds)
12:33rickogden has left IRC (rickogden!~rickogden@146.87.86.186, Quit: Leaving.)
12:36primeministerp has left IRC (primeministerp!~ppouliot@static-71-174-244-28.bstnma.fios.verizon.net, Remote host closed the connection)
12:36rickogden has joined IRC (rickogden!~rickogden@146.87.65.31)
12:41
<wim_>
Hi, I want to get cups working on FAT Clients. What is the best way to do it. I installed cups in the chroot but it did not work. Any ideas where to start? many thanks
12:43primeministerp has joined IRC (primeministerp!~ppouliot@static-71-174-244-28.bstnma.fios.verizon.net)
12:46Parker955_Away is now known as Parker955
12:50Parker955 is now known as Parker955_Away
12:53
<alkisg>
wim_: it should be working, check also the CUPS_SERVER lts.conf variable depending on what you want to do
12:53
Cups should already be installed in the chroot, wasn't it?
12:57alexqwesa has joined IRC (alexqwesa!~alex@109.172.15.11)
13:01alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Read error: Connection reset by peer)
13:01alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
13:23bengoa has joined IRC (bengoa!~bengoa@2001:1291:229:2:216:cbff:feab:6cc9)
13:39Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com)
13:41
<highvoltage>
good morning
13:43
<Hyperbyte>
hi-voltage!!!!!!!!!!!!!!!
13:48Trixboxer has joined IRC (Trixboxer!~Trixboxer@115.124.115.71)
13:49alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
13:51andyGraybeals has joined IRC (andyGraybeals!~andy@h215.212.22.98.dynamic.ip.windstream.net)
13:52bobby_C has joined IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at)
13:55Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Remote host closed the connection)
13:56Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com)
14:06
<wim_>
alkisg: txs it is not really clear to me. should i install cups in the chroot or on the server. I also could not find a lot of info about CUPS_SERVER lts.conf
14:06rickogden has left IRC (rickogden!~rickogden@146.87.65.31, Quit: Leaving.)
14:07rickogden has joined IRC (rickogden!~rickogden@146.87.86.186)
14:07rickogden1 has joined IRC (rickogden1!~rickogden@146.87.65.31)
14:07rickogden has left IRC (rickogden!~rickogden@146.87.86.186, Read error: Connection reset by peer)
14:09rickogden1 has left IRC (rickogden1!~rickogden@146.87.65.31, Client Quit)
14:09rickogden has joined IRC (rickogden!~rickogden@146.87.86.186)
14:14rickogden has left IRC (rickogden!~rickogden@146.87.86.186, Ping timeout: 245 seconds)
14:16rickogden has joined IRC (rickogden!~rickogden@146.87.86.186)
14:16||cw has left IRC (||cw!~chris@phpgroupware/cw, Read error: Connection reset by peer)
14:16||cw has joined IRC (||cw!~chris@gateway.wilsonmfg.com)
14:16||cw has joined IRC (||cw!~chris@phpgroupware/cw)
14:31bengoa has left IRC (bengoa!~bengoa@2001:1291:229:2:216:cbff:feab:6cc9, Quit: Leaving.)
14:32bengoa has joined IRC (bengoa!~bengoa@2001:1291:229:2:216:cbff:feab:6cc9)
14:34
<highvoltage>
hi perbyte :)
14:39toscalix has left IRC (toscalix!~toscalix@212.73.54.48, Ping timeout: 244 seconds)
14:41toscalix has joined IRC (toscalix!~toscalix@212.73.54.48)
14:45bengoa has left IRC (bengoa!~bengoa@2001:1291:229:2:216:cbff:feab:6cc9, Quit: Leaving.)
14:46bengoa has joined IRC (bengoa!~bengoa@2001:1291:229:2:216:cbff:feab:6cc9)
14:46rickogden has left IRC (rickogden!~rickogden@146.87.86.186, Ping timeout: 244 seconds)
14:48bauerski has left IRC (bauerski!~witekb@frodo.psp.opole.pl, Quit: Leaving.)
14:53rickogden has joined IRC (rickogden!~rickogden@146.87.86.186)
14:55* [GuS] Hi!
