| kerframil | yes. you need to select act upon either of those two options. a rapid workaround is to generate the en_IN locale because that will immediately 'fix' things in your current session | 00:00 |
|---|---|---|
| Extreme | kerframil: I did "update-locale LANG=en_US.UTF-8 LC_MESSAGES=POSIX" and I think it worked | 00:01 |
| Extreme | I can see this in the file now: | 00:02 |
| Extreme | LANG=en_US.UTF-8 | 00:02 |
| Extreme | LC_MESSAGES=POSIX | 00:02 |
| kerframil | Extreme: you can "echo $LANG" and run "locale" to check whether these settings are in effect | 00:02 |
| Extreme | kerframil: this is the output: http://pastie.org/6644736 | 00:03 |
| kerframil | Extreme: log in anew and run locale once again | 00:04 |
| kerframil | Extreme: it should change | 00:04 |
| Extreme | login as a different user? | 00:04 |
| kerframil | Extreme: same user, but it shouldn't matter | 00:04 |
| Extreme | okay, that worked. I can now see "en_US.UTF-8" instead of the other one. | 00:05 |
| kerframil | Extreme: as the locale is actually installed, that should be fine then | 00:06 |
| Extreme | great. Is there any way I can test? | 00:06 |
| kerframil | Extreme: um. well, you could run "perl -v". it will warn in no uncertain terms if the locale settings are messed up. | 00:08 |
| kerframil | first line would be "perl: warning: Setting locale failed." in that case | 00:08 |
| Extreme | the first line is This is perl 5, version 14, subversion 2 (v5.14.2) built for... So I guess that worked! | 00:09 |
| kerframil | correct | 00:09 |
| Extreme | kerframil: Thank you very much! :) | 00:09 |
| histo | If I sudo ufw enable while connected via ssh with the default setup block me instantly? | 01:11 |
| histo | or if I sudo ufw enable && sudo ufw enable ssh will I be good? I'm not familiar with ufw's default setup for what it blocks just wondering if someone knows | 01:11 |
| histo | Server is remote so I don't want to get hosed | 01:11 |
| sarnold | histo: I just tested on a VM; ufw enable gives a nice warning that it _might_ disrupt active ssh connections, but mine stayed alive... | 01:15 |
| sarnold | histo: all the same, you might not want to do it until you've got a serial console or local access... | 01:15 |
| histo | sarnold: hrm.. wonder if I should use nohup then on the sudo ufw enable ssh part then | 01:17 |
| histo | Something like sudo ufw enable && sudo nohup ufw enable ssh | 01:18 |
| histo | err | 01:18 |
| histo | ahh piss let's try it and see what happens | 01:18 |
| sarnold | histo: another test, "ufw allow ssh ; ufw enable" also worked.. | 01:19 |
| histo | sarnold: not familiar with ; | 01:21 |
| sarnold | histo: ; is a command separator. try "touch /tmp/foo ; ls -l /tmp/foo" | 01:21 |
| histo | similiar to && ? | 01:22 |
| sarnold | histo: yes | 01:24 |
| sarnold | ; doesn't care about the success or failure of previous commands | 01:25 |
| sarnold | && and || do care | 01:25 |
| histo | ahh | 01:25 |
| histo | anyhoot in your test ufw allow ssh prior to ufw enable worked? | 01:26 |
| sarnold | yeah, I wondered if having the rules might trip some funny established rules or something. It just seemed better to test both directions -- ufw allow ssh ; ufw enable and ufw enable ; ufw allow ssh -- before saying it was fine. :) | 01:27 |
| histo | thank you for you help by the way sarnold | 01:27 |
| jdstrand | histo: you can safely add rules before doing 'ufw enable'. so 'sudo ufw allow ssh && sudo ufw enable' should be safe. the other way works too of course, but then the connection might block | 01:43 |
| histo | Gotcha turns out the server is down anyway. Not sure what the hell the other guy did putting it at the location. Have to wait till tomorrow now. | 01:50 |
| histo | I set it up at offsite while it was at his house. He just had to mvoe it to it's final destination then tells me there's no hw firewall there. So I told him i'd use ufw after he put it in place. He botched that some how lol. | 01:51 |
| === histo2 is now known as histo | ||
| nabblet | hi. where can i find information on best practice about volumen / hard disc management. I am about to set up a data server for me and ~5 other users and was wondering which way would fit me best since I have no practical experience with setting up und maintaining a server (I used to ubuntu, have no problem with commadn line though. usually I know what i am doing ) | 07:17 |
| nabblet | so my question evolves around keywords like RAID, LVM, ZFS | 07:17 |
| === smb` is now known as smb | ||
| AgMo | -j ubuntu-id | 08:35 |
| noaXess | hey server geaks | 09:38 |
| variant | hi all.. ubuntu 12.04 minimal virtual machine here, apt / dpkg completely hangs when installing / removing sssd | 09:39 |
| variant | dpkg says it's "in a very bad state" | 09:39 |
| variant | tried a lot of stuff to fix it, any help would be appreciated | 09:39 |
| noaXess | on a customer server they made a /boot with 228mb.. yes.. now.. /boot is full.. and a new kernel installation is hanging, but can't be installed, cause /boot is full | 09:40 |
| noaXess | i wanted to purge old kernels but until new kernel is installed i can't.. i need to do apt-get -f install.. but this breaks, cause.. /boot is full | 09:40 |
| noaXess | so.. how can i reset the actuall update of the kernel.. so i can remove first old kernels? | 09:40 |
| variant | noaXess: just rm the older kernel from /boot | 09:40 |
| noaXess | variant: what about /boot/grub/grub.cfg? | 09:41 |
| variant | noaXess: once you fixed everything just run update-grub | 09:41 |
| variant | noaXess: rm a couple of older kernels and install the new one, then run update-grub | 09:41 |
| noaXess | variant: aha.. so.. remove manually from /boot, run update-grub.. and then finisch update of newest kernel | 09:41 |
| variant | yes | 09:41 |
| noaXess | variant: remove all files with same verision number, also System.map..., abi-..., config-..., initrd-img-... and so on? | 09:43 |
| variant | noaXess: you just need to free up a little space, you can then use apt to purge the rest of the old files | 09:44 |
| noaXess | moved all of the old kernel files i don't need.. update-grub and apt-get -f install is done.. | 09:45 |
| variant | noaXess: cool | 09:45 |
| noaXess | now purge old kernels.. so also dpkg is clean | 09:45 |
| variant | noaXess: yes, nice | 09:45 |
| variant | if anyone cares, i fixed my dpkg problem by booting to single user mode | 09:46 |
| variant | from where the dpkg command no longer hung when removing sssd. | 09:46 |
| noaXess | variant: your also done ;) | 09:47 |
| variant | jupp | 09:48 |
| noaXess | what if you have a problem such ours and have a m$ win server?.. get fast support in... wait... 5 minutes? | 09:49 |
