IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 27 February 2014 (all times are UTC)
| 00:01 | vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc) | |
| 00:52 | vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Ping timeout: 245 seconds) | |
| 00:59 | adrianor1 is now known as adrianorg | |
| 01:20 | sbalneav2 has left IRC (sbalneav2!~sbalneav@50.71.192.164, Ping timeout: 264 seconds) | |
| 01:23 | willianmazzardo has joined IRC (willianmazzardo!~textual@187.4.15.116) | |
| 01:24 | willianmazzardo has left IRC (willianmazzardo!~textual@187.4.15.116, Client Quit) | |
| 01:47 | sbalneav has joined IRC (sbalneav!~sbalneav@mail.legalaid.mb.ca) | |
| 03:30 | telex has left IRC (telex!~telex@freeshell.de, Remote host closed the connection) | |
| 03:32 | telex has joined IRC (telex!~telex@freeshell.de) | |
| 03:53 | TheProf has joined IRC (TheProf!~smuxi@76-10-176-231.dsl.teksavvy.com) | |
| 03:55 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 03:55 | <alkisg> TheProf: hi
| |
| 03:55 | <TheProf> alkisg: Hello!
| |
| 03:56 | <alkisg> TheProf: sorry I don't have much time, so, could we do this quickly over vnc?
| |
| 03:56 | Ah moment, I'll be back in 5 minutes from a connection that supports vnc...
| |
| 03:56 | <TheProf> Sure - I would just need some instructions on the vnc part.
| |
| 03:56 | vagrantc has joined IRC (vagrantc!~vagrant@2602:47:d67f:1600:215:afff:feed:b991) | |
| 03:56 | vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc) | |
| 03:56 | <alkisg> sudo apt-get install x11vnc, on the server, and I'll be back in 5 minutes for the rest...
| |
| 03:56 | <TheProf> alkisg: OK
| |
| 03:57 | alkisg: Neat - x11vnx already installed and newest version. :)
| |
| 03:57 | <vagrantc> serving the image files over NFS is really quite fast ... and has a lot of the advantages of NFS
| |
| 03:58 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Client Quit) | |
| 04:00 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 04:00 | <alkisg> TheProf: x11vnc -connect alkisg.dyndns.org
| |
| 04:00 | * vagrantc starts editing debian/changelog... | |
| 04:01 | <alkisg> vagrantc: the newer nbd-client supposedly has fixed the persistence issues... but nfs-client is preinstalled in the usual debian/ubuntu initramfs, ok, so that's one advantage...
| |
| 04:01 | <TheProf> alkisg: Ran it and then got an error at the end. X11 MIT shared memory attach failed
| |
| 04:01 | <alkisg> klibc nfsclient doesn't support resolving the server over dns, does it?
| |
| 04:01 | TheProf: if you're using a thin client, also put -noshm: x11vnc -noshm -connect alkisg.dyndns.org
| |
| 04:01 | <TheProf> alkisg: I'm running this on a thin client.
| |
| 04:01 | ah. right
| |
| 04:02 | <vagrantc> alkisg: yeah, no dns as far as i know
| |
| 04:02 | <TheProf> alkisg: what's better for you - on the server or thin client?
| |
| 04:02 | <alkisg> TheProf: same
| |
| 04:02 | Actually better thin client
| |
| 04:02 | <TheProf> OK
| |
| 04:03 | <vagrantc> alkisg: restarting NBD seems to be pretty brutal on running clients ... not sure how NFS images handle being replaced
| |
| 04:03 | i.e. you update the image, does that crash your clients?
| |
| 04:03 | <alkisg> It shouldn't...
| |
| 04:03 | <TheProf> alkisg: looks like a connection is being made...
| |
| 04:03 | <alkisg> vagrantc: But if one updates the image and restarts the nbd-server, then sure, they'd crash
| |
| 04:04 | But not if you do just one of them
| |
| 04:04 | NFS image-based root would crash too, if you did both
| |
| 04:04 | TheProf: sorry dyndns issues, try again with the ip address: x11vnc -noshm -connect 89.210.176.149
| |
| 04:05 | <TheProf> alkisg: 23:04:07 connection failed: Connection timed out
| |
| 04:05 | ah right ok
| |
| 04:05 | <vagrantc> alkisg: did both?
| |
| 04:05 | <TheProf> alkisg: Yup.
| |
| 04:05 | <alkisg> If you stopped nfs-kernel-server, and then updated /opt/ltsp/images/i386.img (which is exposed over NFS), and restarted nfs-kernel-server
| |
| 04:05 | The clients would get connected to a different disk, and probably hang
| |
| 04:06 | <vagrantc> yeah, sure, but if you didn't restart nfs-kernel-server, old clients should continue to use the old file, and new clients should use the new file...
| |
| 04:06 | same with NBD?
| |
| 04:07 | <alkisg> Yes
| |
| 04:07 | * vagrantc wonders about AoE | |
| 04:12 | <alkisg> AoE does that without even needing networking on the client :)
| |
| 04:14 | <vagrantc> heh.
| |
| 04:15 | well ltsp 5.5.x supports NFS, NBD and AoE out of the box, and NFS image support with small updates
| |
| 04:20 | <alkisg> !kvm
| |
| 04:20 | <ltsp`> kvm: Virtual thin client: kvm -vga vmware -ctrl-grab -no-shutdown -kernel /var/lib/tftpboot/ltsp/i386/vmlinuz -initrd /var/lib/tftpboot/ltsp/i386/initrd.img -append "ro init=/sbin/init-ltsp root=/dev/nbd0 ip=:10.160.67.10::"
| |
| 04:26 | TheProf_ has joined IRC (TheProf_!~chatzilla@76-10-176-231.dsl.teksavvy.com) | |
| 04:26 | <TheProf_> alkisg: Hello - sorry I was just saying I was booting up another machine to get on irc
| |
| 04:27 | alkisg: I was also going to mention that the reason the thin clients will not boot now is because a change was made to lts.conf in /var/lib...
| |
| 04:27 | <alkisg> TheProf_: ok, let's revert i
| |
| 04:27 | t
| |
| 04:28 | TheProf_: is that ok?
| |
| 04:28 | If so, boot a client...
| |
| 04:29 | <TheProf_> Should be fine - I'll reboot one now
| |
| 04:30 | Yup works fine now
| |
| 04:30 | 192.168.0.81
| |
| 04:31 | alkisg: last time vnc required the ip
| |
| 04:32 | <alkisg> It takes some time for the router to notify dyndns, that's all...
| |
| 04:32 | It worked now
| |
| 04:32 | <TheProf_> neat :)
| |
| 04:39 | <vagrantc> using lts.conf directly on those earlier lts.conf files produced some unexpected errors.
| |
| 04:40 | er, getltscfg
| |
| 04:40 | it's more fragile than i thought it was
| |
| 04:43 | <TheProf_> alkisg: does the IP address matter? you have 166 and the other machine is different I think.
| |
| 04:44 | <alkisg> TheProf_: I put 192.168.0.1 there, the ip of the server...
| |
| 04:44 | <TheProf_> alkisg: OK. I'll just be quiet :)
| |
| 04:44 | <alkisg> TheProf_: no no do speak :)
| |
| 04:45 | TheProf_: start with, where did that lts.conf come from?
| |
| 04:45 | The one shown in the kvm VM in the server
| |
| 04:45 | Ah, you put that in chroot/etc/lts.conf, ok
| |
| 04:45 | <TheProf_> I made that a few hours ago in an attempt to get the printer working us the other lts.conf file
| |
| 04:45 | Until today that file was empty
| |
| 04:45 | <alkisg> Right, it should be empty...