14:56
<Hyperbyte>
Hi GuS :-)000000
14:57toscalix has left IRC (toscalix!~toscalix@212.73.54.48, Read error: Connection reset by peer)
15:04
<[GuS]>
Hyperbyte: :)
15:07alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
15:12monteslu has left IRC (monteslu!~monteslu@ip68-109-174-213.ph.ph.cox.net, Read error: Operation timed out)
15:12risca has joined IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca)
15:13alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
15:14
<Hyperbyte>
alkisg, I just read on the K12Linux mailing list that someone tried to edit lts.conf via Webmin, and after that the thin clients wouldn't boot. Could it be that LTSP chokes on lts.conf with DOS \r\n, rather than Unix \n?
15:25monteslu has joined IRC (monteslu!~monteslu@ip68-109-174-213.ph.ph.cox.net)
15:34risca has left IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca, Quit: Lämnar)
15:34risca has joined IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca)
15:35xsl has joined IRC (xsl!~silence@unaffiliated/xsl)
15:46loather has joined IRC (loather!~khudson@wsip-98-175-250-115.sd.sd.cox.net)
15:54bobby_C has left IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at, Quit: Goin' down hard)
15:56rickogden has left IRC (rickogden!~rickogden@146.87.86.186, Quit: Leaving.)
16:01LuizAngioletti has joined IRC (LuizAngioletti!c8ef4083@gateway/web/freenode/ip.200.239.64.131)
16:01* LuizAngioletti salutes andyGraybeals
16:02mikkel has joined IRC (mikkel!~mikkel@80-71-132-15.u.parknet.dk)
16:03
<andyGraybeals>
LuizAngioletti, hi hihi
16:04
LuizAngioletti, how are things working out?
16:04
<LuizAngioletti>
slowly
16:04
but on the way
16:04
<andyGraybeals>
that is good.. i feel like since we last talked, i'm still working o nthe same thing, practically in the same spot.
16:05
<LuizAngioletti>
=)
16:07rickogden has joined IRC (rickogden!~rickogden@dab-bhx2-nat-blade-2-1.dab.02.net)
16:14Da-Geek has joined IRC (Da-Geek!~Da-Geek@50.56.228.48)
16:27dobber has left IRC (dobber!~dobber@213.169.45.222, Remote host closed the connection)
16:28Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Remote host closed the connection)
16:29risca has joined IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca)
16:30rickogden has left IRC (rickogden!~rickogden@dab-bhx2-nat-blade-2-1.dab.02.net, Quit: Leaving.)
16:47khildin has joined IRC (khildin!~khildin@ip-80-236-227-45.dsl.scarlet.be)
16:50rthomson has joined IRC (rthomson!~rthomson@mars.pet.ubc.ca)
16:55
<wim_>
Hi, I want to get cups working on FAT Clients. What is the best way to do it.
16:55
In the LTSP Server cups is working, how can I get the fatclient to use cups on the server
16:55
I also installed cups in the chroot and added in lts.conf CUPS_SERVER=192.168.2.1 (LTSP Server) but it did not work.
16:55
Any ideas where to start? many thanks
16:57
<highvoltage>
wim_: I think that if you add the printers to the server and enable printer sharing, it should just work
16:57
<wim_>
i tried but it did not work, I don't see the printers on the fatclient
16:58risca has left IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca, Ping timeout: 240 seconds)
17:01risca has joined IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca)
17:01
<wim_>
highvoltage: txs it worked, i tougt you ment 'Share printers connected to this system'. now I see the printers
17:01wim_ has left IRC (wim_!~chatzilla@WEGC203035.KFUNIGRAZ.AC.AT, Quit: ChatZilla 0.9.88 [Iceweasel 10.0.2/20120217174734])
17:12
<LuizAngioletti>
highvoltage: Occam's razor does the trick most of times, right?
17:12
*most times
17:20
andyGraybeals: Have to go. Good luck with your stuff.
17:20
=)
17:21
See you later.