| noaXess | hehe | 09:49 |
| variant | noaXess: yeah no chance hehe | 09:49 |
| variant | noaXess: you just have to be a bit carefull with who you listen to | 09:49 |
| noaXess | other question, about kernels.. is there a way, that.. if a new kernel s installed, that apt-get just keep last eg. two kernels and not all? | 09:49 |
| variant | noaXess: some people will give advice without understanding possible quesntions | 09:49 |
| variant | noaXess: yeah i think you can do that.. don't remmeber how of the top of my head though | 09:50 |
| variant | noaXess: looks like 2 kernels being saved is default | 09:53 |
| variant | noaXess: look at the comment at the top of /etc/kernel/postinst.d/apt-auto-removal | 09:53 |
| variant | noaXess: i would just leave it as it is if i were you. you could look at expanding the /boot partition which you can do wihtout rebooting | 09:54 |
| variant | assuming you have space | 09:54 |
| noaXess | variant: ok.. thanks.. | 09:54 |
| noaXess | the server will be reinstalled in the future.. so in the moment manually removing kernels is good enough | 09:55 |
| variant | afk.. | 09:55 |
| jamespage | rbasak, mongodb working on arm in raring now thanks for your work on that! | 10:38 |
| rbasak | jamespage: great! I've since got the smoke tests to pass with another patch I need to send you. It's another SIGBUS failure case. | 10:39 |
| jamespage | rbasak, so we can enable the testing for arm during package build? | 10:39 |
| rbasak | jamespage: also the smoke test is producing some warnings that I think might actually be errors that are arm-specific and should fail the test | 10:39 |
| jamespage | if so fantastic! | 10:39 |
| jamespage | rbasak, oh - its very verbose | 10:39 |
| rbasak | so I'm running an amd64 build to compare | 10:39 |
| jamespage | stacktraces and all | 10:39 |
| rbasak | There are stack traces | 10:39 |
| rbasak | Ah OK | 10:39 |
| jamespage | but it does fail tests when they are unexpected | 10:39 |
| jamespage | rbasak, I have a 2.4.0 update in testing | 10:40 |
| jamespage | rbasak, just trying to get some test coverage with ceilometer | 10:40 |
| rbasak | jamespage: you're confident in their error handling? In that case I have one final patch. | 10:40 |
| jamespage | rbasak, send it to me - I'll tie up a PPA build for ~6 hours again! | 10:41 |
| one | how do I find out why I was banned? | 11:02 |
| ikonia | one: you're in #ubuntu-ops asking, so I suspect you know | 11:04 |
| one | oh | 11:05 |
| one | well why was I banned? | 11:05 |
| one | I wish youse no harm. | 11:13 |
| one | Why am I shunned? | 11:14 |
| one | There is no other way to life and eternal. | 11:15 |
| rbasak | jamespage: build of mongodb 2.4 on armhf fails. It's a pretty common x86 assumption about signed/unsigned chars. I can fix but there may be others - I'll keep working on it | 11:28 |
| jamespage | rbasak, great - thanks | 11:32 |
| jamespage | rbasak, I uploaded your extra arm fixes and testing enablement btw | 11:32 |
| rbasak | Great - thanks! | 11:32 |
| jamespage | rbasak, please timebox this - if its looking like to alot of work then we can stick with 2.2.3 | 11:43 |
| jamespage | and push 2.4.0 out to S | 11:44 |
| rbasak | jamespage: ack | 11:44 |
| wiehan | Question: I want my ubuntu server box to be happy. Mounted in a 6u swing frame wall mounted cabinet, with only 2 fans at the top of the 6u wall box, should they blow in or suck out? | 12:00 |
| wiehan | this is the box http://goo.gl/68stC | 12:00 |
| jamespage | zul, just re-reading the pep8 thread on openstack-dev | 12:01 |
| jamespage | zul, we should not do PEP-8 checking in the package builds | 12:01 |
| jamespage | zul, code formatting is an upstream concern IMHO | 12:01 |
| jamespage | (and yes I know pep8 does more that just check formats) | 12:01 |
| jamespage | zul, what do you think? something for next cycle anyways | 12:01 |
| jamespage | zul, when you start can you take a look at https://bugs.launchpad.net/ubuntu/+source/glance/+bug/1158247 | 12:06 |
| uvirtbot | Launchpad bug 1158247 in glance "test_server_monkey_patch leaves orphaned process" [Medium,Triaged] | 12:07 |
| jamespage | its whats breaking the glance build - I had to get webops to poke it in the right way inthe distro after 21 hours to make the build go through | 12:07 |
| jamespage | zul, havana ppa etc.. - please hold off - I want to automate the branch creation and ppa seeding | 12:11 |
| jamespage | otherwise we forget how todo it each cycle | 12:11 |
| === txwikinger2 is now known as txwikinger | ||
| zul | jamespage: i was double checking last night and there are some packages that do pep8 and some that dont so turn if off in h | 12:36 |
| zul | jamespage: ill take a look at glance | 12:36 |
| jamespage | zul, ta | 12:40 |
| jamespage | zul, I'm re-syncing any mismatched deps in the ca for grizzly | 12:44 |
| zul | ack | 12:44 |
| zul | jamespage: can you check to see if we carry this python-netaddr patch please? | 12:46 |
| jamespage | zul, http://people.canonical.com/~jamespage/ca-updates/ last load of dep updates | 12:55 |
| zul | jamespage: who updated python-django-openstack-auth ? | 12:56 |
| jamespage | zul, me | 12:56 |
| zul | jamespage: okies | 12:57 |
| jamespage | zul, bug with multple project in horizon | 12:57 |
| zul | ack | 12:57 |
| zul | jamespage: +1 | 12:58 |
| jamespage | zul, ack | 13:01 |
| jamespage | ta | 13:01 |
| LargePrime | HEYO1 /sbin/modprobe tun gives FATAL: Could not load /lib/modules/3.2.13-grsec-xxxx-grs-ipv6-64/modules.dep: No such file or directory THOUGHTS? | 13:11 |
| vila | hallyn: ping | 13:22 |
| hallyn | vila: . | 13:22 |
| vila | hallyn: argh, brought the bug page to copy it here, found your comments, let me read ;) | 13:23 |
| zul | jamespage: glance fixed locally ;) | 13:30 |
| jamespage | zul, how? | 13:30 |
| zul | changed self.config(workers=1) to self.config(workers=0) | 13:30 |
| vila | hallyn: bug #1157589 updated, sorry, I missed your comments and that delayed my answer, no pressure on you ;) | 13:31 |
| uvirtbot | Launchpad bug 1157589 in kvm "kvm hangs for freebsd guests since raring upgrade" [Medium,Incomplete] https://launchpad.net/bugs/1157589 | 13:31 |
| hallyn | vila: checking | 13:33 |
| hallyn | vila: first, could you just edit the xml to set the emulator to kvm instead of kvm-spice? | 13:34 |
| hallyn | that shouldn't do it, but... | 13:34 |
| vila | hallyn: indeed, no change (sorry for the delay had some vms I shut down before doing 'service restart libvirt-bin') | 13:39 |
| mardraum | oh, I know about that one :D | 13:40 |
| mardraum | freebsd guests with serial console configured do fail to boot in raring | 13:40 |