| |
| 04:46 | <TheProf_> I put it in /opt... as /var/lib wasn't working
| |
| 04:46 | Ark74 has left IRC (Ark74!~Ark74@189.214.42.128.cable.dyn.cableonline.com.mx, Quit: Saliendo) | |
| 04:46 | Ark74 has joined IRC (Ark74!~Ark74@189.214.42.128.cable.dyn.cableonline.com.mx) | |
| 04:46 | <TheProf_> As those were the two places lts.conf existed.
| |
| 04:50 | <alkisg> TheProf_: btw, now if you try to boot a real client, it won't boot?
| |
| 04:50 | <TheProf_> I don't know - I will go try
| |
| 04:50 | <alkisg> OK
| |
| 04:50 | If it works ok, you should see the tty2 shell
| |
| 04:52 | <TheProf_> alkisg: nope - hangs on the Ubuntu loading screen with the dots
| |
| 04:52 | <alkisg> The VM booted fine...
| |
| 04:52 | Lets see what happens without the dots...
| |
| 04:53 | <TheProf_> I'm using another thin client in another room -that OK?
| |
| 04:55 | <alkisg> TheProf_: is it on the same subnet?
| |
| 04:55 | (try again rebooting it)
| |
| 04:55 | <TheProf_> Yes - I meant it's not the one you were logged into before.
| |
| 04:55 | All the machines are on the same subnet and identical hardware. Rebooting...
| |
| 04:57 | alkisg: Booted into a shell
| |
| 04:57 | client 119
| |
| 04:57 | <alkisg> OK so the problem is that the clients don't see the correct server?!
| |
| 04:57 | <TheProf_> I only have the one server.
| |
| 04:57 | <alkisg> Let's recap
| |
| 04:57 | You now see the tty2 shell, right?
| |
| 04:57 | <TheProf_> Yes
| |
| 04:58 | <alkisg> Not initramfs> or anything else...
| |
| 04:58 | <TheProf_> No
| |
| 04:58 | Just the root shell login
| |
| 04:58 | <alkisg> If you use `cat /proc/cmdline`, you see ip=::192.168.0.1:
| |
| 04:58 | <TheProf_> I can go try hold on it's down the hall :)
| |
| 04:58 | <alkisg> No no
| |
| 04:58 | Wait until the next step
| |
| 04:58 | <TheProf_> OK
| |
| 04:58 | <alkisg> Now, I'll remove that ip= part from your server configuration, and then you reboot the client
| |
| 04:58 | And then the client won't boot!??
| |
| 04:59 | So, let me do the change...
| |
| 04:59 | <TheProf_> OK
| |
| 04:59 | <alkisg> Done it, so (1) check cat /proc/cmdline, (2) reboot client
| |
| 04:59 | <TheProf_> Alright off to reboot
| |
| 04:59 | <alkisg> ok
| |
| 05:02 | <TheProf_> alkisg: Hello. cat /proc/cmdline gives the whole bootup line with the ip
| |
| 05:02 | <alkisg> Nice
| |
| 05:02 | <TheProf_> when I rebooted it did boot successfully and re-running cat/proc/cmdline gives the whole line without the ip
| |
| 05:02 | <alkisg> Errr
| |
| 05:02 | * vagrantc would be interested in a summary once y'all figure this out... | |
| 05:03 | <alkisg> So, the previous change I made was to remove quiet splash
| |
| 05:03 | I'll put it back. If it boots again, we're in the same state we were a few minutes ago, when the client didn't boot :D
| |
| 05:03 | <TheProf_> maybe it was in shell behind the boot screen?
| |
| 05:03 | right.
| |
| 05:03 | <alkisg> Hmm
| |
| 05:03 | Yes, that's quite possible
| |
| 05:05 | TheProf_: reboot the client
| |
| 05:05 | <TheProf_> alkisg: OK
| |
| 05:06 | <alkisg> TheProf_: you can reboot it from epoptes too
| |
| 05:06 | I just don't know which one it is... :D
| |
| 05:07 | <TheProf_> I'm losing weight this way :)
| |
| 05:07 | vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc) | |
| 05:07 | <TheProf_> so it does not boot - it hangs on the dots. They stop blinking
| |
| 05:08 | I have a feeling the shell is running the background. crl+alt+del reboots
| |
| 05:08 | vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc) | |
| 05:08 | <alkisg> TheProf_: let's see if we can see it from epoptes
| |
| 05:08 | TheProf_: did the dots go away?
| |
| 05:08 | <TheProf_> I will go check... :)
| |
| 05:09 | no
| |
| 05:10 | <alkisg> I tried "poweroff"
| |
| 05:10 | Did the client shut down?
| |
| 05:10 | <TheProf_> checking.
| |
| 05:10 | <alkisg> If it did, boot it
| |
| 05:10 | It should now display LDM, *and* give you tty2 if you press alt+ctrl+f2
| |
| 05:12 | <TheProf_> I checked - it did not shutdown. I manually rebooted it and it loaded the GUI fine. Client 63
| |
| 05:12 | I didn't check the tty2 - I can go check?
| |
| 05:12 | <alkisg> Please do
| |
| 05:13 | <TheProf_> alkisg: Yes I get the tty2
| |
| 05:13 | <alkisg> OK so everything is fine now...
| |
| 05:13 | And the bottom line was that your clients/setup have issues with only SCREEN_02=shell
| |
| 05:14 | <TheProf_> So it's reading lts.conf in /var/lib/tftpboot...
| |
| 05:14 | <alkisg> Sure
| |
| 05:14 | The only major change I did was symlink amd64 to i386 in tftp
| |
| 05:14 | Let me revert that, and retry...
| |
| 05:14 | <TheProf_> Hmm...that's strange. I put in screen_02 in order to test if it was reading lts.conf as that was a recommended way to do so
| |
| 05:14 | <alkisg> I rebooted ltsp63
| |
| 05:15 | Sure you did fine
| |
| 05:15 | Everything else is just bugs...
| |
| 05:15 | <TheProf_> So it's expecting to find an lts.conf in i386 even if there's no 386 image
| |
| 05:15 | <alkisg> No no
| |
| 05:15 | <TheProf_> And the symlink fulfills that?
| |
| 05:15 | <alkisg> I only tried that while troubleshooting
| |
| 05:15 | <TheProf_> Ah ok.
| |
| 05:15 | <alkisg> I removed the symlink and it still works
| |
| 05:16 | 00:23:7D:1F:F9:CD
| |
| 05:16 | Ah, the clients get a different IP each time?
| |
| 05:16 | <TheProf_> I can go check if that's the MAC
| |
| 05:16 | <alkisg> It was 63 before, now 2xx..
| |
| 05:16 | <TheProf_> it's all DHCP served
| |
| 05:16 | I only specified one machine for its printer
| |
| 05:16 | <alkisg> I thought it would give the same IP until the lease expired...
| |
| 05:17 | Anyways
| |
| 05:19 | Yeah it gets a new ip each time
| |
| 05:19 | TheProf_: so, everything working ok, i also removed kvm
| |
| 05:19 | This took waaay longer than I expected... :)
| |
| 05:20 | <TheProf_> I apologize. I'm not sure why it occurred. So I could test it by putting sound=false in lts.conf and the thin client would get no sound right?
| |
| 05:20 | The purpose of this was to get the thin client's printer working.
| |
| 05:21 | It is a clean installation of edubuntu with no modifications to the configuration files except lts.conf
| |
| 05:21 | <vagrantc> alkisg: i spent a good hour trying to troubleshoot the other day as well ... it was perplexing...
| |
| 05:21 | well, spent trying to help TheProf_
| |
| 05:22 | <alkisg> TheProf_: sure, try SOUND=False or whatever else, lts.conf is working,
| |
| 05:22 | although there were some issues with the lts.conf printer directives
| |
| 05:22 | <TheProf_> alkisg: awesome. Is this fix something that would revert if I do updates, etc?
