00:01 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 248 seconds) | |
00:19 | Jd0302 has joined IRC (Jd0302!~Jd0302@86.137.17.124) | |
00:19 | kd0204 has left IRC (kd0204!~Jd0302@host86-137-17-124.range86-137.btcentralplus.com, Ping timeout: 276 seconds) | |
00:20 | vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving) | |
00:36 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
01:18 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Remote host closed the connection) | |
01:29 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
01:33 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 250 seconds) | |
01:56 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
02:29 | teknkik has left IRC (teknkik!~tek@kapsi.fi, Ping timeout: 250 seconds) | |
02:30 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
02:34 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 246 seconds) | |
03:30 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
03:35 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 246 seconds) | |
04:03 | kjackal has joined IRC (kjackal!~kjackal@2a02:587:3110:5800:9478:a37d:a27e:c26b) | |
04:12 | kjackal has left IRC (kjackal!~kjackal@2a02:587:3110:5800:9478:a37d:a27e:c26b, Ping timeout: 246 seconds) | |
04:27 | <cyberorg> alkisg, hi, new precise-i386.img does not boot when root=/dev/etherd/e*, looks like aoe module is missing
| |
04:29 | r3pek, you can bzr checkout each ltsp component and run mkdst tar to create tarballs
| |
04:31 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
04:33 | <vagrantc> oh yeah, mkdst ...
| |
04:35 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 246 seconds) | |
04:36 | <cyberorg> vagrantc, hi :)
| |
04:37 | r3pek, ldm, ltsp and ltspfs is all you need, and epoptes once you have everything working
| |
05:00 | dberkholz has left IRC (dberkholz!~dberkholz@gentoo/developer/dberkholz) | |
05:11 | * vagrantc waves to cyberorg | |
05:12 | * cyberorg waves back | |
05:32 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
05:36 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 246 seconds) | |
05:42 | FGXR6 has joined IRC (FGXR6!~phantom@ppp121-44-213-216.lns20.syd7.internode.on.net) | |
05:43 | F-GT has left IRC (F-GT!~phantom@ppp121-44-85-28.lns20.syd4.internode.on.net, Ping timeout: 250 seconds) | |
05:53 | <alkisg> cyberorg: if you boot with e.g. nbd, and run modprobe aoe, doesn't it load and show the aoe devices?
| |
05:53 | vmlintu has joined IRC (vmlintu!~vmlintu@a88-112-3-40.elisa-laajakaista.fi) | |
05:53 | <alkisg> The module is there, but I haven't tested booting it with aoe
| |
05:54 | <cyberorg> alkisg, i tried modprobe aoe from busybox prompt
| |
05:54 | <alkisg> Good enough, and ?
| |
05:54 | Aaaah
| |
05:54 | modules=netboot might not have that
| |
05:54 | OK, let me test if that's what it is about
| |
05:54 | <cyberorg> the image before this one worked
| |
05:54 | nothing, no /dev/ether... show up
| |
05:54 | <alkisg> Yup yup I think that's it
| |
05:56 | <cyberorg> while you are at it, why not add some light DE so when it boots in fatclient mode it gets something when user log in
| |
05:57 | i have to remove the defaulting to fc, the auto setting of LTSP_FATCLIENT bit
| |
05:59 | <alkisg> cyberorg: I didn't get that... suppose we install something like openbox there, what good will it do? we don't want that thin chroot to serve fat sessions, do we?
| |
06:01 | <cyberorg> alkisg, booting it defaults to LTSP_FATCLIENT=True, when user logs in nothing happens
| |
06:01 | <alkisg> cyberorg: this shouldn't be so
| |
06:02 | It should default to false
| |
06:02 | <cyberorg> /usr/share/ltsp/ltsp_config.d/20fatclients
| |
06:03 | FAT_RAM_THRESHOLD=${FAT_RAM_THRESHOLD:-300}
| |
06:03 | <alkisg> Only if xsessions are found
| |
06:03 | And there are none
| |
06:03 | Right?