17:21LuizAngioletti has left IRC (LuizAngioletti!c8ef4083@gateway/web/freenode/ip.200.239.64.131, Quit: Page closed)
17:26monteslu has left IRC (monteslu!~monteslu@ip68-109-174-213.ph.ph.cox.net, Read error: Operation timed out)
17:29risca has left IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca, Quit: Lämnar)
17:37ogra_ has left IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de, Read error: Operation timed out)
17:38alexqwesa has left IRC (alexqwesa!~alex@109.172.15.11, Quit: Хана X'ам !!!)
17:39ogra_ has joined IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de)
17:40xsl has left IRC (xsl!~silence@unaffiliated/xsl, Remote host closed the connection)
17:40alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
17:44monteslu has joined IRC (monteslu!~monteslu@ip68-109-174-213.ph.ph.cox.net)
17:44
<alkisg>
Hyperbyte: yes it's quite possible that getltscfg chokes on CRLFs... but I wonder why those K12Linux people don't file bugs in the ltsp bug tracker :)
17:49komunista has joined IRC (komunista!~slavko@adsl-195-168-242-217.dynamic.nextra.sk)
18:12
<alkisg>
Meh. Using btrfs instead of squashfs in a chroot requires 5 MB more RAM (with SCREEN_02=shell and all services deleted, from 16 MB to 21 MB).
18:15bengoa has left IRC (bengoa!~bengoa@2001:1291:229:2:216:cbff:feab:6cc9, Quit: Leaving.)
18:15bengoa1 has joined IRC (bengoa1!~bengoa@alberto.propus.com.br)
18:17komunista has left IRC (komunista!~slavko@adsl-195-168-242-217.dynamic.nextra.sk, Ping timeout: 260 seconds)
18:17Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com)
18:29Trixboxer has left IRC (Trixboxer!~Trixboxer@115.124.115.71, Quit: "Achievement is not the end, its the beginning of new journey !!!")
18:33dgroos has joined IRC (dgroos!~dgroos@x-134-84-69-105.uofm-secure.wireless.umn.edu)
18:33toscalix has joined IRC (toscalix!~toscalix@178.139.143.31)
18:45khildin has left IRC (khildin!~khildin@ip-80-236-227-45.dsl.scarlet.be, Quit: I'm gone, bye bye)
18:51Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Ping timeout: 276 seconds)
18:53Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com)
18:58
<alkisg>
Ah no the main cause for the RAM difference was that the one chroot was thin and the other fat
19:01Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Ping timeout: 276 seconds)
19:02Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com)
19:16Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Ping timeout: 276 seconds)
19:17toscalix has left IRC (toscalix!~toscalix@178.139.143.31, Ping timeout: 240 seconds)
19:18Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com)
19:31Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Ping timeout: 276 seconds)
19:33Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com)
19:42bobby_C has joined IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at)
19:46Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Ping timeout: 276 seconds)
19:47Steve_The_Pirate has joined IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com)
19:49komunista has joined IRC (komunista!~slavko@adsl-195-168-242-217.dynamic.nextra.sk)
19:52Da-Geek has left IRC (Da-Geek!~Da-Geek@50.56.228.48, Remote host closed the connection)
19:54killermike has left IRC (killermike!~killermik@2.26.103.83, Remote host closed the connection)
19:55killermike has joined IRC (killermike!~killermik@2.26.103.83)
19:55alexqwesa has joined IRC (alexqwesa!~alex@alexo-veto.broker.freenet6.net)
19:57Steve_The_Pirate has left IRC (Steve_The_Pirate!~Gary@cpc16-haye17-2-0-cust169.haye.cable.virginmedia.com, Quit: Leaving)
19:58[GuS] has left IRC ([GuS]!~MysT@unaffiliated/gus/x-663402, Remote host closed the connection)
20:14toscalix has joined IRC (toscalix!~toscalix@212.166.252.1)
20:15wim_ has joined IRC (wim_!~chatzilla@91-114-178-247.adsl.highway.telekom.at)
20:18andyGraybeals has left IRC (andyGraybeals!~andy@h215.212.22.98.dynamic.ip.windstream.net, Read error: Operation timed out)
20:21vagrantc has joined IRC (vagrantc!~vagrant@c-76-115-60-19.hsd1.or.comcast.net)
20:21andyGraybeals has joined IRC (andyGraybeals!~andy@h215.212.22.98.dynamic.ip.windstream.net)
20:23* vagrantc waves to alkisg
20:23
<alkisg>
Hey there!