| |
| 05:22 | <alkisg> At some cases the could prevent the client from booting
| |
| 05:23 | TheProf_: after reverting all the changes, the client respects lts.conf
| |
| 05:23 | So the only real change left was the --verbose to tftpd...
| |
| 05:23 | I hope tftpd-hpa doesn't have the same issue that dnsmasq had years ago...
| |
| 05:23 | That hanged while downloading a few specific lts.conf's...
| |
| 05:23 | <TheProf_> OK...so if I was to ask what did you do to fix it, you'd say? :)
| |
| 05:23 | <alkisg> Let me revert that too, in a while...
| |
| 05:24 | <TheProf_> Don't put screen?
| |
| 05:24 | <vagrantc> tagged ltsp 5.5.1, and uploaded to Debian!
| |
| 05:24 | alkisg: do your worst!
| |
| 05:25 | <alkisg> Cool!
| |
| 05:25 | vagrantc: let's make an ltsp5 branch :)
| |
| 05:25 | And head on to ltsp6!
| |
| 05:26 | <TheProf_> Man I really need to learn Epoptes!
| |
| 05:26 | <alkisg> TheProf_: if it boots now, the only change I did was install epoptes-client in the chroot, and run ltsp-update-image
| |
| 05:26 | <TheProf_> I'd save myself miles of running.
| |
| 05:26 | <vagrantc> alkisg: maybe it's time...
| |
| 05:27 | * vagrantc had wished to get ltsp6 really specced out in october :( | |
| 05:27 | <TheProf_> alkisg: booted fine
| |
| 05:28 | <alkisg> vagrantc: I would actually prefer ltsp6 to start with a clean history, to be able to remove things that haven't been maintained in years
| |
| 05:28 | And if anyone is interested, they're welcome to put them back
| |
| 05:28 | <vagrantc> alkisg: hmmm...
| |
| 05:28 | alkisg: so then we don't have to bother with forking ltsp-trunk ... we just start ltsp6
| |
| 05:30 | <alkisg> vagrantc: sounds fine to me
| |
| 05:32 | Good morning to all, bb in 1h.
| |
| 05:32 | <TheProf_> alkisg: Thanks very much!!
| |
| 05:40 | TheProf has left IRC (TheProf!~smuxi@76-10-176-231.dsl.teksavvy.com) | |
| 05:47 | TheProf_ has left IRC (TheProf_!~chatzilla@76-10-176-231.dsl.teksavvy.com, Remote host closed the connection) | |
| 05:49 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Remote host closed the connection) | |
| 05:54 | <vagrantc> stgraber: so, there were a few non-bugfixes, but overall relatively minor changes with ltsp 5.5.1-1, you might want to consider merging
| |
| 05:55 | stgraber: the big bugfix is an issue with 5.5.0 and ltsp-pnp causing the server to have a non-bootable default boot option
| |
| 05:55 | so that's pretty important!
| |
| 05:59 | prem has joined IRC (prem!~prem@218.248.24.19) | |
| 06:05 | <stgraber> vagrantc: cool, I'll try and look at that on Friday
| |
| 06:17 | vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Quit: leaving) | |
| 06:29 | andygraybeal_ has left IRC (andygraybeal_!~andy@h52.210.22.98.dynamic.ip.windstream.net, Ping timeout: 264 seconds) | |
| 06:49 | work_alkisg is now known as alkisg | |
| 07:06 | Grembler has joined IRC (Grembler!~Ben@cpc29-aztw22-2-0-cust128.18-1.cable.virginm.net) | |
| 07:26 | alexxtasi has joined IRC (alexxtasi!~alex@unaffiliated/alexxtasi) | |
| 07:34 | goodevening has joined IRC (goodevening!~Adium@vpn.marinetec.com) | |
| 07:35 | khildin has joined IRC (khildin!~khildin@ip-213-49-84-52.dsl.scarlet.be) | |
| 07:37 | datube has joined IRC (datube!~ibuijs@192.87.54.162) | |
| 07:40 | Ark74 has left IRC (Ark74!~Ark74@189.214.42.128.cable.dyn.cableonline.com.mx, Read error: Operation timed out) | |
| 07:42 | datube has left IRC (datube!~ibuijs@192.87.54.162, Quit: Syntax error) | |
| 07:44 | datube has joined IRC (datube!~datube@192.87.54.162) | |
| 07:44 | flo1546796 has joined IRC (flo1546796!~flo154679@2001:660:5001:156:26be:5ff:fe13:f2db) | |
| 07:44 | flo1546796 has joined IRC (flo1546796!~flo154679@unaffiliated/flo1546796) | |
| 07:52 | Ark74 has joined IRC (Ark74!~Ark74@189.214.42.128.cable.dyn.cableonline.com.mx) | |
| 07:54 | gbaman_ has left IRC (gbaman_!~gbaman@host81-130-45-110.in-addr.btopenworld.com, Ping timeout: 240 seconds) | |
| 08:05 | Ark74 has left IRC (Ark74!~Ark74@189.214.42.128.cable.dyn.cableonline.com.mx, Quit: Saliendo) | |
| 08:31 | <goodevening> hey folks, can anyone one help with menu script located under screen.d? http://www.filewatcher.com/p/ltsp-client-core_5.4.2-2_mips.deb.222726/usr/share/ltsp/screen.d/menu.html Thin client stuck at Trying to load: pxelinux.cfg/default ok and nothing happens. Looks like script cant initialise. LTSP running under Ubuntu 12.04
| |
| 08:39 | <alkisg> !quiet-splash
| |
| 08:39 | <ltsp`> quiet-splash: to disable the splash screen in Ubuntu, in order to see any boot error messages, run sudo gedit /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default and remove quiet splash plymouth:force-splash vt.handoff=7
| |
| 08:39 | <alkisg> Remove quiet splash and see the last messages displayed
| |
| 08:41 | mips? do mips support pxelinux?
| |
| 08:48 | <goodevening> thx, i'll try
| |
| 08:49 | link above just for example, i have same script out the box in ubuntu 12.04 with ltsp-server-standalone
| |
| 09:04 | thanks! i've solved problem! after removing splash screen, menu comes up
| |
| 09:05 | <alkisg> You're welcome
| |
| 09:06 | <goodevening> ty a lot!
| |
| 09:32 | bobby_C has joined IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at) | |
| 09:33 | Grembler has left IRC (Grembler!~Ben@cpc29-aztw22-2-0-cust128.18-1.cable.virginm.net, Quit: I Leave) | |
| 10:00 | flo1546796 has left IRC (flo1546796!~flo154679@unaffiliated/flo1546796, Quit: Quitte) | |
| 10:45 | gdi2k has joined IRC (gdi2k!~gdi2k@222.127.48.175) | |
| 10:46 | <gdi2k> hi all, I would like to install Teamviewer on a fat client, so I installed it in the chroot, but when I run it, it says the demon isn't started, and trying to start the demon shows that the /etc/init.d/teamviewerd is missing from the chroot - why would that be? is ltsp doing something weird with /etc/init.d ?
| |
| 10:55 | <Hyperbyte> gdi2k, not weird. You're using ltsp-pnp I guess?
| |
| 10:58 | <datube> Hi all, I'm (trying) setting up a Ubuntu 12.04 LTSP cluster. I've was able to succesfully setup the basic environment (no suprise there), however, I was not able to enable NBD_SWAP. So I started to debug the scripts in (chroot) /usr/share/ltsp/init-ltsp.d. As I found out none of the variables defined on the cluster controller were available in those scripts. I've did some tweaking in the original scripts and now it works as I expected.
| |
| 10:59 | <alkisg> NBD_SWAP doesn't need to be enabled, it's pre-enabled...
| |
| 11:00 | <datube> akward, I did an out-of-the-box installation... And no NBD_SWAP at all...
| |
| 11:00 | <alkisg> How much RAM did the client have?