| |
06:04 | <cyberorg> ah ok
| |
06:05 | you are right, missed that
| |
06:12 | <alkisg> MODULES=netboot adds all the net/ modules, but misses aoe because it's in the block/ dir
| |
06:12 | I'll put it manually somehow
| |
06:14 | vagrantc: do you think it's worth a bug report to initramfs-tools there? That MODULES=netboot should include the aoe module?
| |
06:15 | <cyberorg> alkisg, not sure what changed recently, even opensuse now uses just 70M at ldm screen
| |
06:15 | <alkisg> Are you using 22light-defaults there?
| |
06:20 | Putting "aoe" to /etc/initramfs-tools/modules and running `update-initramfs -u` did it
| |
06:22 | <cyberorg> alkisg, how do you get terminal at LDM?
| |
06:22 | <alkisg> cyberorg: eeer do you mean SCREEN_07=xterm, or SCREEN_07=shell, or exporting the display and running xterm *over* LDM?
| |
06:23 | Like right clicking execute > open terminal from epoptes?
| |
06:23 | <cyberorg> ah, epoptes :)
| |
06:23 | <alkisg> You can also do it from the command line
| |
06:25 | <cyberorg> no, export DISPLAY does not work
| |
06:25 | <alkisg> export $(/usr/share/epoptes-client/get-display)
| |
06:26 | It's a bit more complicated than plain export DISPLAY, that's why we created a script for it
| |
06:28 | <cyberorg> /usr/share/epoptes-client/get-display: line 73: ck-list-sessions: command not found
| |
06:28 | <alkisg> It shouldn't matter
| |
06:28 | It should still work, just run xterm now
| |
06:31 | <cyberorg> yup, worked
| |
06:34 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
06:38 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 246 seconds) | |
06:58 | ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz) | |
07:04 | <cyberorg> alkisg, 60M for 32bit http://paste.opensuse.org/38215871 and 80M for 64bit http://paste.opensuse.org/10639276
| |
07:05 | <alkisg> cyberorg: how much does precise-i386 need at that point?
| |
07:06 | <cyberorg> image is updated?
| |
07:06 | <alkisg> For aoe? to download it again?
| |
07:06 | <cyberorg> yes
| |
07:07 | <alkisg> in 1 minute....
| |
07:07 | (it also needed a newer kernel...)
| |
07:08 | cyberorg: are you using a VM to test those images? For opensuse 32 bit, what's the less RAM that it boots with?
| |
07:08 | E.g. can it boot with 64MB RAM?
| |
07:09 | <cyberorg> its taking 140M
| |
07:09 | <alkisg> The precise-i386.img image is ready for download
| |
07:09 | OK, precise-i386.img does boot with 64MB
| |
07:10 | <cyberorg> booting normally, in virtualbox
| |
07:10 | <alkisg> Yup, I can boot a vbox client with 64 MB RAM with the precise-i386.img image
| |
07:12 | <cyberorg> i am interested in precise for i386-i586, suse only supports i686 now i think
| |
07:13 | <alkisg> We have a lot of pentium 3 PCs here with only a little RAM
| |
07:13 | Like, thousands of them
| |
07:13 | So if we can save e.g. 100 MB RAM for the thin client session, things will be smoother...
| |
07:14 | It's not worth it to spend money to buy SDRAM, we only upgrade the clients that have DDR-1 RAM or newer
| |
07:15 | <cyberorg> yup, that is why precise-i386 will be very useful
| |
07:16 | suse won't boot below 512M
| |
07:19 | main distro, leap has dropped 32bit, now only tumbleweed offers it
| |
07:19 | <alkisg> I think that 16.04 will be the last time where we'll be supporting i386 fat clients
| |
07:20 | <cyberorg> i am hoping you will keep precise alive :)
| |
07:20 | <alkisg> After that we'll tell people to download the precise-i386 thin chroot for all i386 clients
| |
07:20 | mikkel has joined IRC (mikkel!~mikkel@mail.dlvs.dk) | |
07:35 | tek__ has joined IRC (tek__!tek@kapsi.fi) | |
07:38 | <vagrantc> alkisg: sure, can add a bug about aoe modules ...