20:23
Thanks for the release
20:25
<vagrantc>
sure thing
20:25
i see you put the request in to sync with ubuntu
20:26
<alkisg>
Yes I synced it an hour or so ago
20:26* alkisg has been checking memory usage in ubuntu 12.04... doesn't look very well for low-ram clients
20:27
<vagrantc>
uh-oh.
20:27
<alkisg>
I'd like to compare that with the memory usage of debian, maybe it's lighter
20:28
I'll put some more services in RM_THIN_SYSTEM_SERVICES, like network-manager, but I'm afraid that 128MB RAM will be the lowest bearable client
20:28
Ah btw stgraber, vagrantc, we still have a problem with shutdown/reboot when using nbd, because ltsp-client-core is only installed for runlevel2, no K* links. Any reason for that?
20:28
<vagrantc>
alkisg: i finally figured out a workaround for /var/cache/apt/*.bin getting recreated at boot ... apparently apt-get clean cleans those out now too, but running apt-get update or apt-cache show ltsp-client-core or just about any apt command will recreate them.
20:29
alkisg: because it never ran in that level before?
20:29
alkisg: changing the lsb headers in it might fix that
20:29
<alkisg>
But why put `dh_installinit -a --no-start -u"start 25 2 ."` instead of leaving the defaults?
20:29
At least that's what I see in debian/rules in Ubuntu
20:30
<vagrantc>
alkisg: --no-start was to prevent it from restarting during package install.
20:30
alkisg: those rules are essentially ignored for several debian releases.
20:30
<alkisg>
Ah, we don't need that anymore, we have policy-rc.d and the grep /proc/cmdline there
20:30
<vagrantc>
alkisg: insserv completely rewrites them based on lsb headers
20:31
<alkisg>
Hmm so the problem is Ubuntu-specific... stgraber?
20:31
<vagrantc>
alkisg: no, it's an issue, just where to fix it is thhe question.
20:32
alkisg: or, at least, nbd root dies badly post-5.3.1 in debina.
20:32
<alkisg>
If we just remove that line in rules there, won't the default lsb headers be used?
20:32
I tried that (update-rc.d ...defaults), and it worked fine afterwards
20:33
Hey, so apt-get clean can save us 40 MB of the squashfs image, even if we use --mount-package-cache...
20:34
Another thing... that ssh-hostkeychecker... if we used LDM_SERVER=server, instead of LDM_SERVER=server-ip, wouldn't that be redudant?
20:34zevlag has left IRC (zevlag!~zevlag@intranetgateway.etv.net, Quit: Leaving)
20:35
<vagrantc>
alkisg: the defaults will be used, which will not default to lsb headers, except that it will, unless they're not using insserv. hah.
20:35
<alkisg>
With LDM_SERVER=server, the hostname is checked, so the server can change its ip without giving login problems to the clients
20:36
<vagrantc>
alkisg: that assumes you're connecting to "server"
20:36
alkisg: with multiple application servers, that's not the case.
20:36
<alkisg>
Indeed, but one can use dns names there too
20:36
<vagrantc>
which, i guess, i do.
20:36
i don't want to get too distracted by lots of redesign today, though
20:39
<alkisg>
(10:35:12 μμ) vagrantc: alkisg: the defaults will be used, which will not default to lsb headers ==> not a problem in ubuntu though, the defaults are fine for our case
20:39
And since insserv reads the lsb headers on debian, ...we just need to do it?
20:40
<vagrantc>
not sure what the exactly correct thing to do is...
20:41
alkisg: basically, we need to set the lsb headers appropriately, in any case.
20:41
for debian
20:41
<alkisg>
I think they're fine as they are now
20:41
<vagrantc>
and then ideally, we match the lsb headers with the dh_install arguments.
20:41
otherwise we get obnoxious lintian warnings... although it's been a while since i've had that problem.
20:43
<alkisg>
Another unrelated thing... I tried btrfs + copyonwrite nbd, our init-bottom code is fine, I automatically didn't get a tmpfs/aufs overlay.