| |
| 11:01 | By default, they only ask for swap if they have little RAM
| |
| 11:01 | <datube> the one I'm testing on 1G
| |
| 11:01 | <alkisg> Right, so it wouldn't ask for swap if it's a thin client
| |
| 11:02 | With NBD_SWAP=True in lts.conf you could enforce the swap for all clients
| |
| 11:02 | <gdi2k> Hyperbyte, thanks. no, just regular ltsp installed on ubuntu 12.04. It seems like it may be a common Ubuntu issue with Teamviewer in general, and the daemon requires root to start. I can solve that with some sudoers magic I think?
| |
| 11:02 | <alkisg> No server-side changes should be necessary, at least for normal LTSP, I've no idea about cluster
| |
| 11:03 | <datube> well, it's a cluster client which doesn't uses lts.conf. all config should be done by cluster controller, which is also tru for "USE_LOCAL_SWAP"
| |
| 11:04 | <alkisg> OK, how it gets the ltsp config variables doesn't matter as long as they reach the client
| |
| 11:04 | <datube> I also noticed that I needed to put back "mkswap $swap" (like it was in 10.04) to be able to use the /dev/nbd${num} as swap
| |
| 11:04 | well, that the problem... they don't...
| |
| 11:05 | <alkisg> If the variables do not reach the client, it's an ltsp-cluster problem, not an ltsp-client problem
| |
| 11:05 | I don't think ltsp cluster is actively maintained anymore...
| |
| 11:05 | <datube> ow...
| |
| 11:06 | well, than i'm sorry to bother you...
| |
| 11:06 | <alkisg> In ltsp 6 many parts of it will be superceded by more general solutions
| |
| 11:06 | Like e.g. ltspd instead of lts.conf
| |
| 11:08 | (01:04:46 μμ) datube: I also noticed that I needed to put back "mkswap $swap" (like it was in 10.04) to be able to use the /dev/nbd${num} as swap ==> where?
| |
| 11:09 | $ grep mkswap /usr/sbin/nbdswapd
| |
| 11:09 | # Default to running mkswap
| |
| 11:09 | mkswap "$SWAP" > /dev/null
| |
| 11:09 | <datube> chroot: /usr/share/ltsp/init-ltsp.d/50-swap
| |
| 11:10 | <alkisg> What are the contents of this file on the server? /etc/nbd-server/conf.d/swap.conf
| |
| 11:10 | <datube> I known, but when command "swapon $swap" isseud on client, it produces an block I/O error
| |
| 11:11 | [swap]
| |
| 11:11 | exportname = /tmp/nbd-swap/%s
| |
| 11:11 | prerun = nbdswapd %s
| |
| 11:11 | postrun = rm -f %s
| |
| 11:11 | <alkisg> Sounds fine
| |
| 11:11 | And nbdswapd does call mkswap, so if you need to call it again, you have some other issue elsewhere
| |
| 11:12 | If you run nbdswapd on the server to generate a swap file and then run swapon for it, does it still complain about io errors?
| |
| 11:13 | <datube> tried that also, that worked fine
| |
| 11:13 | I know, i don't understand it either
| |
| 11:14 | to do the debugging i added "bash -l" to the "/sbin/init-ltsp" and issued the commands manually
| |
| 11:16 | <alkisg> You can use e.g. ltsp.break=50-fstab in the kernel command line to get a shell
| |
| 11:17 | ltsp.break=50-swap will give you a shell just before that script
| |
| 11:17 | <datube> I know...
| |
| 11:17 | <alkisg> Do you use encrypted swap?
| |
| 11:17 | <datube> nope
| |
| 11:17 | <alkisg> swap files are sparse, any weird file system on the server?
| |
| 11:19 | <datube> well, the only thing I can come up with is that is a KVM machine, where to host has the image on a remote storage
| |
| 11:20 | <alkisg> ...on the other hand, the problem would appear with the local `swapon` test too...
| |
| 11:22 | <datube> hmm, now I issued a "nbd-client <IP> -N swap -swap /dev/nbd12 && swapon /dev/nbd12" on a fully booted client and it doesn't gives me I/O errors
| |
| 11:26 | <alkisg> Are you using nbd-proxy?
| |
| 11:26 | ps aux|grep proxy...
| |
| 11:26 | <datube> no I don't...
| |
| 11:26 | <alkisg> (client side)
| |
| 11:26 | It's weird...
| |
| 11:27 | <datube> oeps... sorry... (moment)
| |
| 11:27 | client doesn't has nbd-proxy running
| |
| 11:28 | <alkisg> Having io errors at init-ltsp.d and not on a fully booted system is surely weird...
| |
| 11:32 | bobby_C has left IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at, Ping timeout: 240 seconds) | |
| 11:36 | <datube> I also noticed that the file "/run/resolvconf/interface/LTSP" created by "10-resolv-conf" isn't used, so the client isn't able to resolve anyting, because there is a symlink for "/etc/resolv.conf" /run/resolvconf/resolv.conf ..->
| |
| 11:37 | oeps: .. for "/etc/resolv.conf" -> /run/resolvconf/resolv.conf ..
| |
| 12:11 | Grembler has joined IRC (Grembler!~Ben@cpc29-aztw22-2-0-cust128.18-1.cable.virginm.net) | |
| 12:13 | andygraybeal_ has joined IRC (andygraybeal_!~andy@h52.210.22.98.dynamic.ip.windstream.net) | |
| 12:14 | <andygraybeal_> it's offtopic, but does anyone use an ldap directory app, like zentyal? can you recommend something ?
| |
| 12:14 | for managing user data and auth
| |
| 12:17 | <Hyperbyte> 389-DS.