| mardraum | that was my next bug report heh | 13:40 |
| hallyn | hm, serial console eh | 13:41 |
| mardraum | vila: I bet you have console="comconsole" in loader.conf ? | 13:41 |
| hallyn | no i've got serial and it works... | 13:41 |
| hallyn | ah | 13:42 |
| vila | mardraum: loader.conf rings no bell... where should I look for that ? | 13:42 |
| mardraum | vila: /boot/loader.conf | 13:42 |
| vila | mardraum: hmpf, can't look into that without booting first :-} | 13:42 |
| mardraum | heh yeah | 13:42 |
| jamespage | zul, just keystone and ceilometer pendings rc's right? | 13:43 |
| hallyn | but i can try adding that to my working one :) | 13:43 |
| mardraum | I'm only assuming it's the same bug since you don't know about it - but all my freebsd guests die on the kernel load using that | 13:43 |
| vila | mardraum: ha ha ! Sound workflow to the rescue ! I happen to have a copy of /boot/loader.conf, no console there | 13:44 |
| mardraum | hm | 13:44 |
| mardraum | guess I'm opening another one... | 13:44 |
| vila | mardraum: only kern.hz=100 and debug.witness.watch="-1" the last time I took a copy (which should the last time I modified it) | 13:44 |
| mardraum | lemme try it again now | 13:44 |
| vila | hallyn: should I try removing mine ? Would that be the <console> section only or something else ? | 13:45 |
| hallyn | vila: i don't think that would work, | 13:46 |
| LargePrime | Ubuntu Server 12.10; /sbin/modprobe tun gives FATAL: Could not load /lib/modules/3.2.13-grsec-xxxx-grs-ipv6-64/modules.dep: No such file or directory THOUGHTS? | 13:46 |
| mardraum | vila: I didn't have to edit the guest config to work around it | 13:46 |
| hallyn | i have a console entry in the xml | 13:46 |
| hallyn | i don't seem to have a /boot/loader.conf though | 13:46 |
| mardraum | it's not created by default in freebsd | 13:46 |
| mardraum | if it doesn't exist, you simply don't have one | 13:46 |
| vila | hallyn: not required AFAIK but I'm a freebsd noob, only use it for regression testing | 13:46 |
| vila | mardraum: note that the vm pause very early in the boot process, it stopped after displaying: 'ACPI APIC Table: <BOCHS BXPCAPIC>' | 13:48 |
| hallyn | apparently i did it wrong :) now it just hangs after loading loader.conf | 13:48 |
| vila | hallyn: how far in the boot process ? | 13:49 |
| hallyn | vila: I think you'll do better if you (1) download git://git.qemu.org/qemu.git and build it, (2) run with 'kvm -hda freebsd.img -serial stdio, and then report failures upstream | 13:49 |
| hallyn | I think adding libvirt to the mix obscures it, and I don't know enough abou freebsd OR seabios to probably help you anyway | 13:50 |
| hallyn | once you can reproduce from cmdline, you can check exactly where it's dying and see if it's a freebsd bug or a qemu one | 13:50 |
| mardraum | ok, console="comconsole" works fine on 12.10 and will not load the guest on 13.10. I remove that single line from loader.conf and it loads | 13:51 |
| mardraum | er. 13.04 | 13:51 |
| hallyn | and by 'won't load' - how does it die? | 13:52 |
| hallyn | you're I assume connecting to the serial console to see it die? | 13:52 |
| vila | hallyn: 'kvm -hda freebsd-8.0-64bits.qcow2 -serial stdio' works ! | 13:52 |
| mardraum | it doesn't get to load the kernel | 13:52 |
| mardraum | the process running is tiny | 13:52 |
| hallyn | vila: interesting, so it might be libvirt's serial console that's the problem? huh | 13:53 |
| mardraum | let me add a vnc console | 13:53 |
| vila | hallyn: yeah or something else but not kvm itself which gives me some target and a way to isolate | 13:54 |
| mardraum | openbsd serial console still works fine btw 12.10 -> 13.04 | 13:54 |
| mardraum | different implementation to freebsd | 13:54 |
| hallyn | vila: actually, could you try with libvirt again, do virsh dumpxml <vmname>, check the /dev/pts/N assigned to serial, and connect to it with minicom? | 13:55 |
| hallyn | I'm wondering whether freebsd is hanging until you connect before continuing | 13:55 |
| plars | sbeattie: can you look at http://paste.ubuntu.com/5634101/ - it was from a run I did for an oneiric kernel SRU, but I had to run by hand | 13:55 |
| vila | hallyn: argh, not that *kind* again ;) | 13:55 |
| mardraum | use screen or cu, not bloody minicom :P | 13:55 |
| hallyn | (look for 'console type='pty' and /dev/pts entry there) | 13:55 |
| plars | sbeattie: unfortunately the preinstalled images from oneiric don't lend themselves well to automation | 13:55 |
| hallyn | mardraum: screen can do that? | 13:56 |
| mardraum | yep | 13:56 |
| plars | sbeattie: that was the only failure I've seen, but it's my first time running it in that environment | 13:56 |
| mardraum | screen does serial | 13:56 |
| hallyn | cool | 13:56 |
| vila | mardraum: tell me more so I don't have to install minicom ;) | 13:56 |
| hallyn | i've already got screen -e^Aa inside screen -e^Yy, what's one more | 13:56 |
| mardraum | man screen? :p | 13:56 |
| plars | example: screen /dev/ttyUSB0 115200 | 13:57 |
| hallyn | plars++ | 13:57 |
| vila | plars: thx | 13:57 |
| plars | then normal screen commands to manage the session (such as ^A-k) to end it | 13:57 |
| plars | and mardraum: *big* +1 to not using minicom | 13:57 |
| === wedgwood_away is now known as wedgwood | ||
| mardraum | minicom is a sickness, it must be eradicated | 13:58 |
| plars | cu was my go-to tool for that for a long time, but I've gotten to kinda like screen | 13:59 |
| hallyn | h4t3rz :) | 13:59 |
| hallyn | but if i can use screen all the better. (haven't used anything but telnet to qemu serial tcp port in prolly 5 years) | 13:59 |
| plars | hallyn: nostalgic for your bbs days? | 13:59 |
| plars | hallyn: minicom reminds me too much of that, and downloading linux floppy install images over 1200 baud modem with xmodem | 13:59 |
| * plars shudders a bit | 14:00 | |
| mardraum | plars: haha yeah, I think that's part of it | 14:00 |
| zul | jamespage: yep | 14:00 |
| hallyn | heh, 2008 was not bbs days, at least for me :) | 14:00 |
| zul | jamespage: ill bug ttx and nijaba about it | 14:00 |
| mardraum | pPP WHY YOU NO WORK!? | 14:00 |
| jamespage | zul, lol | 14:00 |
| hallyn | but yes, i do get nostalgic for my vt100 over acoustically cooupled 300bad modem :) | 14:00 |
| zul | hallyn: eee eeeeern! | 14:01 |
| hallyn | *that* was when I learned about screen, btw :) | 14:01 |
| vila | hallyn: nope, nothing happened. Note that the vm is paused as in 'virsh list' says '3 freebsd9 paused' | 14:01 |
| hallyn | 9 vts on vt100, i couldn't understand why ppl wanted the whole windows thing :) | 14:01 |