20:45
<vagrantc>
nice!
20:49toscalix has left IRC (toscalix!~toscalix@212.166.252.1, Remote host closed the connection)
20:50bengoa1 has left IRC (bengoa1!~bengoa@alberto.propus.com.br, Quit: Leaving.)
20:55toscalix_ has joined IRC (toscalix_!~toscalix@62.87.111.143)
20:56
<highvoltage>
!seen asmok
20:56
<ltsp`>
highvoltage: asmok was last seen in #ltsp 3 days, 11 hours, 25 minutes, and 14 seconds ago: <asmok> I found Gentoo howtos very usefull: http://en.gentoo-wiki.com/wiki/Category:OpenSSH
20:59toscalix_ has left IRC (toscalix_!~toscalix@62.87.111.143, Remote host closed the connection)
21:00wim_ has left IRC (wim_!~chatzilla@91-114-178-247.adsl.highway.telekom.at, Ping timeout: 245 seconds)
21:10wim_ has joined IRC (wim_!~chatzilla@91-114-178-247.adsl.highway.telekom.at)
21:14risca has joined IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca)
21:19wim_ has left IRC (wim_!~chatzilla@91-114-178-247.adsl.highway.telekom.at, Remote host closed the connection)
21:20dgroos has left IRC (dgroos!~dgroos@x-134-84-69-105.uofm-secure.wireless.umn.edu, Quit: dgroos)
21:20
<highvoltage>
stgraber: would you have some time tonight (or tomorrow night (or soon(ish))) to help me with the ltsp-cluster-lbagent packaging mess? I tried to fix it last week but there seems to be a few good options and I'd like your input. I think it might be a good idea to get rid of cdbs too and make it pure dh, but I'd like to show you a few things and get your input first
21:22
<stgraber>
highvoltage: sure, I should have some time tonight
21:22
<highvoltage>
great
21:23
<stgraber>
alkisg: Ubuntu doesn't use insserv and is trying to get rid of it
21:23
alkisg: so the right runlevels should be set in the packaging
21:23
alkisg: or we could ship an upstart job, whatever is easiest
21:23
<alkisg>
stgraber: the problem is debian/rules
21:24
We manually call dh_installinit there with S 2 only
21:24
dh_installinit -a --no-start -u"start 25 2 ."
21:24
If we change that to "defaults", I think it'll be fine
21:25
I.e. dh_installinit -a --no-start
21:25
<stgraber>
defaults will use 2345 and 016
21:25
<alkisg>
Hmm we only want 06
21:25
<stgraber>
alkisg: can you confirm it works?
21:26
alkisg: update-rc.d -f ltsp-client-core remove
21:26
<alkisg>
I tried the defaults and they worked,
21:26
<stgraber>
alkisg: update-rc.d -f ltsp-client-core defaults
21:26
<alkisg>
but I didn't try switching to single runlevel
21:26
(to see if that could kill the nbd-client connection)
21:26
OK let me try it...
21:27
<stgraber>
it probably would
21:27
though single on LTSP seems a bit silly
21:27
<alkisg>
Ah no it wouldn't, the nbd-disconnect script checks for runlevel 06
21:27
so it should be fine
21:28
<stgraber>
alkisg: would start on 2 and stop on 0 and 6 work?
21:28
<alkisg>
stgraber: I believe so, but why omit the others?
21:28
<stgraber>
so we only define what we actually know to work :)
21:29
<alkisg>
Isn't 2-5 the same in debian/ubuntu?
21:29
<stgraber>
Ubuntu doesn't really support any other runlevel and I don't think we want to see how badly it breaks if people start switching to some runlevels, then back to 2
21:29
<alkisg>
I haven't seen any sysv script that descriminates between them...
21:29
<stgraber>
and have ltsp-client-core executed twice
21:31
<alkisg>
I think that if someone tried runlevel=3, he should still get ltsp-client-core running
21:31
As it's supposed to be the same with runlevel=2
21:31killermike has left IRC (killermike!~killermik@2.26.103.83, Ping timeout: 252 seconds)
21:31
<alkisg>
(without going through runlevel=2 first, but directly to 3)
21:32
<stgraber>
I'm actually pretty tempted to just port ltsp-client-core to upstart, it'd be much shorted, much cleaner and should give us a slightly faster boot
21:32
I'll have a look at that once I'm done with my current meeting
21:32
<vagrantc>
ltsp-client-core mainly handles the screen sessions?