| |
| 12:19 | prem has left IRC (prem!~prem@218.248.24.19, Ping timeout: 252 seconds) | |
| 12:19 | PhoenixSTF has joined IRC (PhoenixSTF!~rudi@78.29.155.245) | |
| 12:20 | <andygraybeal_> ;)
| |
| 12:20 | http://simplesamlphp.org/
| |
| 12:20 | hi Hyperbyte
| |
| 12:36 | willianmazzardo has joined IRC (willianmazzardo!~textual@187.4.15.116) | |
| 12:37 | prem has joined IRC (prem!~prem@218.248.24.19) | |
| 12:47 | prem has left IRC (prem!~prem@218.248.24.19, Ping timeout: 252 seconds) | |
| 12:54 | gadi has left IRC (gadi!~gadi@pool-108-21-187-8.nycmny.fios.verizon.net, Quit: gadi) | |
| 12:55 | gadi has joined IRC (gadi!~gadi@pool-108-21-187-8.nycmny.fios.verizon.net) | |
| 12:58 | alkisg is now known as work_alkisg | |
| 13:04 | goodevening has left IRC (goodevening!~Adium@vpn.marinetec.com, Quit: Leaving.) | |
| 13:09 | flo1546796 has joined IRC (flo1546796!~flo154679@unaffiliated/flo1546796) | |
| 13:23 | goodevening has joined IRC (goodevening!~Adium@vpn.marinetec.com) | |
| 13:26 | andygraybeal_ has left IRC (andygraybeal_!~andy@h52.210.22.98.dynamic.ip.windstream.net, Read error: Operation timed out) | |
| 13:31 | <datube> work_alkisg: in case you may need or want it; this is what I did to enable swap and cluster controller values in init-ltsp.d: http://pastebin.com/mLQRRKGk
| |
| 13:55 | adrianorg has left IRC (adrianorg!~adrianorg@179.187.25.143.dynamic.adsl.gvt.net.br, Read error: Connection reset by peer) | |
| 13:58 | adrianorg has joined IRC (adrianorg!~adrianorg@179.187.30.170.dynamic.adsl.gvt.net.br) | |
| 14:05 | bennabiy has joined IRC (bennabiy!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com) | |
| 14:09 | flo1546796 has left IRC (flo1546796!~flo154679@unaffiliated/flo1546796, Quit: Quitte) | |
| 14:14 | goodevening has left IRC (goodevening!~Adium@vpn.marinetec.com) | |
| 14:15 | andygraybeal_ has joined IRC (andygraybeal_!~andy@50.120.109.126) | |
| 14:47 | Phantomas has joined IRC (Phantomas!~Phantomas@ubuntu/member/phantomas) | |
| 15:02 | dead_inside has left IRC (dead_inside!~taylor@76.75.3.174, Quit: Leaving.) | |
| 15:04 | dead_inside has joined IRC (dead_inside!~taylor@76.75.3.174) | |
| 15:07 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 15:16 | NeonLich1 has joined IRC (NeonLich1!~NeonLicht@darwin.ugr.es) | |
| 15:17 | bennabiy1 has joined IRC (bennabiy1!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com) | |
| 15:21 | andygraybeal_ has left IRC (andygraybeal_!~andy@50.120.109.126, *.net *.split) | |
| 15:21 | bennabiy has left IRC (bennabiy!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com, *.net *.split) | |
| 15:21 | datube has left IRC (datube!~datube@192.87.54.162, *.net *.split) | |
| 15:21 | khildin has left IRC (khildin!~khildin@ip-213-49-84-52.dsl.scarlet.be, *.net *.split) | |
| 15:21 | NeonLicht has left IRC (NeonLicht!~NeonLicht@darwin.ugr.es, *.net *.split) | |
| 15:28 | datube has joined IRC (datube!~datube@541B07B1.cm-5-4a.dynamic.ziggo.nl) | |
| 15:28 | khildin has joined IRC (khildin!~khildin@ip-213-49-84-52.dsl.scarlet.be) | |
| 15:29 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Ping timeout: 264 seconds) | |
| 15:30 | andygraybeal_ has joined IRC (andygraybeal_!~andy@50.120.109.126) | |
| 15:34 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 15:39 | willianmazzardo has left IRC (willianmazzardo!~textual@187.4.15.116, Quit: Computer has gone to sleep.) | |
| 15:45 | adrianorg has left IRC (adrianorg!~adrianorg@179.187.30.170.dynamic.adsl.gvt.net.br, Ping timeout: 265 seconds) | |
| 16:00 | imox has joined IRC (imox!~imox@p57A1F4CD.dip0.t-ipconnect.de) | |
| 16:05 | andygraybeal_ has left IRC (andygraybeal_!~andy@50.120.109.126, Ping timeout: 240 seconds) | |
| 16:18 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.) | |
| 16:30 | Grembler has left IRC (Grembler!~Ben@cpc29-aztw22-2-0-cust128.18-1.cable.virginm.net, Quit: I Leave) | |
| 16:46 | vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc) | |
| 16:51 | bennabiy1 has left IRC (bennabiy1!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com, Quit: http://www.twelvetribes.org) | |
| 16:51 | bennabiy has joined IRC (bennabiy!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com) | |
| 16:56 | willianmazzardo has joined IRC (willianmazzardo!~textual@187.4.15.116) | |
| 17:04 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 17:13 | PhoenixSTF has left IRC (PhoenixSTF!~rudi@78.29.155.245, Quit: Leaving) | |
| 17:18 | <alkisg> vagrantc: I tested 5.5.1 a bit, seems fine, thanks! :)
| |
| 17:19 | <vagrantc> yay
| |
| 19:08 | Phantomas1 has joined IRC (Phantomas1!~Phantomas@ubuntu/member/phantomas) | |
| 19:08 | Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas, Ping timeout: 240 seconds) | |
| 19:14 | <gadi> Phantomas1: ping
| |
| 19:14 | <Phantomas1> Ping 20s
| |
| 19:15 | Phantomas1 is now known as Phantomas | |
| 19:15 | <Phantomas> heyo
| |
| 19:15 | <gadi> 1...2...3... ;)
| |
| 19:15 | hey
| |
| 19:15 | alkisg tells me you may be charged with putting together an ls.conf web service of sorts
| |
| 19:15 | *lts.conf
| |
| 19:16 | <alkisg> Phantomas: gadi has been a long time contributor to ltsp, and is interested in helping out with ltspd
| |
| 19:16 | <gadi> ah - it even has a name
| |
| 19:16 | :)
| |
| 19:17 | * gadi waves to alkisg | |
| 19:17 | <alkisg> vagrantc: look who's actively back! ^ ;)
| |
| 19:17 | <Phantomas> yeah I remember his nickname in this channel :D
| |
| 19:17 | <alkisg> Hey gadi, welcome back to action
| |
| 19:17 | imox has left IRC (imox!~imox@p57A1F4CD.dip0.t-ipconnect.de, Quit: imox) | |
| 19:17 | <gadi> dusting off the neurons...
| |
| 19:17 | :)
| |
| 19:17 | feels good
| |
| 19:17 | <Phantomas> gadi: Yes, I am working on it
| |
| 19:18 | * vagrantc waves to gadi | |
| 19:18 | <vagrantc> !gadi
| |
| 19:18 | <ltsp`> I do not know about 'gadi', but I do know about these similar topics: 'gpxe', 'g'
| |
| 19:18 | <gadi> Phantomas: cool. can I play?
| |
| 19:18 | * vagrantc cheers on gadi! | |
| 19:18 | <gadi> !g
| |
| 19:18 | <ltsp`> g: Gadi!!!!!!!!!!!!!!!!!!!!!!!!
| |
| 19:18 | <alkisg> We have many things we want to do for LTSP6, and we just got one of the last ltsp5 versions out yesterday...
| |
| 19:18 | <gadi> ah....
| |
| 19:18 | <Phantomas> Sure, bring your athletic shoes! :D
| |
| 19:19 | <vagrantc> i'd like to think of ltsp5 in maintenance mode, although it's hard to resist not implementing small new features
| |
| 19:19 | <gadi> cool. I hacked together a python/flask lts.conf web service the other day
| |
| 19:19 | <alkisg> Phantomas: have you pushed ltspd anywhere?
| |
| 19:19 | <gadi> (as in Tue, I think)
| |
| 19:19 | <alkisg> (e.g. launchpad...)
| |
| 19:19 | <Phantomas> alkisg: That was what I was thinking right now
| |
| 19:19 | Nope, I need to do it :)
| |
| 19:19 | <gadi> but, then I figured alkisg and his gang may aleardy be doing something
| |
| 19:20 | have you guys written a spec?
| |
| 19:21 | <alkisg> No, but we're trying to use standards whenever we can
| |
| 19:21 | The python ConfigParser has a relatively standarized format for .ini files...
| |
| 19:21 | Which is quite similar to lts.conf
| |
| 19:21 | <gadi> yeah - that's what I used
| |
| 19:21 | <alkisg> Instead of LIKE=xxx, we decided to switch to INCLUDE=x,y,z
| |
| 19:22 | * gadi nods | |
| 19:22 | <alkisg> And in version 1 we'll support [HOSTNAME*], [IP], [MAC],
| |
| 19:22 | <gadi> are you caring for writing as well as reading?
| |
| 19:22 | <alkisg> but we're also thinking of [RAM > 500] sections
| |
| 19:22 | <gadi> and security
| |
| 19:23 | <alkisg> Each client will only get the parts it needs, over http, or https if we want security
| |
| 19:23 | But we don't have anything about client identification etc drafted
| |
| 19:23 | <vagrantc> alkisg: what about [IP*] or [MAC*] ?
| |
| 19:23 | i've definitely made use of [IP*] before
| |
| 19:23 | <gadi> and writing?