| hallyn | vila: while attached to the serial, can you resumem it with virsh? | 14:02 |
| hallyn | and/or do another virsh dumpxml and pastebin the output | 14:02 |
| mardraum | hallyn: so with that single line in loader.conf in a freebsd 9.1 guest, it hangs after the "Loading /boot/defaults/loader.conf" line with a single line showing "-", which is the beginning onf the kernel load spinner | 14:03 |
| mardraum | I remove that line, it boots fine, and I can still use virsh console vm | 14:03 |
| hallyn | mardraum: ok - i wonder why i didn't get the error msgs vila got on the terminal where i started kvm though | 14:03 |
| hallyn | mardraum: and you would insist that with quantal's qemu it would proceed? | 14:04 |
| hallyn | i can build an older qemu and try it | 14:04 |
| vila | hallyn: attempt to resume from the window thingy receives an error: Error unpausing domain: internal error unable to execute QEMU command 'cont': Restting the Virtual Machine is required | 14:04 |
| mardraum | hallyn: yep same image on shared storage, just tested it with a 12.10 host ok | 14:04 |
| hallyn | mardraum: bleh. thx | 14:04 |
| hallyn | mardraum: i still suspect seabios over qemu, but let's see... | 14:05 |
| mardraum | hallyn: this is the other issue I noticed when I started testing 13.04 besides the migration one, but I had a workaround for this so it wasn't an issue so far | 14:05 |
| mardraum | hallyn: do you want me to open a bug for it? | 14:06 |
| mardraum | I'll do a fresh freebsd 9.1 amd64 install on 12.10 if so and replicate it again first if you like | 14:06 |
| hallyn | mardraum: we're working from bug 1157589 so no need for new one | 14:08 |
| mardraum | though probably tomorrow night since it's after midnight here now | 14:08 |
| uvirtbot | Launchpad bug 1157589 in kvm "kvm hangs for freebsd guests since raring upgrade" [Medium,Incomplete] https://launchpad.net/bugs/1157589 | 14:08 |
| hallyn | thx | 14:08 |
| mardraum | ok | 14:08 |
| mardraum | I'm using virtio ok in the freebsd guest still in 13.04. I suspected that at first | 14:08 |
| hallyn | yeah that didn't work for me | 14:10 |
| hallyn | livecd installer refused to recognize it | 14:10 |
| mardraum | yeah it's in ports in freebsd 9, so you need a working system, then install the port (requires full src, ugh) | 14:10 |
| mardraum | freebsd 10 finally has it in tree though | 14:11 |
| mardraum | I can upload a 9.1 amd64 build of the virtio port though, it has no depends AFAIK and pkg_add will install it | 14:12 |
| mardraum | otherwise you'll need to build a machine to build the port which has the full src tree | 14:12 |
| hallyn | wait. it didn't crash with lastest git head. maybe i was too impatient with rarings' qemu, one more try | 14:14 |
| zul | jamespage: patch sent upstream for glance | 14:15 |
| jamespage | zul, so I see | 14:15 |
| jamespage | thanks muchly for picking that up | 14:15 |
| jamespage | zul, I'm concerned we have not updated libvirt inthe cloud archive | 14:16 |
| jamespage | zul, its going to make it more difficult to track security updates etc... | 14:16 |
| zul | jamespage: im wasnt...but you are right | 14:16 |
| jamespage | zul, I might stick it in the trunk testing ppa and see how it gies | 14:16 |
| zul | jamespage: please | 14:17 |
| jamespage | zul, ack | 14:17 |
| zul | jamespage: im going to get the folsom stuff passing again on jenkins | 14:19 |
| hallyn | vila: mardraum: for me it crashes with -enable-kvm, and not without. | 14:20 |
| jamespage | zul, +1 - we really need to push the 2.3 release through | 14:20 |
| hallyn | both upstream and in raring pkg. now lemme try 1.2 - to see if it's the kernel or qemu | 14:20 |
| jamespage | we keep getting jumped by security updates three weeks into the process. | 14:20 |
| jamespage | Daviey, around? I'd like to discuss openstack SRU's if you have a moment | 14:20 |
| * jdstrand would love for the updates to stop rolling in ;) | 14:21 | |
| Daviey | jamespage: o/ | 14:22 |
| jamespage | jdstrand, you and me both :-) | 14:22 |
| jamespage | Daviey, ola - one second | 14:22 |
| * Daviey ponders. | 14:22 | |
| jamespage | Daviey, right - back | 14:24 |
| jamespage | Daviey, adam_g and I where discussing whether we could approach the first part of the SRU process a bit better | 14:24 |
| jamespage | Daviey, we agreed after the 2012.2.1 release how we would verify bugs | 14:24 |
| vila | hallyn: not sure I follow your experiments :-) Do you mean you have a way to reproduce ? | 14:24 |
| jamespage | Daviey, but its taking 2+ weeks to actually get into proposed and the rate of security bugs means we generally get superceded between initial preparation and verification | 14:25 |
| jamespage | Daviey, any thoughts? | 14:25 |
| hallyn | vila: yes, with the loader.conf line, kvm -hda freebsd.img -m 512 -serial stdio -vnc :1 -enable-kvm does it for me | 14:26 |
| hallyn | vila: without -enable-kvm does not | 14:26 |
| hallyn | sigh, but older qemu isn't compiling. | 14:26 |
| Daviey | jamespage: well, there has been a general backlog in SRU handling. | 14:27 |
| jamespage | Daviey, so this is not *normal*? | 14:27 |
| jamespage | Daviey, I just don't want to ask adam_g to recut the packages if we are going to be in the same place again in 3 weeks | 14:27 |
| Daviey | Whilst i got started with SRU processing, i wanted to try and avoid larger ones... However, once i am up to speed.. I can probably prioritise these based on priority of them blocking other work. | 14:28 |
| Daviey | I've been trying to help reduce the queue to free up others to review it. | 14:28 |
| Daviey | In addition, slangasek was chewing his fair share with MAAS SRU. | 14:28 |
| Daviey | So it's a backlog caused by the point release IMO. | 14:29 |
| Daviey | jamespage: I think it makes sense to check with jdstrand that there is nothing likely to trump an SRU shortly, before crafting an upload..Once that is signed off, we should try and get it accepted same day. | 14:30 |
| jamespage | Daviey, okies | 14:30 |
| jdstrand | I'm fine with that | 14:31 |
| Daviey | Yeah, really crappy situation on the essex uploads we had | 14:31 |
| jdstrand | note that most of these are embargoed first, so I can't necessarily be very specific | 14:31 |
| === chuck_ is now known as zul | ||
| jamespage | jdstrand, its would be enough to know that you have a security update in process for package XX | 14:31 |
| jdstrand | yeah | 14:32 |
| jdstrand | (though, not publicly) | 14:32 |
| jdstrand | we don't want to jeopardize our access to advanced notice | 14:32 |
| Daviey | for sure. | 14:33 |
| jamespage | jdstrand, agreed | 14:33 |
| jdstrand | anyway, just ask and I'll give you what you need to make a decision | 14:33 |