21:33
<alkisg>
Yes
21:33
I think I'd prefer less delta from debian, to some msec faster booting...
21:33
<stgraber>
yeah, it's basically running start_screen_sessions, start_sound, configure_sound_volume
21:34
alkisg: don't worry, we're trying to push upstart to Debian ;)
21:34
<alkisg>
:)
21:34
<stgraber>
which seems more likely to be considered than systemd at least ;)
21:34
<alkisg>
btw, any reason for us not to delete ltsp-setup?
21:35artista-frustrad has left IRC (artista-frustrad!~fernando@200.247.43.2, Quit: Leaving)
21:35
<alkisg>
I don't think anyone uses it at all now...
21:35
<stgraber>
I don't think so, if both Ubuntu and Debian don't ship it, it's probably not used anymore
21:35
not sure about Gentoo/Fedora, are they using it?
21:35
<alkisg>
Gentoo doesn't, and I _think_ fedora doesn't use either
21:36
<vagrantc>
alkisg: we'd need to get the LSB headers updated to include Default-Stop to get it to do what you want on debian
21:36
sound_volume is in udev now
21:37
<Hyperbyte>
I'm quite certain Fedora doesn't use ltsp-setup.
21:37
<alkisg>
vagrantc: OK, so, Default Start 2 3 4 5, and Default Stop 0 6 ?
21:37
<Hyperbyte>
stgraber, too bad... systemd is a beautiful thing.
21:38
<alkisg>
vagrantc: Indeed, we need to remove it from ltsp-client-core then
21:38
<Hyperbyte>
Fedora completely switched to systemd, and I absolutely love it.
21:38
<alkisg>
And I think start_sound belongs in the ldm screen script
21:38
<Hyperbyte>
Granted it took a while, definitely wasn't 'love at first sight'.
21:39
<vagrantc>
alkisg: sound support is used in non-ldm LTSP environments.
21:39
<alkisg>
vagrantc: pulseaudio? how?
21:39
<stgraber>
alkisg: can't xfreerdp use pulseaudio?
21:40
<alkisg>
stgraber: I think people turn it off deliberately to have rdesktop/xfreerdp use alsa
21:40
<vagrantc>
i even used to have a headless thin client that played music
21:40
<alkisg>
vagrantc: start_sound mainly sets up pulseaudio for ldm sessions, right?
21:40
<vagrantc>
you booted it, it started playing music fromm a randomized playlist
21:40
<alkisg>
So how would people use it without ldm?
21:41
Alsa is there, no need to "start" it...
21:41
<stgraber>
alkisg: it sets up pulseaudio as a system wide daemon
21:41
<vagrantc>
i haven't looked deeply into it.
21:41
<stgraber>
alkisg: anything running locally on the thin client can use it
21:42
<alkisg>
stgraber: normally, pulseaudio is started inside user sessions
21:42
<vagrantc>
alkisg: i'm testing with default stop added and using the installinit defaults
21:42
<alkisg>
So, for kiosk setups etc it's best to leave it to the defaults
21:43
Now, for some things running as root, like e.g. a headless media player like vagrantc says, alsa should be more than enough...
21:43
And it lags less with rdesktop/xfreerdp, and saves ram too
21:43
<stgraber>
the session should be able to detect and use the system wide puleaudio, at least it did last I check
21:44
if PULSE_SERVER=127.0.0.1 is in the environment that usually prevents another pulseaudio from starting
21:44
<alkisg>
I think Gadi was planning to move it to the user session anyway, because running it system-wide is not recommented
21:44
<vagrantc>
alkisg: regarding dh_installinit, i've only got the ltsp-client-setup using non-default values in debian/rules.
21:45
alkisg: and i'm going to just remove ltsp-client-setup.