| |
| 19:23 | <alkisg> *.*.*.* is IP, *:*:*:*:*:* is MAC,
| |
| 19:24 | <gadi> i.e. PUT'ing to the web service
| |
| 19:24 | and writing the conf file programmatically
| |
| 19:24 | <alkisg> and alphanumerics without three . or five : are hostnames or included sections
| |
| 19:24 | * can go anywhere
| |
| 19:24 | We haven't thought about writing
| |
| 19:24 | <vagrantc> alkisg: so 192.168.*.* not 192.168.* ?
| |
| 19:25 | <alkisg> vagrantc: it's still debatable, but we need a clean way to separate all those
| |
| 19:25 | <vagrantc> alkisg: ok.
| |
| 19:25 | <alkisg> To avoid having to write [HOST = xx] or [MAC = yy]
| |
| 19:25 | <gadi> do you have notes somewhere?
| |
| 19:25 | <alkisg> We're also thinking of supporting variables (part of ConfigParser), e.g. %serverload%
| |
| 19:26 | <vagrantc> there is something to be said for explicit, though!
| |
| 19:26 | <gadi> preferably not in Greek ;)
| |
| 19:26 | <Phantomas> Would we need writing through the web service?
| |
| 19:26 | <vagrantc> i kind of like [MAC = yy]
| |
| 19:26 | <gadi> Phantomas: it opens up a lot of options
| |
| 19:26 | <Phantomas> and holes? :D
| |
| 19:26 | <gadi> and easy to implement
| |
| 19:26 | that's what security's for
| |
| 19:27 | and the "read-only" mode by default
| |
| 19:27 | lol
| |
| 19:27 | <alkisg> vagrantc: I think it's a good opportunity to decide on these things. We just thought to be compatible with the existing lts.conf syntax, but we can surely require the [MAC = xx] part...
| |
| 19:27 | <gadi> Phantomas: are you using flask?
| |
| 19:28 | <Phantomas> Yeah, my opinion is that [MAC = 192*] would be a better idea
| |
| 19:28 | gadi: no atm
| |
| 19:28 | <alkisg> gadi: are you talking about a web-based editor, or a service that would e.g. allow a client to define its own hostname?
| |
| 19:29 | <gadi> alkisg: I was thinking more about the former (admin tools - be they web-based, mobile, etc) - easy to write a front end once you have the web service
| |
| 19:29 | well, easier
| |
| 19:29 | :)
| |
| 19:29 | <alkisg> In that case the service can be plain command line, and leave the web part to php
| |
| 19:29 | E.g. ltsp-config get "option" or ltsp-config set "option", with --mac, --host etc
| |
| 19:30 | <gadi> but, you are writing a web service anyway
| |
| 19:30 | <alkisg> (and the security part)
| |
| 19:30 | It's a SimpleHTTPServer, not an apache server :)
| |
| 19:31 | <Phantomas> gadi: I am using http.server.HTTPServer currenlty but wouldn't have any particular problems switching to flask
| |
| 19:31 | maestro_ has joined IRC (maestro_!bbac70f7@gateway/web/freenode/ip.187.172.112.247) | |
| 19:31 | <Phantomas> currently*
| |
| 19:31 | <maestro_> hi there!
| |
| 19:32 | <alkisg> Using http is good because it opens a lot of possibilities, but we shouldn't make ltsp rely on a full web server. And writing a good front-end should require a full web server with UI libraries etc, I think that's a separate task
| |
| 19:32 | Hi maestro_
| |
| 19:32 | <gadi> alkisg: I have the code written in flask and it's about 100 lines of code. Pretty sure a python web server will do the job fine
| |
| 19:32 | it's not like you are going to have 100's of mobile clients editing it at once
| |
| 19:32 | ;)
| |
| 19:33 | unless you tie it to flappybird
| |
| 19:33 | :D
| |
| 19:33 | <vagrantc> EDITS_VIA_TWITTER_FEED=true
| |
| 19:33 | <alkisg> gadi: flask is also a web server?
| |
| 19:33 | <Phantomas> Segmentation fault.
| |
| 19:34 | flask is a python web framework
| |
| 19:34 | kinda like django
| |
| 19:34 | AFAIK
| |
| 19:34 | <alkisg> I don't understand what that means
| |
| 19:34 | So, we'll also need a web server, right?
| |
| 19:35 | With hooks to run python scripts?
| |
| 19:36 | So anyways, this is about the editor
| |
| 19:36 | I think that's further ahead in the future
| |
| 19:36 | maestro_ has left IRC (maestro_!bbac70f7@gateway/web/freenode/ip.187.172.112.247, Quit: Page closed) | |
| 19:36 | <alkisg> First we should implement the configuration daemon, and then go for the editor part...
| |
| 19:36 | I think we should be deciding the new lts.conf syntax now...
| |
| 19:37 | <gadi> right. but I am just saying write it as a web service rather than a daemon
| |
| 19:37 | which you are
| |
| 19:37 | and if that web service supports writing
| |
| 19:37 | you are good to go
| |
| 19:37 | even if you are using simplehttp server
| |
| 19:38 | <alkisg> You're proposing that we would use simplehttpserver + flask together? Would that work?
| |
| 19:38 | <gadi> sure
| |
| 19:39 | flask uses it (I believe) as its built in for dev
| |
| 19:39 | I'll check
| |
| 19:40 | <alkisg> gadi, what features would flask has to offer to ltspd?
| |
| 19:40 | s/would//
| |
| 19:41 | <gadi> it is an extensible framework that makes all of the web service stuff easy
| |
| 19:41 | that's all
| |
| 19:41 | <alkisg> Our client is mostly shell
| |
| 19:41 | So it doesn't support sessions or similar stuff
| |
| 19:42 | The server gets the client request, processes it, and echoes the output
| |
| 19:42 | <gadi> right. so you could either view this as a "daemon" and build up from shell, or view it as a web service and build up from a framework that makes that easy
| |
| 19:42 | as a web service, it can all be in python, more testable, and more extensible
| |
| 19:42 | <alkisg> I've no idea about web frameworks, but I don't see in which part we'd need a framework...
| |
| 19:43 | The big part is the processing, which isn't web-related
| |
| 19:43 | The other is just standard url decoding etc
| |
| 19:43 | <gadi> how about I throw my quick implementation somewhere, you throw yours somewhere, and we can stare at them
| |
| 19:43 | <alkisg> Sure, that'd help a lot
| |
| 19:43 | <gadi> cool deal
| |
| 19:46 | * alkisg checks the features of web frameworks at http://en.wikipedia.org/wiki/Comparison_of_web_application_frameworks... ajax, i18n, databases, templates, caching... do we need any of those? | |
| 19:53 | <gadi> depends whatcha want to do. In the beginning, we don't need a UI or a database
| |
| 19:53 | we can just leverage flask for the URI routes, etc
| |
| 19:53 | and we can adhere to some good api standards
| |
| 19:54 | moving forward though, being able to swap out the conf file with a db backend or ldap backend is something we have talked about before
| |
| 19:54 | and would be made easy if we code it right
| |
| 19:54 | <alkisg> Why not use gconf settings then? :D
| |
| 19:55 | slashroot has left IRC (slashroot!~pyroman@ip-46-21-210-66.nette.pl, Ping timeout: 264 seconds) | |
| 19:56 | slashroot has joined IRC (slashroot!~pyroman@ip-46-21-210-66.nette.pl) | |
| 19:56 | <alkisg> I think that in the initial ltspd implementation, the web part will be very small, so I don't really care how it's written as long as it doesn't have many dependencies, so I'll leave you and Phantomas to figure it out
| |
| 19:57 | * gadi nods | |
| 19:57 | <alkisg> The non-web backend that does the actual work can be made a flask extension at any time, I don't think that will be a problem...
| |
| 19:57 | <gadi> is that written in python?