| rbasak | jamespage: build success for 2.4, with one extra patch needed: http://paste.ubuntu.com/5634220/ | 14:42 |
| rbasak | I'm a bit suspicious of what I saw around there though. I wouldn't be surprised to find a runtime problem. | 14:42 |
| rbasak | (alignment SIGBUS again) | 14:42 |
| jamespage | rbasak, lol - I don't think upstream have much interest in arm right now | 14:43 |
| rbasak | sorry, had an xpra/xchat hang | 14:45 |
| hallyn | vila: mardraum: it's the kernel | 14:45 |
| hallyn | i did the same tests in a precise container on raring host, same results | 14:46 |
| hallyn | with kvm hangs, without kvm works | 14:46 |
| Extreme | Hi, can anyone link me to a good article/Ubuntu help page that describes how to set up VPN on an Ubuntu server? | 14:56 |
| LargePrime | lol i was just doing that | 14:57 |
| LargePrime | https://help.ubuntu.com/11.10/serverguide/openvpn.html | 14:57 |
| holstein | !vpn | 14:57 |
| ubottu | For more information on vpn please refer to https://wiki.ubuntu.com/VPN | 14:57 |
| LargePrime | but i hit fail | 14:57 |
| LargePrime | because my provider had removed the kernal info | 14:58 |
| LargePrime | so i cannot /sbin/modprobe tun | 14:58 |
| holstein | LargePrime: are you using ubuntu? or something custom from your provider? | 14:59 |
| Extreme | LargePrime: aha | 14:59 |
| LargePrime | ubuntu, from my provider | 14:59 |
| jamespage | xnox, rbasak has the mongodb test suite running on ARM | 14:59 |
| LargePrime | I guess I could hack around it | 15:00 |
| LargePrime | but not that ubuntu savy | 15:00 |
| xnox | jamespage: noticed the upload. very nice =) | 15:00 |
| jamespage | xnox, yeah - its looking better on arm now | 15:01 |
| jamespage | rbasak, did you mean to pastebin the patch for 2.4.0? | 15:01 |
| rbasak | jamespage: I meant to show you the stacktraces in the test run :) | 15:02 |
| rbasak | I'll pastebin the patch for you | 15:02 |
| jamespage | rbasak, ta | 15:02 |
| jamespage | rbasak, I've been thinking about it and I'm not sure we should do this upgrade | 15:02 |
| jamespage | we really need to bump pymongo as well for compat | 15:02 |
| jamespage | feels a little risky | 15:02 |
| rbasak | Any other rdepends? | 15:03 |
| * jamespage looks | 15:04 | |
| rbasak | jamespage: http://paste.ubuntu.com/5634304/ - against your PPA 2.4 so re-adds the newest changes; arm-signed-char.patch is the new one | 15:04 |
| rbasak | I want to send this upstream if I can figure out where they want it | 15:05 |
| jamespage | kirkland, I see hockeypuck depends on mongodb - have you guys tested with 2.4? | 15:05 |
| rbasak | jamespage: I'm worried that staying with 2.2 will make the mongodb packing irrelevant. | 15:07 |
| rbasak | packaging | 15:07 |
| jamespage | rbasak, probably only for 9 months | 15:07 |
| jamespage | :-) | 15:07 |
| rbasak | Which will increase the number of people who find the packaging not useful and not come back | 15:07 |
| jamespage | rbasak, so the python driver release is due tomorrow - https://jira.mongodb.org/browse/PYTHON/fixforversion/11981 | 15:08 |
| jamespage | this all feels very .0 | 15:08 |
| LargePrime | is this not off topic chater? | 15:11 |
| LargePrime | holstein: | 15:13 |
| LargePrime | what are the risks of adding back the kernal libraries to the repository on my server | 15:14 |
| ttx | jamespage: keystone MP branch up | 15:17 |
| patdk-wk | largeprime, your oviously inside an openvz type thing, there isn't any hope unless you ask your provider to do something about it | 15:17 |
| holstein | LargePrime: yeah.. you want to make sure the provider supports what you are wanting to do | 15:18 |
| jamespage | ttx, great - thanks | 15:22 |
| Extreme | I'm getting this error when I do ./build-rsa http://pastie.org/6930703 | 15:23 |
| jamespage | zul: want me todo the CA backports for swift, glance, nova and keystoneclient? | 15:31 |
| jamespage | zul, also note keystone just cut a milestone-proposed | 15:31 |
| zul | jamespage: if you are so inclined ;) | 15:31 |
| jamespage | I've switch the lab | 15:31 |
| zul | jamespage: i saw | 15:31 |
| jamespage | zul, its like four commands | 15:31 |
| zul | jamespage: im still working on folsom | 15:32 |
| jamespage | zul, actually lets hold off glance until that fix lands | 15:32 |
| zul | jamespage: sure im just getting the red balls blue | 15:33 |
| jamespage | zul, great! | 15:33 |
| zul | jamespage: my ocd is kicking in | 15:33 |
| jamespage | lol | 15:33 |
| === Ursinha_ is now known as Ursinha | ||
| zul | jamespage: ill take care of keystone | 15:41 |
| jamespage | zul, lovely | 15:41 |
| === gema_ is now known as gema | ||
| jamespage | zul, any idea why python-keystoneclient builds in distro but not in sbuild locally? | 15:53 |
| jamespage | looks like its trying to use $HOME | 15:53 |
| RoyK | how can I see why ubuntu tells me the system needs a restart_ | 16:05 |
| RoyK | ? | 16:05 |
| jpds | RoyK: I think it's something in /var/lib/update-notifier/ . | 16:08 |
| jamespage | zul, http://people.canonical.com/~jamespage/ca-updates/ | 16:10 |
| eagles0513875 | hey guys im workign on enabling mod_status on apache and for some blessed reason when i use my domain it pulls up my site saying the page doesnt exist can anyone tell me what im doign wrong | 16:15 |
| eagles0513875 | i have everythign setup according to the documentation i have read on the apache site | 16:16 |
| histo | eagles0513875: have you tried in #apache ? | 16:18 |
| RoyK | jpds: hm.. can't find anything relevant there | 16:18 |
| histo | eagles0513875:nvm not on freenode | 16:19 |
| eagles0513875 | histo: there is httpd here for that and i have asked htere | 16:19 |
| ikonia | eagles0513875: what does the logs say | 16:23 |
| zul | jamespage: +1 | 16:28 |
| jamespage | zul, ta | 16:32 |
| kirkland | jamespage: not yet, but it's on our to-do list | 16:41 |
| kirkland | utlemming: okay, so I never quite got a bootable AMI | 16:42 |
| kirkland | utlemming: hoping you have some advice there | 16:42 |
| jamespage | kirkland, no rush - I was considering going for a FFe for raring but no-one has done any real testing yet | 16:42 |
| jamespage | so deferring until S | 16:42 |
| kirkland | jamespage: cool | 16:45 |
| rbasak | jamespage: so is the decision made? We're definitely sticking with 2.2 then, regardless? | 16:49 |
| jamespage | rbasak, I think so yes; I spoke with some of the ceilometer guys and they have done 0 testing as well | 16:50 |
| rbasak | OK | 16:50 |
| jamespage | rbasak, can you try one more 2.2.3 build on ARM for me? just add libssl-dev to the BD's | 16:51 |
| rbasak | Sure | 16:51 |
| jamespage | it will enable SSL support - I've not checked it on ARM | 16:51 |