| |
07:38 | alkisg: easy enough to add it to ours, too
| |
07:49 | <cyberorg> alkisg, you need to mount precise-i386.img /mnt && btrfs filesystem defragment -r -v -c /mnt/ && umount /mnt && virt-sparsify --in-place precise-i386.img
| |
07:49 | du -h after that: 402M precise-i386.img
| |
07:49 | <alkisg> cyberorg: it's possible that the last upload was done while precise-i386.img was still mounted
| |
07:49 | Are you talking about that?
| |
07:49 | Or to minimize its size?
| |
07:50 | <cyberorg> 641M precise-i386.img without that
| |
07:50 | <alkisg> I want it fixed to 640M so that it's possible for the sysadmins to upgrade it etc
| |
07:50 | <cyberorg> yes minimize size
| |
07:50 | <alkisg> I don't want them to have to expand it manually
| |
07:50 | <cyberorg> alkisg, sparsify does not take that away
| |
07:50 | <alkisg> If we want to minimize it, we should make it squashfs
| |
07:51 | sparse files over the internet don't make much sense
| |
07:51 | And locally, I don't think they'd care for the 150 MB difference
| |
07:51 | <cyberorg> df will still show: 640M 401M 162M 72%
| |
07:52 | you can put it in tar cvf precise-i386.img.tar precise-i386.img
| |
07:52 | so downloading saves 162M :)
| |
07:54 | <alkisg> If we are going to tar it, then why not squashfs it instead, and just tell them one command to convert it to btrfs *if* they want?
| |
07:54 | Instead of untar, they could convert it to btrfs...
| |
07:54 | Same overhead, 1 command, and most of them won't even need it
| |
07:55 | Or they could just unsquashfs it
| |
07:55 | <cyberorg> no overhead, just normal tar without compression
| |
07:55 | <alkisg> And use it as a normal chroot in /opt/ltsp/i386
| |
07:55 | Overhead of typing commands
| |
07:55 | And of waiting for them to finish
| |
07:55 | <cyberorg> yeah :)
| |
07:56 | never dealt with sparse over http, but rsync handles it find
| |
07:57 | <alkisg> cyberorg: so... do you prefer a btrfs image or a squashfs image?
| |
07:57 | <cyberorg> btrfs is much better
| |
07:58 | just optimized sparse bit
| |
07:58 | <alkisg> it's easy to convert from squashfs to btrfs, just a couple of commands...
| |
07:58 | We could even have them inside the image
| |
08:00 | <cyberorg> alkisg, i know, for suse, default is squashfs, kiwi-ltsp script does the conversion if user need it
| |
08:01 | alkisg, http://paste.opensuse.org/23194965 error on boot
| |
08:02 | <alkisg> cyberorg: cat /proc/cmdline
| |
08:02 | What command line was used there?
| |
08:02 | <cyberorg> append ro initrd=boot-precise/initrd.img init=/sbin/init-ltsp forcepae root=/dev/etherd/e3.2 quiet splash
| |
08:02 | <alkisg> And, try: mkdir /tmp/e3.2; mount /dev/etherd/e3.2 /tmp/e3.2
| |
08:03 | <cyberorg> works
| |
08:04 | <alkisg> Hrm
| |
08:10 | vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving) | |
08:11 | <cyberorg> alkisg, using earlier initramfs works, new one does not
| |
08:13 | <alkisg> cyberorg: I'll test aoe locally today or tomorrow and ensure it works fine
| |
08:14 | <cyberorg> new error, etherd/e3.2 unkown partition table
| |
08:15 | manually mounting still works
| |
08:15 | thanks, see you later :)
| |
08:35 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
08:40 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 246 seconds) | |
08:46 | gvy has joined IRC (gvy!~mike@altlinux/developer/mike) | |
08:52 | kjackal has joined IRC (kjackal!~kjackal@onopfy.static.otenet.gr) | |
09:42 | <alkisg> cyberorg: can you try with this one? http://ts.sch.gr/repo/livecd/images/precise-i386.squashfs
| |
09:43 | It's 261 MB :D
| |
09:43 | I'm starting to like btrfs less... :D
| |
09:43 | <cyberorg> alkisg, hi, is it booting aoe?