| |
| 19:57 | <alkisg> Yup
| |
| 19:57 | <gadi> good
| |
| 19:58 | <alkisg> I think it would be much better if it stayed in RAM, and not called on request from a web server
| |
| 19:58 | It would serve the clients a lot faster then, with less overhead
| |
| 20:00 | <gadi> we already do that
| |
| 20:00 | on the client
| |
| 20:00 | <alkisg> I'm talking about server-side
| |
| 20:01 | So that it doesn't have to re-read files, parse them, cache results etc
| |
| 20:01 | It would have everything in a few tables in RAM, and just match the client request without reading anything from the disk
| |
| 20:01 | <gadi> ah, sure
| |
| 20:01 | <alkisg> Only an inotify to check for writes in /etc/ltsp/config/
| |
| 20:01 | <vagrantc> check for updates to files via inotify...
| |
| 20:01 | * vagrantc nods | |
| 20:01 | PhoenixSTF has joined IRC (PhoenixSTF!~rudi@78.29.155.245) | |
| 20:05 | <Phantomas> I am generally focusing on the parser part currently so, I haven't done much work on the web part
| |
| 20:06 | So, if we decide to use flask, It won't be a big deal for me
| |
| 20:07 | btw, I am using python3 for this.
| |
| 20:08 | telex has left IRC (telex!~telex@freeshell.de, Ping timeout: 240 seconds) | |
| 20:08 | gbaman has joined IRC (gbaman!~gbaman@host81-130-45-110.in-addr.btopenworld.com) | |
| 20:09 | <gbaman> Remind me again, in a fat client boot, to get to the desktop, what actually is transferred to the client?
| |
| 20:10 | <vagrantc> nearly everything?
| |
| 20:10 | <gbaman> well, the os image is a few GBs
| |
| 20:10 | so I doubt all that is getting transfered
| |
| 20:10 | <vagrantc> you start program x, program x itself + anything program x depends on
| |
| 20:11 | so all the executed binaries and libraries used by those libraries
| |
| 20:11 | <gbaman> on boot though
| |
| 20:13 | telex has joined IRC (telex!~telex@freeshell.de) | |
| 20:15 | slashroot has left IRC (slashroot!~pyroman@ip-46-21-210-66.nette.pl, Ping timeout: 240 seconds) | |
| 20:16 | slashroot has joined IRC (slashroot!~pyroman@ip-46-21-210-66.nette.pl) | |
| 20:16 | telex has left IRC (telex!~telex@freeshell.de, Remote host closed the connection) | |
| 20:16 | <gbaman> brb
| |
| 20:16 | gbaman has left IRC (gbaman!~gbaman@host81-130-45-110.in-addr.btopenworld.com, ) | |
| 20:18 | gbaman has joined IRC (gbaman!~gbaman@host81-130-45-110.in-addr.btopenworld.com) | |
| 20:18 | telex has joined IRC (telex!~telex@freeshell.de) | |
| 20:18 | <gbaman> back
| |
| 20:18 | <Phantomas> gadi: I'll upload a zip somewhere for now, it's not anything special, just the base concept
| |
| 20:19 | I'll create a LP project tomorrow and a branch
| |
| 20:19 | or maybe just a branch
| |
| 20:19 | <gadi> Phantomas: sounds like a plan. LTSP is just a concept, too ;)
| |
| 20:20 | <Phantomas> gadi, alkisg: http://people.ubuntu.com/~phantomas/ltspd.zip
| |
| 20:21 | a sample client request would be wget -q -O - --post-data "MAC=00:11:25:84:CE:BA" localhost:40571
| |
| 20:21 | willianmazzardo has left IRC (willianmazzardo!~textual@187.4.15.116, Quit: Computer has gone to sleep.) | |
| 20:21 | <gbaman> ltspd?
| |
| 20:22 | <Phantomas> it only supports POST atm, and there is no detection between MAC/IP/Hostname or wildcard support yet. :(
| |
| 20:22 | As I said, nothing special :P
| |
| 20:27 | <gadi> Phantomas: http://pastebin.com/SKfZ6YAz
| |
| 20:27 | the LtsConfigParser is in a separate file, but it basically just an extension of ConfigParser, like you did
| |
| 20:28 | This guy takes JSON for the editing actions - returns text on the GETs
| |
| 20:29 | <stgraber> hey gadi, been a while!
| |
| 20:29 | vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Ping timeout: 245 seconds) | |
| 20:32 | <Phantomas> gadi: I actually like the idea of using custom routes in the http uris, and offering an edit capability.
| |
| 20:33 | PhoenixSTF has left IRC (PhoenixSTF!~rudi@78.29.155.245, Quit: Leaving) | |
| 20:35 | <Phantomas> I've never used flask, though. Do you have work done in your parser like INCLUDEs parsing (recursiveness detection etc)?
| |
| 20:36 | <gadi> stgraber: hey! only since November... :)
| |
| 20:36 | Phantomas: I didn't add the "LIKE" stuff
| |
| 20:36 | er, includes
| |
| 20:36 | :)
| |
| 20:36 | I just did some scrubbing
| |
| 20:36 | so, for example, ConfigParser doesn't barf if you have lines with leading tabs
| |
| 20:41 | Phantomas: securing the different pieces becomes easy in flask via HttpAuth
| |
| 20:41 | khildin has left IRC (khildin!~khildin@ip-213-49-84-52.dsl.scarlet.be, Quit: I'm gone, bye bye) | |
| 20:43 | <Phantomas> Sounds good
| |
| 20:46 | gadi: I'll work on the LIKE/INCLUDE stuff parsing and commit my work in LP, then we can discuss about the section format [MAC = xyz] or [xyz] etc
| |
| 20:46 | I think flask is a good option for the web service
| |
| 20:49 | another problem is that if we want things like RAM > 500 as the plan was, we need to use different operators for strings/integers/decimals etc
| |
| 20:49 | bennabiy has left IRC (bennabiy!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com, Read error: Connection reset by peer) | |
| 20:50 | <Phantomas> Though, I can't find a sane use for hostname > fatclient
| |
| 20:52 | bennabiy has joined IRC (bennabiy!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com) | |
| 20:53 | lebelges1 has joined IRC (lebelges1!52d4aa31@gateway/web/freenode/ip.82.212.170.49) | |
| 20:54 | vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc) | |
| 20:55 | lebelges1 has left IRC (lebelges1!52d4aa31@gateway/web/freenode/ip.82.212.170.49, Client Quit) | |
| 20:56 | lebelges1 has joined IRC (lebelges1!52d4aa31@gateway/web/freenode/ip.82.212.170.49) | |
| 20:58 | lebelges1 has left IRC (lebelges1!52d4aa31@gateway/web/freenode/ip.82.212.170.49, Client Quit) | |
| 20:58 | <gadi> Phantomas: sounds like overcomplicating
| |
| 20:58 | Phantomas: and perhaps at the wrong level
| |
| 20:59 | <Phantomas> the string comparison part or the RAM > 500 thing?
| |
| 20:59 | <gadi> using alternate operators
| |
| 21:00 | * gadi thinks | |
| 21:00 | <gbaman> roughly, how much data is needed to be pulled across the network to get a fat client to login screen?
| |
| 21:01 | <alkisg> 50 mb
| |
| 21:01 | * gadi was thinking 42 | |
| 21:01 | <gadi> :)
| |
| 21:02 | <alkisg> Haha
| |
| 21:02 | it's the answer to everything :)
| |
| 21:02 | <gbaman> 50mb, that it?
| |
| 21:02 | <alkisg> With compressed nbd
| |
| 21:02 | <gadi> so, any good yo mama jokes since I have been away?
| |
| 21:02 | <gbaman> nfs
| |
| 21:02 | not nbd
| |
| 21:02 | <alkisg> Nah, you're the expert on those
| |
| 21:02 | gbaman: you can still use squashfs with nfs :D
| |
| 21:02 | OK, about 125 them
| |
| 21:02 | then
| |
| 21:03 | Probably less in debian...