| kirkland | jamespage: that would be nice :-) | 16:52 |
| kirkland | (ssl support) | 16:52 |
| jamespage | zul, yolanda: ceilometer branch needed a patch unfuzz - I've just refreshed and pushed | 16:52 |
| jamespage | kirkland, yeah | 16:52 |
| zul | jamespage: whoops...i was doing that too | 16:52 |
| jamespage | kirkland, it would - just this minor niggle of GPL + OpenSSL licensing to resolve | 16:52 |
| eagles0513875 | hey ikonia ended up figuring it out the issue was caused with mod rewrite rules intercepting the server-status address and redirecting it to wordpress page not found | 16:53 |
| eagles0513875 | on my site | 16:53 |
| yolanda | jamespage, great, i'm just testing the ceilometer charms, something is not working in the ceilometer-agent one | 16:53 |
| jamespage | zul, yolanda: I think that for a simple unfuzz no peer review should be required - I make the commit with [trivial] | 16:53 |
| jamespage | make/mark | 16:53 |
| zul | jamespage: ack | 16:53 |
| kirkland | jamespage: ah, yeah, that | 16:53 |
| jamespage | :-) | 16:53 |
| jamespage | yeah - that | 16:53 |
| * jamespage head in hands | 16:53 | |
| kirkland | jamespage: has 10gen not written an ssl exception into their license header? | 16:54 |
| kirkland | jamespage: we did that for ecryptfs | 16:55 |
| kirkland | jamespage: well, IBM did | 16:55 |
| jamespage | kirkland, no - I've asked for one but not yet happened | 16:55 |
| kirkland | jamespage: interesting | 16:55 |
| kirkland | jamespage: does it not compile against gnutls or libnss or libgcrypt? | 16:55 |
| zul | jamespage: did you kick off a new ceilometer as well? | 16:57 |
| jamespage | kirkland, no explicit support in the codebase other than openssl; that next path to investigate | 16:57 |
| jamespage | zul, yep | 16:57 |
| zul | jamespage: coolio | 16:58 |
| kirkland | jamespage: yeah, then in that case and in my non-lawyer opinion, compilation against ssl is implied as allowed | 16:58 |
| kirkland | jamespage: if the (GPLed) code is written specifically against openssl.h, then, um yeah, I reckon the authors intended on allowing you to actually run that code | 16:59 |
| kirkland | jamespage: but that's just me :-) | 16:59 |
| jamespage | kirkland, lol - indeed | 16:59 |
| kirkland | jamespage: ie, when upstream distributed the code in that manner | 17:00 |
| kirkland | jamespage: on the other hand, if you, jamespage, wrote a big fat patch to mongodb that added ssl support inline in the code, but required compiling against ssl | 17:00 |
| kirkland | jamespage: for that, I'd say you'd need to acquire an exception from the copyright holders | 17:00 |
| kirkland | jamespage: but if they're distributing the code as such | 17:00 |
| jamespage | kirkland, I tend to agree - but this appears to be a legal grey area | 17:01 |
| kirkland | jamespage: and it doesn't actually compile against any of the gnu tls libraries...then what option do you really have? | 17:01 |
| * kirkland just watched Lincoln... most of what Lincoln did was legal grey area -- but it worked itself out :-) | 17:01 | |
| kirkland | jamespage: do you have any expertise creating AMIs? | 17:02 |
| kirkland | jamespage: I'm looking for smoser or utlemming to help me finish something I started yesterday, but I'm blocked on now | 17:02 |
| jamespage | kirkland, zip I'm afraid - that's smoser/utlemming territory | 17:02 |
| kirkland | jamespage: dang | 17:02 |
| jamespage | sorry | 17:02 |
| smoser | kirkland, i'm here. whats up? | 17:02 |
| * utlemming is here now | 17:03 | |
| utlemming | kirkland: what's up? | 17:03 |
| kirkland | smoser: oh, hey | 17:03 |
| kirkland | utlemming: okay, so my AMI is currently not bootable | 17:03 |
| utlemming | kirkland: did you dd from /dev/nbd0 to the volume? | 17:04 |
| kirkland | utlemming: yep | 17:04 |
| * utlemming tries to do this | 17:04 | |
| kirkland | utlemming: did I need to do any resize2fs or tune2fs or fsck on that? | 17:04 |
| utlemming | kirkland: nope | 17:04 |
| kirkland | utlemming: and what aki- should I be using? | 17:04 |
| kirkland | utlemming: I've tried Kernel ID: aki-b4aa75dd | 17:05 |
| kirkland | utlemming: I *think* maybe smoser helped me with something like this before, and pointed me to that aki | 17:05 |
| smoser | console output ? | 17:05 |
| utlemming | kirkland: that is the wrong one...you need the hd0 variant | 17:05 |
| kirkland | smoser: http://paste.ubuntu.com/5632736/ | 17:06 |
| kirkland | utlemming: k -- is that the -825 one? | 17:06 |
| utlemming | kirkland: you have the wrong AKI for sure. | 17:07 |
| kirkland | utlemming: cool-- which one should I be using? | 17:07 |
| kirkland | utlemming: and the root device -- is that /dev/sda or /dev/sda1? | 17:07 |
| utlemming | kirkland: aki-88aa75e1 and whatever you used for your device mapping when you registered it | 17:08 |
| utlemming | kirkland: for the device mapping, it doesn't matter | 17:08 |
| kirkland | utlemming: okay, created and launching | 17:09 |
| * kirkland whistles the jeopardy tune | 17:09 | |
| kirkland | utlemming: smoser: okay, so I'm curious on both of your takes on this... what do you reckon is the best way to flag that I want to run an instance with an overlayroot (either tmpfs or encrypted), WITHOUT first launching/editing/rebooting | 17:11 |
| kirkland | utlemming: smoser: right now, I'm creating an AMI that has a one-line change in /etc/overlayroot.conf | 17:11 |
| kirkland | utlemming: smoser: I can add support into the initramfs hooks for overlayroot that sets up networking and fetches the metadata | 17:12 |
| utlemming | smoser: correct me if I am wrong, but does cloud-init look for OVERLAY_ROOT or something like that and just do it? | 17:12 |
| === _gugaua_ is now known as gugaua | ||
| kirkland | utlemming: that's too late | 17:12 |
| kirkland | utlemming: it has to be setup in initramfs, not userspace | 17:12 |
| kirkland | utlemming: cloud-init runs in userspace | 17:12 |
| kirkland | utlemming: by that point, we've already written stuff to the root disk | 17:13 |
| smoser | kirkland, its not acceptabale to change the default amis. so your one off with the one line change is not unreasonable. | 17:13 |
| kirkland | utlemming: still not booting, http://paste.ubuntu.com/5634659/ | 17:14 |
| smoser | the other option is to use the original image bit for bit and boot with a --block-device-mapping | 17:14 |
| smoser | and that disk that is attached can probably manage to do what you want. | 17:14 |
| kirkland | cmagina: how so? | 17:14 |
| kirkland | smoser: ^ | 17:14 |
| kirkland | cmagina: sorry... and howdy, btw :-) | 17:15 |
| cmagina | kirkland: np :) | 17:15 |