| |
09:43 | <alkisg> I haven't tried, it's the same chroot, but without the btrfs partition,
| |
09:44 | <Hyperbyte> You guys playing Age of Empires again?
| |
09:44 | <alkisg> and I'm thinking that it might have been an issue with btrfs getting loaded after initramfs-tools needed it
| |
09:44 | Hyperbyte: it was a fun game indeed, back in '98!
| |
09:45 | <cyberorg> it works with the image i have from before your update
| |
09:45 | <alkisg> It might be some race condition with btrfs and aoe and the kernel version etc etc
| |
09:46 | If the same thing works without btrfs, then maybe I should be maintaining it in squashfs form...
| |
09:46 | <cyberorg> yes, after rmmod aoe and modprobe aoe the device shows up and i can manually mount it
| |
09:47 | <alkisg> cyberorg: can you try with the squashfs one? if it works, then we drop the btrfs bit
| |
09:47 | <cyberorg> so may be modprobe aoe is happening before the network is up
| |
09:47 | 50Min to download, slow net here :)
| |
09:51 | <alkisg> cyberorg: do you want to make it yourself? loop-mount your precise-i386.img to /opt/ltsp/precise-i386 and then run ltsp-update-image precise-i386
| |
09:51 | Then get the result from /opt/ltsp/images/precise-i386.img and test if that one works without issues
| |
09:51 | I.e. if the change from btrfs to squashfs solves them...
| |
09:54 | <cyberorg> trying
| |
10:00 | alkisg, no luck
| |
10:00 | <alkisg> cyberorg: what client are you using? vbox?
| |
10:02 | <cyberorg> yes
| |
10:05 | http://paste.opensuse.org/96276341 same on real hardwares also
| |
10:05 | tried bootdelay=10 as well
| |
10:06 | <alkisg> OK I'll test after I finish some other work, today or tomorrow as I said... thanks for testing
| |
10:07 | <cyberorg> np, thanks
| |
10:37 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
10:42 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 246 seconds) | |
10:47 | lbssousa has joined IRC (lbssousa!~laercio@177.143.57.35) | |
11:35 | alkisg is now known as alkisg_away | |
11:37 | <cyberorg> alkisg_away, this is the idea https://sourceforge.net/p/kiwi-ltsp/wiki/Using%20Ubuntu%20Precise%20images%20for%20old%20Thinclients/
| |
11:38 | that image is the one that worked
| |
11:54 | Faith has joined IRC (Faith!~paty_@unaffiliated/faith) | |
12:28 | Faith has left IRC (Faith!~paty_@unaffiliated/faith, Quit: Leaving) | |
12:38 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
12:43 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 246 seconds) | |
12:44 | kjackal has left IRC (kjackal!~kjackal@onopfy.static.otenet.gr, Ping timeout: 276 seconds) | |
12:59 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
13:18 | kjackal has joined IRC (kjackal!~kjackal@2a02:587:3110:5800:85f6:a9d0:6495:fdd7) | |
13:50 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 246 seconds) | |
13:54 | Faith has joined IRC (Faith!~paty_@unaffiliated/faith) | |
14:51 | ben_roose has joined IRC (ben_roose!~roose@roose.cs.wichita.edu) | |
15:50 | Phantomas has joined IRC (Phantomas!~ftsamis@ubuntu/member/phantomas) | |
15:55 | mikkel has left IRC (mikkel!~mikkel@mail.dlvs.dk, Quit: Leaving) | |
16:10 | gvy has left IRC (gvy!~mike@altlinux/developer/mike, Quit: Leaving) | |
17:20 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
18:05 | robb_nl has joined IRC (robb_nl!~robb_nl@62.235.103.62) | |
18:24 | Faith has left IRC (Faith!~paty_@unaffiliated/faith, Ping timeout: 276 seconds) | |
18:24 | Faith has joined IRC (Faith!~paty_@200.144.182.219) | |
18:25 | Faith has left IRC (Faith!~paty_@200.144.182.219, Changing host) | |
18:25 | Faith has joined IRC (Faith!~paty_@unaffiliated/faith) | |
18:31 | Faith has left IRC (Faith!~paty_@unaffiliated/faith, Ping timeout: 250 seconds) | |
18:47 | Faith has joined IRC (Faith!~paty_@143.107.231.49) | |