| |
| 21:03 | <gbaman> is debian on pi
| |
| 21:03 | <alkisg> Ubuntu tends to be a bit more fat
| |
| 21:03 | ...opposite to its mama...
| |
| 21:03 | <gadi> NICE
| |
| 21:04 | <gbaman> big talk on it is tomorrow, so finishing off slides
| |
| 21:04 | going to demo on stage live...
| |
| 21:04 | <alkisg> You can just run iperf and boot a client
| |
| 21:04 | And measure it for yourself...
| |
| 21:05 | <gbaman> did just pack it all away
| |
| 21:05 | how easy is iperf to use...
| |
| 21:06 | * vagrantc bets gbaman is using raspbian, not debian :P | |
| 21:06 | <gbaman> well, it is debian
| |
| 21:06 | just with a slight bit of raspberries added in
| |
| 21:07 | * vagrantc guffaws | |
| 21:07 | <gbaman> makes debian taste a bit better, everyone loves raspberries
| |
| 21:08 | any idea to get to LXDE loaded state how much data might have gone across the network?
| |
| 21:08 | * vagrantc sheds a tear | |
| 21:08 | <gadi> Phantomas: maybe we should have a reserved "HARDWARE-" prefix in lts.conf that always gets looked at if hardware arguments are part of the request
| |
| 21:09 | Phantomas: and under each HARDWARE action, you can put configs and tags
| |
| 21:09 | tag is like:
| |
| 21:09 | RAM = 500
| |
| 21:09 | <vagrantc> (would that be implicitly RAM >= 500 ?)
| |
| 21:09 | <gadi> and the incoming request can match on hardware sections with the appropriate tags
| |
| 21:09 | <gbaman> sadly my 32mb sd card has decided to die, so have to resort to my 128mb sd card for the demo tomorrow :(
| |
| 21:10 | * vagrantc 's netbook may be finally giving up the ghost | |
| 21:10 | <Phantomas> why not include the tags in the sections' names?
| |
| 21:10 | <gbaman> correct me if I wrong on any of this
| |
| 21:10 | pi loads kernel off sd card
| |
| 21:10 | <gadi> vagrantc: nah, I am thinking I *ask* for RAM >= 500 and I look for a hardware section's RAM tag to see if it matches
| |
| 21:10 | <gbaman> is then pointed to the server via the ip address in cmdline file
| |
| 21:11 | <gadi> hmm
| |
| 21:11 | guess that's backwards...
| |
| 21:11 | <Phantomas> i mean why name them a separate thing, "tags"
| |
| 21:11 | <gbaman> then mounts the server drive as its root file system
| |
| 21:11 | <gadi> Phantomas: you're right
| |
| 21:11 | still, something feels funny
| |
| 21:11 | <gbaman> (basically tricking it into thinking that it is a normal hdd)
| |
| 21:12 | then loads ldm and user logs in
| |
| 21:12 | that is it?
| |
| 21:12 | getting to LDM stage is about 100mb
| |
| 21:12 | <Phantomas> [MAC = 00:11:25:84:CE:BA]
| |
| 21:12 | VAR=value
| |
| 21:12 | [RAM >= 500]
| |
| 21:12 | VAR=value2
| |
| 21:12 | this is what we were thinking with alkisg
| |
| 21:13 | <vagrantc> will fuzzy matching be implicity with =, or will you want exact matching with == ?
| |
| 21:13 | <Phantomas> client would pass the RAM=xyz thing in its http request
| |
| 21:13 | <alkisg> gbaman: yup
| |
| 21:14 | * alkisg notes that there are server-side variables too, like server(s) load, and that configparser supports %(variables) syntax | |
| 21:14 | <gbaman> ok, all is good then
| |
| 21:15 | right, time to hit the sack, got a 3:30am start tomorrow
| |
| 21:15 | thanks alkisg and vagrantc and everyone else :)
| |
| 21:15 | gbaman has left IRC (gbaman!~gbaman@host81-130-45-110.in-addr.btopenworld.com, ) | |
| 21:15 | * vagrantc waves | |
| 21:15 | <Phantomas> vagrantc: what do you mean fuzzy matching? RAM = 500 and client gives RAM=501?
| |
| 21:16 | <vagrantc> Phantomas: more like [IP = 192.168.*]
| |
| 21:16 | <alkisg> glob or regex matching
| |
| 21:16 | <vagrantc> yeah, either
| |
| 21:16 | <Phantomas> oh, we were thinking something like ~= for that
| |
| 21:16 | but haven't concluded, I guess
| |
| 21:16 | <vagrantc> so = is effectively == ?
| |
| 21:16 | <Phantomas> yes
| |
| 21:18 | <vagrantc> suppose it could get expensive to default to globbing/regex matches
| |
| 21:18 | lebelges1 has joined IRC (lebelges1!52d4aa31@gateway/web/freenode/ip.82.212.170.49) | |
| 21:18 | <gadi> the whole operator thing sounds expensive
| |
| 21:18 | :)
| |
| 21:18 | <Phantomas> yes it is
| |
| 21:18 | <vagrantc> depends on what the minimum wage laws are
| |
| 21:19 | <gadi> monotonically increasing
| |
| 21:19 | :P
| |
| 21:19 | * vagrantc needs to reboot and run memtest for a while | |
| 21:20 | <vagrantc> keep it up, y'all!
| |
| 21:20 | <Phantomas> good luck
| |
| 21:20 | * gadi has some paint for you to watch dry | |
| 21:20 | <Phantomas> :P
| |
| 21:20 | vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Quit: leaving) | |
| 21:20 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Remote host closed the connection) | |
| 21:21 | <lebelges1> Hello, now I have implemented LTSP for i386 clients but I would like know if it is possible to build a client for ARM ? Thank you
| |
| 21:21 | <gadi> lebelges1: there is a how to somewhere that vagrantc wrote
| |
| 21:21 | https://wiki.debian.org/DebianEdu/LTSPArm
| |
| 21:21 | that may be it
| |
| 21:22 | <lebelges1> Ok thanks a lot :)
| |
| 21:23 | <Phantomas> bbl
| |
| 21:24 | Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas, Quit: Leaving.) | |
| 21:53 | willianmazzardo has joined IRC (willianmazzardo!~textual@187.4.15.116) | |
| 21:53 | willianmazzardo has left IRC (willianmazzardo!~textual@187.4.15.116, Client Quit) | |
| 22:00 | adrianorg has joined IRC (adrianorg!~adrianorg@179.180.171.81) | |
| 22:06 | lebelges1 has left IRC (lebelges1!52d4aa31@gateway/web/freenode/ip.82.212.170.49, Ping timeout: 245 seconds) | |
| 23:24 | PhoenixSTF has joined IRC (PhoenixSTF!~rudi@78.29.155.245) | |
| 23:37 | bennabiy has left IRC (bennabiy!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com, Ping timeout: 272 seconds) | |
| 23:38 | adrianorg has left IRC (adrianorg!~adrianorg@179.180.171.81, Ping timeout: 272 seconds) | |
| 23:38 | adrianorg has joined IRC (adrianorg!~adrianorg@179.180.171.81) | |
| 23:41 | bennabiy has joined IRC (bennabiy!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com) | |
| 23:45 | NeonLicht has joined IRC (NeonLicht!~NeonLicht@darwin.ugr.es) | |
| 23:45 | Leolo_3 has left IRC (Leolo_3!~fil@24-54-35-233.mg.cgocable.ca, Ping timeout: 272 seconds) | |
| 23:45 | PhoenixSTF has left IRC (PhoenixSTF!~rudi@78.29.155.245, Ping timeout: 272 seconds) | |
| 23:45 | NeonLich1 has left IRC (NeonLich1!~NeonLicht@darwin.ugr.es, Ping timeout: 272 seconds) | |
| 23:47 | PhoenixSTF has joined IRC (PhoenixSTF!~rudi@78.29.155.245) | |