21:45
<alkisg>
Ah, so it's an Ubuntu-only problem
21:45
vagrantc: ls -l /opt/ltsp/i386/etc/rc*/*ltsp*
21:45
<stgraber>
alkisg: moving to the user session won't work with multiple ldm sessions opened as you can't listen to the same port multiple times
21:45
alkisg: that's why it's been made system wide to start with
21:46
so you can have multiple ldm session, rdesktop sessions, ... all using the same daemon on the same port
21:46
<alkisg>
I think the main problem was that normally there wasn't a local user, to start pulseaudio in its "session"
21:47
The port could be configured according to the display number etc
21:47
<vagrantc>
alkisg: that lists ltsp-cleint-core in runlevels 2-5 and ltsp-client-setup as starting in S
21:47
<alkisg>
vagrantc: do you also have K* links?
21:48
<vagrantc>
alkisg: nope, all S links.
21:48
alkisg: will test with my upcoming patches.
21:48
<alkisg>
So Default Stop == empty should become 0 6...
21:49* vagrantc just found out yesterday that ubuntu only has the most recent 10 changes in the installed /usr/share/doc/*/changelog.debian.gz ... interesting.
21:49
<stgraber>
vagrantc: yep, space saving :)
21:50
<vagrantc>
as a developer, it would drive me crazy, but i can see the logic of it.
21:50
<stgraber>
vagrantc: "apt-get changelog ltsp-server" will give you the whole changelog
21:50
<vagrantc>
but only while online
21:50
<stgraber>
indeed
21:51
<vagrantc>
i like it in general, but i'd still like to have a way to install the whole changelog, but that's wildly off-topic here :)
21:51
does it install full upstream changelogs?
21:51
<alkisg>
A more on-topic topic would be how to handle policy-rc.d :D
21:53
<vagrantc>
policy-rc.d is considered to be a local admin tool, so it's all sorts of murky when you talk about doing it as a package.
21:54
<alkisg>
We could do it as part of ltsp-chroot, as stgraber proposed then
21:54
<stgraber>
vagrantc: I believe only the Debian changelog is stripped yes
21:55
<vagrantc>
doing policy-rc.d from ltsp-chroot makes some sense ... as long as we did it well.
21:56
unfortunately doesn't handle people using it with plain chroot, but that isn't handled all that well anyways.
21:57
<alkisg>
Should we do start-stop-daemon from ltsp-chroot too?
21:58
stgraber: when is the last date for small ltsp patches to hope to make it for precise?
21:58
<vagrantc>
now i need all sorts of additional lintian overrides with these new stop runlevels ... i think it's appropriate to override, at least.
21:58
<stgraber>
alkisg: I'm not planning on doing much LTSP updates in April, so try and get everything done in March :)
21:59
<alkisg>
Should suffice, thanks
22:08
<vagrantc>
i'm *really* confused now. nothing mentions ltsp-client-setup in my packaging, but it's failing to install because of it. :(
22:09
aha. ltsp-client-core references it
22:09
<stgraber>
I was just about to paste you that :)
22:09
./client/initscripts/ltsp-core:# Required-Start: ltsp-client-setup $remote_fs $network $syslog
22:10
which Ubuntu happily ignores ;)
22:11
<vagrantc>
i actually tested RM_SYSTEM_SERVICES=ltsp-client-setup to quick-and-dirty test how it would work without it the other day
22:12* stgraber commits a "fix" dropping any remaining reference to usplash
22:13
<stgraber>
that'll make client-core a bit shorter
22:13
<vagrantc>
heh
22:14
<stgraber>
pushed
22:15
<vagrantc>
with this upload of ltsp, my biggest fear is creating the changelog ... once i'm over that ... wheee.
22:15
<stgraber>
also updated the lsb head of -core
22:15
that should fix it for vagrantc
22:15
<vagrantc>
stgraber: i was testing it first, but ok :)
22:15
<stgraber>
I dropped the reference to -setup, added -setup as a provide and made it stop on 06
22:15
<vagrantc>
on 0 6 ?
22:15
<stgraber>
right
22:16
<vagrantc>
stgraber: at which point, just kill off ltsp-setup
22:16
<stgraber>
I also dropped the reference to hal which everyone started killing a while ago and isn't a dependency anymore
22:17
vagrantc: removed and pushed
22:19
<vagrantc>
haven't heard much about the apt-get clean bug, other than it fixes caches where the *.bin are corrupted.