| smoser | well, the overlayroot config disk executes (it is sourced) in the initramfs. | 17:15 |
| smoser | so you can actually do anything you want. | 17:15 |
| kirkland | utlemming: i did dd from /dev/nb0, and not from the .img -- that's correct, right? | 17:15 |
| utlemming | kirkland: yup. I'm going through the exercise now. Give me a few minutes... | 17:16 |
| kirkland | utlemming: thanks! | 17:16 |
| kirkland | utlemming: I can share this ami with you too, if it helps | 17:16 |
| kirkland | utlemming: the *only* change is that /etc/overlayroot.conf has: "crypt:dev=/dev/xvdb" | 17:17 |
| kirkland | utlemming: make sure you add one ephemeral disk at xvdb when you register the ami | 17:17 |
| sbeattie | plars: sorry, I just noticed your ping. Odd that af_bluetooth is being reported as unsupported | 17:19 |
| plars | sbeattie: any chance it's just because of age? This is oneiric we're talking about.. | 17:21 |
| plars | and on panda | 17:21 |
| sbeattie | well oneiric/omap4 ; it's possible I guess, though I'm not sure why that would show up just now. | 17:21 |
| smoser | kirkland, does it make sense to you how you'd do that ? | 17:22 |
| plars | sbeattie: it'll take some time, but I could go back and reinstall and see if it failed before the update.. I have no idea what it looked like before | 17:22 |
| sbeattie | plars: any chance you can run [path to qa-r-t]/scripts/apparmor/test-net.py --domain bluetooth | 17:22 |
| plars | sbeattie: as soon as it finishes what it's doing now, I can | 17:22 |
| plars | board is tied up at the moment | 17:22 |
| sbeattie | plars: that'd be great. if that fails, then it's not an apparmor issue, but an issue with that kernel's bluetooth support. | 17:25 |
| utlemming | kirkland: launching... | 17:26 |
| kirkland | smoser: without modifying Ubuntu's stock, default overlayroot.conf? | 17:26 |
| kirkland | smoser: I know that if I modify overlayroot.conf myself in my own AMI, then yeah, sure, I can do whatever I want | 17:27 |
| kirkland | smoser: I'm trying though to find a way where I can continue to use the stock Ubuntu AMI, and instead just launch with some additional option or configuration or something | 17:27 |
| kirkland | smoser: heck, even just configuring a "security group" would work for me -- though that would still involve reading the security-group from metadata | 17:28 |
| smoser | kirkland, yes. you can do that. | 17:29 |
| smoser | you just have to use your imagination a bit :) | 17:29 |
| plars | sbeattie: I don't remember for certain, but that was my suspicion - that we didn't get bluetooth working properly there until later | 17:31 |
| smoser | kirkland, bzr+ssh://bazaar.launchpad.net/~smoser/+junk/ovroot-snapshot-create/ is the branch i pointed you at once showing how i could create a partitioned disk. | 17:32 |
| smoser | inside that (lines 24) is a '_find_dev' that goes looking for a disk that it should tell the initramfs is the overlay root disk. | 17:33 |
| smoser | you can do the same thing, but instead of looking for disks, look at the metadata service. | 17:33 |
| sbeattie | plars: looking at http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-oneiric.git;a=blob;f=debian.master/config/config.common.ubuntu vs http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-oneiric.git;a=blob;f=debian.ti-omap4/config/config.common.ubuntu;h=f0724990db3a6b4c44ca0fa8ddc49462a1b5ff96;hb=7bf8c235abf0058e8a7a4d6fc927656c74cfbf93 I see differences in some of the CONFIG_BT sub options. | 17:33 |
| smoser | the "real root" is mounted read-only at this point. so you can do just about anything actually. | 17:33 |
| smoser | (including mounting rw and chroot to it!) | 17:33 |
| smoser | you can use it to bring up networking, look at user-data and act appropriately. | 17:34 |
| kirkland | smoser: cool, thanks | 17:38 |
| smoser | kirkland, you'll probably want to put a flag on the root disk in some sense so you dont do that every boot if it wasn't desired. (or you more confidently do it if it *was*) | 17:39 |
| kirkland | utlemming: did yours launch successfully? | 17:42 |
| kirkland | smoser: right -- so currently, I'm just looking to do this at every reboot (for my current project) | 17:42 |
| kirkland | smoser: but eventually, yes | 17:43 |
| smoser | thats fine. you might as well mark somewhere that you've decided "do it" | 17:43 |
| kirkland | smoser: also, I'd like to in the next cycle MIR molly-guard and have the overlayroot binary package recommend it, and install a mollyguard script | 17:43 |
| utlemming | mkirkland: no...looking why | 17:44 |
| kirkland | utlemming: awesome, thanks | 17:46 |
| kirkland | smoser: I'd like overlayroot to install something like http://paste.ubuntu.com/5634755/ in /etc/molly-guard/run.d/02-overlayroot | 17:47 |
| kirkland | smoser: basically, to prevent inadvertent reboots that lose or destroy data (in tmpfs or encrypted filesystems) | 17:47 |
| kirkland | smoser: are you familiar with molly-guard? | 17:47 |
| xerxas | Hi all | 17:48 |
| xerxas | can I preseed my install with protocol mirror and http://mirrors.ubuntu.com/mirrors.txt , for the apt configuration ? | 17:48 |
| yolanda | jamespage, pushed a working version - finally - of ceilometer agent charm | 17:49 |
| smoser | answer=$(head -n1) is done without a fork like: | 17:53 |
| smoser | read answer | 17:53 |
| smoser | i dont have a conceptual issue with that. as long as the implementation doesn't require hijacking /sbin/reboot | 17:55 |
| jamespage | yolanda, great! | 17:56 |
| jamespage | zul, adam_g: libvirt 1.0.2 in the trunk testing PPA - working OK for me | 17:57 |
| jamespage | I can boot instances, live migrate them etc... | 17:57 |
| zul | jamespage: push it | 17:57 |
| === Ursinha is now known as Ursinha-afk | ||
| zul | jamespage: just a note from that libvirt release python-libvirt was broken with a patch that was backported ;) | 17:59 |
| jamespage | zul, live migration dropped 3 packets - that sound reasonable to me | 18:01 |
| zul | jamespage: tolearable | 18:01 |
| jamespage | zul, from with libvirt release? confused | 18:01 |
| zul | 1.0.2 | 18:01 |
| zul | so thats why i was a bit leary about it its fixed in 1.0.3 but we backported the patch | 18:02 |
| zul | just an fyi | 18:03 |
| === Ursinha-afk is now known as Ursinha | ||
| rbasak | jamespage: mongodb 2.2.3 build with libssl-dev succeeded on armhf. Build log: mongodb_2.2.3-0ubuntu4~basak1_armhf.build - looks like it picked it up OK. | 18:37 |
| rbasak | EOD | 18:37 |
| === airtonix_ is now known as airtonix | ||
| xerxas | ajmitch: I saw you wrote several time about using mirror uri scheme in sources.list, can I preseed this for installation ? | 19:09 |