18:48 | bennabiy has left IRC (bennabiy!~bennabiy@unaffiliated/bennabiy, Ping timeout: 268 seconds) | |
18:49 | gehidore has left IRC (gehidore!~username@unaffiliated/man, Ping timeout: 268 seconds) | |
18:51 | pppingme has left IRC (pppingme!~pppingme@unaffiliated/pppingme, Ping timeout: 250 seconds) | |
19:01 | bennabiy has joined IRC (bennabiy!~bennabiy@unaffiliated/bennabiy) | |
19:04 | gehidore has joined IRC (gehidore!~username@unaffiliated/man) | |
19:04 | pppingme has joined IRC (pppingme!~pppingme@unaffiliated/pppingme) | |
19:23 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
19:26 | gbaman_ has joined IRC (gbaman_!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
19:27 | gbaman_ has left IRC (gbaman_!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Remote host closed the connection) | |
19:29 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 250 seconds) | |
19:36 | tek__ is now known as teknkik | |
19:38 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
19:42 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 248 seconds) | |
19:56 | lbssousa has left IRC (lbssousa!~laercio@177.143.57.35, Quit: Konversation terminated!) | |
20:09 | robb_nl has left IRC (robb_nl!~robb_nl@62.235.103.62, Ping timeout: 246 seconds) | |
20:27 | <sebd> Hello, often the LTSP boot is slowed down by DHCP requests that get replied too fast. The server is so fast that the client does not hear the first answer. Is there something one can do to avoid this ? In the case of ISC dhcpd I read about "min-secs", should we use that ? And what about NOT asking for a DHCP lease more than once per boot ? From an iPXE script one can forward the network conf through kernel para
| |
20:27 | meter "ip=" (instead of "ip=dhcp"). I'll try that.
| |
20:39 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
20:40 | Faith has left IRC (Faith!~paty_@143.107.231.49, Quit: Leaving) | |
20:44 | gbaman has left IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net, Ping timeout: 250 seconds) | |
21:04 | gbaman has joined IRC (gbaman!~gbaman@cpc15-belf9-2-0-cust171.2-1.cable.virginm.net) | |
21:24 | alkisg_away is now known as alkisg | |
21:24 | <alkisg> sebd: you can enable IPAPPEND 3 in /etc/ltsp/update-kernels.conf
| |
21:24 | I don't think it's possible for the "server to be so fat that the client does not hear the first answer"
| |
21:25 | But anyway ipappend will prevent the second dhcp request...
| |
21:27 | cyberorg: I'd prefer it if the image wasn't downloaded through sourceforge but through ts.sch.gr, so that it's used as proof that it's worth it to maintain that image
| |
21:28 | I can keep older versions for as long as you like them, e.g. if we can't make aoe work with the last kernel, we can keep the older one, or both of them, no space or bandwidth restrictions there...
| |
21:29 | vmlintu has left IRC (vmlintu!~vmlintu@a88-112-3-40.elisa-laajakaista.fi, Ping timeout: 276 seconds) | |
21:45 | <alkisg> sebd: ah sorry with ipxe you're probably downloading the kernel directly instead of using pxelinux, right? Then yup you'd need to emulate ipappend 3 yourself, with ${next_server} etc etc
| |
21:45 | https://github.com/codehelp/klibc-aarch64/blob/master/usr/kinit/ipconfig/README.ipconfig
| |
21:45 | ip=<client-ip>:<server-ip>:<gw-ip>:<netmask>:<hostname>:<device>:<autoconf>
| |
21:47 | ip=${ip}:${next-server}:${gateway}:${netmask}:${hostname}:eth0:none or something....
| |
21:49 | alkisg is now known as alkisg_away | |
22:15 | kjackal has left IRC (kjackal!~kjackal@2a02:587:3110:5800:85f6:a9d0:6495:fdd7, Ping timeout: 246 seconds) | |
22:41 | ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Quit: Leaving) | |
23:08 | Phantomas has left IRC (Phantomas!~ftsamis@ubuntu/member/phantomas) | |