22:19
so i may need to implement a workaround.
22:25
alkisg: NBD still hangs after ltsp-client-core is called in runlevel 6...
22:25* alkisg reads the logs...
22:26
<vagrantc>
rebooting/shutting down was workingg fine in 5.3.1 ... although it would leave dead nbd-server processes
22:27
if only there was a way to say to the nbd-server to please start the timeout process, but continue to work indefinitely.
22:27
<alkisg>
vagrantc: you've packaged nbd-disconnect, right?
22:27
What does ` grep KILLALL /opt/ltsp/i386/etc/nbd-client` give you?
22:28
<vagrantc>
alkisg: should be in ltsp-client-core?
22:28
<alkisg>
if [ -x /usr/share/ltsp/nbd-disconnect ]; then
22:28
/usr/share/ltsp/nbd-disconnect &
22:28
fi
22:29
ltsp-client-core calls nbd-disconnect if it's there, which does the actual job
22:29
<vagrantc>
well, that's the story.
22:29
not packaged.
22:29
<alkisg>
stgraber: we need that too ^
22:32
In my tests it always managed to shutdown/reboot without leaving connections open, even if it sometimes showed nbd read errors for a few msecs. If it doesn't work ok with insserv, we can remove the "&" at the end if the nbd-disconnect call, and use a smaller delay, e.g. 2 secs instead of 5.
22:32
<vagrantc>
will actually test it this time :)
22:33
i really need to update NEWS.Debian and/or README.Debian
22:34
to mention the big switch to writeable /
22:34risca has left IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca, Ping timeout: 244 seconds)
22:37
<vagrantc>
alkisg: yay. seems to work
22:37
alkisg: although it sends my qemu-kvm instance into a perpetual reboot cycle :(
22:37
<alkisg>
Ouch
22:39
Is that a problem with qemu-kvm and the sysrq keys?
22:39
<vagrantc>
might be
22:39
shutdown works great
22:39
i'm hoping that's just a bug in qemu-kvm...
22:40
but i sure use that feature a lot... although with NFS root mostly :)
22:40
alkisg: so nbd-disconnect works because it's using built-in code already in ram? no calls to outside binaries?
22:41
<alkisg>
vagrantc: yes, hopefully
22:41
<vagrantc>
heh
22:41
<alkisg>
It only does an "echo" after the disconnection, which is supposedly embedded in sh, or cached in ram just above
22:41
<vagrantc>
alkisg: worst case we could re-write it in C or something...
22:42
<alkisg>
Nah, we should just send a patch to wouter
22:42
<vagrantc>
that's actually the right thing to do, yes.
22:42
<alkisg>
About a special nbd-client signal which requests a disconnect *after* 10 seconds :D
22:42
<vagrantc>
alkisg: some sort of way to say : disconnect if you don't hear from us after X time
22:42
<alkisg>
Right
22:43
stgraber: about r2119, log_action_end_msg was previously log_end_msg, and now it's ...log_action_end_msg. It's a small change, but I don't think it matters, just mentioning it...
22:45komunista has left IRC (komunista!~slavko@adsl-195-168-242-217.dynamic.nextra.sk, Quit: Leaving.)
22:47* vagrantc is really happy about the improvements to initramfs hooks and the resulting simplification
22:47
<vagrantc>
init-ltsp.d is really good to.
22:47
<alkisg>
Hehe, I hope you'll say the same next year with the new configuration system :D
22:48
...which we'll design at BTS
22:48risca has joined IRC (risca!~risca@wi-secure-5637.cc.umanitoba.ca)
22:51mikkel has left IRC (mikkel!~mikkel@80-71-132-15.u.parknet.dk, Quit: Leaving)
22:55
<alkisg>
Goodnight guys
22:55alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
23:05bobby_C has left IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at, Quit: Goin' down hard)
23:06
<vagrantc>
although, now i've got to figure out how to purge ltsp-client-setup...
23:06
or just ship a no-op ltsp-client-setup