| kantlivelong | hey all.. im trying to install 12.04.2 and the installer doesnt see any disks.. but when going to the installer shell i see /dev/sda /dev/sdb and also see in dmesg.. any idea whats going on? | 19:12 |
| lamawithonel | Is it possible to encrypt a single partition with preseeding? I'd like to encrypt swap, but nothing else. | 19:36 |
| scalability-junk | hey I have a problem when I try to run lvdisplay I only get "read_urandom: /dev/urandom: open failed: No such file or directory" returned | 19:41 |
| scalability-junk | additionally all mv vms under /dev/vg0/* are gone ... (but they are still working( | 19:41 |
| scalability-junk | any idea please | 19:41 |
| scalability-junk | no idea at all? | 19:51 |
| scalability-junk | when trying to run virt-install with /dev/vg0/something I get an apparmor error: internal error cannot load AppArmor profile 'libvirt-XXX' | 20:51 |
| scalability-junk | any idea? | 20:51 |
| fabiofranco | how to add an existing user to an existing group? | 20:56 |
| sw | fabiofranco: $ sudo useradd -G <group> <user> | 20:57 |
| fabiofranco | sw the command returns an error msg - useradd: user 'thiago' already exists | 20:58 |
| sw | fabiofranco: did the user get added to the group though? ($ groups <user>) | 20:59 |
| fabiofranco | sw sadly no | 21:00 |
| sw | fabiofranco: erm, -a -G maybe, sorry been a while since I've done it, if all else fails see $ man useradd :b | 21:00 |
| Underbyte | how do you convert /etc/init.d scripts to upstart jobs? | 21:01 |
| fabiofranco | sw: the -a parameter doesnt exist... i'll have a look at the manual, thanks | 21:02 |
| sarnold | scalability-junk: some of the protections for virtual machines / lxc containers via libvirt is handled via apparmor profiles; see /etc/apparmor.d/{libvirt,lxc,lxc-containers} ... (not great information there, but .. information.) | 21:04 |
| Trudko | guys i run netstat -tulpn | grep :80 and i got tcp 0 0 0.0.0.0:80 0.0.0.0:* LISTEN - how to stop this service ? | 21:15 |
| sarnold | Trudko: re-run as root and find the process | 21:16 |
| Trudko | ok thx | 21:17 |
| === Ursinha is now known as Ursinha-afk | ||
| scalability-junk | sarnold: yeah but it doesn't find the profile for some libvirt-uuid and then fails | 22:17 |
| scalability-junk | using aa-complain didn't help | 22:17 |
| scalability-junk | and disabling apparmor seems like bad... | 22:18 |
| sarnold | scalability-junk: do you have DENIED messages in your dmesg or /var/log/audit/audit.log ? | 22:19 |
| scalability-junk | nothing in the log | 22:21 |
| scalability-junk | nope no DENIED | 22:22 |
| scalability-junk | sarnold: retried and got a denied [178696.696032] init: /tmp/tmpYLM9OL/etc/init: Configuration directory deleted | 22:24 |
| scalability-junk | [179194.781900] type=1400 audit(1363904669.177:199): apparmor="DENIED" operation="open" parent=18775 profile="/usr/lib/libvirt/virt-aa-helper" name="/dev/dm-29" pid=18174 comm="virt-aa-helper" requested_mask="r" denied_mask="r" fsuid=0 ouid=0 | 22:24 |
| sarnold | scalability-junk: ha, I've got an explicit deny for those files in my version of that profile: deny /dev/dm-* r, | 22:27 |
| scalability-junk | sarnold: any workaround? haven't worked with apparmor at all | 22:28 |
| sarnold | hallyn: scalability-junk here has had some problems with virt-install, it's beyond me :) care to look at it? ^^^ | 22:28 |
| sarnold | scalability-junk: you could put that profile into complain mode, but you'd lose the confinment. someone put that 'deny' rule in place for a good reason... | 22:29 |
| sarnold | scalability-junk: you could add a rule to grant read access to that device (and, maybe, soon it'd need write as well....) -- if you're fine with containers being able to modify that device. | 22:29 |
| scalability-junk | sarnold: yeah but I don't see any issues with my virt-install command, or isn't is allowed to use /dev/vg0/something as image for example | 22:30 |
| * scalability-junk is a bit frustrated tried vmbuilder first -> failed, then it worked after some patches then the vm got stuck in booting from disk and then tried virt-install and apparmor doesn't like me :D | 22:31 | |
| scalability-junk | srsly it complained about the directory of the .iso file not the block device yeah I love errors :D | 22:36 |
| scalability-junk | anyone using virt-install? I'm stuck how to configure sshd to be present after startup? | 22:50 |
| scalability-junk | virt-install -r 1024 --accelerate -n demo.oc -f /dev/vg0/demo.oc --graphics vnc --network bridge=virbr1 --cdrom /var/ubuntu-12.04.2-server-amd64.iso | 22:50 |
| scalability-junk | was my try, but first graphics vnc needs virt-viewer o0 what was used by the vmbuilder? | 22:51 |
| tgm4883 | Mythbuntu site broke | 22:55 |
| tgm4883 | Daviey needs to fix DNS records | 22:55 |
| tgm4883 | Then I will be glad | 22:55 |
| Trudko | Guys i need to copy something to local/bin but my user does not have persmissions. ls -l /bin | grep bin i get lrwxrwxrwx 1 root root what to do to have rights to write? | 23:09 |
| Trudko | i found that i should do usermod -G www-data user which should put user to www data group does have to data group exists already? is user param user litterallly or it is name of user | 23:10 |
| shauno | that's a fun handful. "ls -l /bin | grep bin" is giving you a red herring; it's returning a result within /bin that happens to be symlinked somewhere (the leading 'l' on the permissions) | 23:14 |
| shauno | ls -ld /usr/local/bin will show you what the perms are there, which seems a lot more relevant; default will be read-all, write-root | 23:15 |
| Trudko | shauno i see so what should i do so user X can have write permissions too? | 23:15 |
| shauno | you very rarely would; usually that's left alone and sudo is used to escalate where needed | 23:17 |
| Trudko | shauno where I am copying using winscp | 23:17 |
| scalability-junk | YESSSSSSSS!!!!! finally it worked | 23:18 |
| scalability-junk | now I know why some people like the cloud :D virtual machines do just work :D | 23:18 |
| shauno | personally, I'd scp the files into the user's home folder, and then move them into place with sudo afterwards. it just works out a whole lot easier than fighting the design, which usually causes loose threads to unravel | 23:18 |
| sarnold | scalability-junk: woo :) | 23:19 |
| histo | scalability-junk: I wish I had a cpu that supported vmx or svm myself | 23:40 |
| scalability-junk | histo: get a server ;) | 23:43 |
| histo | scalability-junk: I will just want a tablet right now to stuff ubuntu on. | 23:59 |
| scalability-junk | histo: servers are so much better :) | 23:59 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!