=== wedgwood_away is now known as wedgwood | ||
=== wedgwood is now known as wedgwood_away | ||
=== airtonix_ is now known as airtonix | ||
=== arrrghhh is now known as arrrghhhAWAY | ||
=== arrrghhhAWAY is now known as arrrghhh | ||
gustav | Gmail doesn't accept e-mails from my exim4 host. Wat do kil self? | 07:12 |
---|---|---|
gustav | Oh, it deblacklisted me now. That's awkward. | 07:14 |
=== smb` is now known as smb | ||
RoyK | hm... on precise, where can I find why a reboot may be required? /var/run/reboot-required doesn't give any hits | 07:44 |
=== Sargun_ is now known as Sargun | ||
zokko | hello folks | 10:26 |
=== Nigel_ is now known as G | ||
student-2013 | hello | 12:54 |
student-2013 | Can i run an application on amazon EC2? | 12:55 |
student-2013 | Why no answer? | 12:57 |
premera | student-2013: what appllication ? | 13:11 |
greppy | premera: they left the channel. | 13:12 |
premera | k | 13:12 |
pmatulis | using resolvconf on 12.04, say i make a change to interfaces file by adding dns-nameservers line. is restarting networking service the only way to enable that change (update the resolv.conf symlink)? | 15:18 |
SpamapS | pmatulis: I think you can just re-run 'ifup ethX' | 15:27 |
adam_g_ | zul: http://people.canonical.com/~agandelman/ca/grizzly/libvirt/ | 15:29 |
DasIch | I'm running the latest ubuntu and would like to serve my website using spdy with nginx. I'm already using nginx and ssl but following the nginx documentation I get '[...]invalid parameter "spdy"[...]' in the error.log | 15:48 |
DasIch | Does the nginx version in the repositories not support spdy? | 15:49 |
DasIch | If so is there are reasonable and secure way to get a more recent nginx version without compiling nginx myself? | 15:49 |
jacobw | DasIch: see if someone else has packaged it or use the vendor's binaries | 16:00 |
Nafallo | DasIch: launchpad has a ppa for latest stable. | 16:01 |
DasIch | Nafallo: if I understand https://launchpad.net/~nginx/+archive/stable correctly the ppa was last updated in february, there have been several nginx releases since then | 16:12 |
=== wedgwood_away is now known as wedgwood | ||
Nafallo | DasIch: looks like you're correct. might be worth mailing the maintainer and ask what's up? | 16:32 |
hallyn_ | jjohansen: hi, have apparmor controls on setns() been considered? (sorry i haven't read your email yet, so sorry if it's answered in there) | 17:00 |
jjohansen | hallyn_: yes | 17:01 |
DasIch | Nafallo: yeah, I will try that | 17:01 |
hallyn_ | jjohansen: cool, considered as in it's on the roadmap? | 17:01 |
jjohansen | hallyn_: we have a work item for it | 17:02 |
hallyn_ | cool, thanks. | 17:02 |
jjohansen | I'm not sure which bp its in, probably july or august | 17:02 |
hallyn_ | awesome | 17:03 |
=== wedgwood is now known as wedgwood_away | ||
hallyn_ | stgraber: for cloning ubuntu-cloud images, smoser would like to have a lxc.clonehook, which could point to a script to be executed (with all arguments passed to lxc-clone after '--') at clone time to update the rootfs. Do you see any issues with that/ | 18:17 |
stgraber | hallyn_: I'm fine with lxc.hooks.clone though the arguments part will make it a bit different from the other hooks, do we actually have variable arguments for lxc-clone or could we get away with a few additional env variables? | 18:19 |
smoser | lxc-clone -o orig -n new -- --my-clonehook-arg1=value ... | 18:20 |
smoser | then "my-clonehook" gets passed: | 18:20 |
smoser | /var/lib/lxc/orig /var/lib/lxc/new --my-clonehook-arg=1value | 18:20 |
hallyn_ | stgraber: yeah it would be a bit different, and i'm not entirely comfortable with it, but it's doable | 18:21 |
hallyn_ | confile would just store it as one long char*, and lxcapi_clone() would parse it | 18:21 |
stgraber | hallyn_: I think I'd rather add a LXC_CLONE_ARGUMENTS env variable that we pass to the hook. That way we keep the hook parameters consistent. | 18:27 |
stgraber | hallyn_: so basically have the clone() call in the API take the extra parameters (similar to create and start) and just push those into LXC_CLONE_ARGUMENTS before calling the hook (if there's one set) | 18:28 |
stgraber | (remember that I just got back from a week of doing client stuff, I haven't yet had time to catchup on the various patches, so the above is solely based on what I read on IRC ;)) | 18:29 |
hallyn_ | feels hacky though. Perhaps these should be kept semantically separate from container startup hooks. So we have container create hooks (which right now is just the template) and clone hooks, and heck maybe destroy hooks | 18:31 |
hallyn_ | anyway, it's something to think about and hopefully discuss at vUDS next week | 18:31 |
hallyn_ | stgraber: there's a way around this, at any case - using a pre-start hook which checks whether /var/lib/lxc/$c/NOTNEW exists, if so exits, if not creates it and does the update. But that doesn't have the arguments, so it's not quite sufficient | 18:33 |
=== Ursinha-afk is now known as Ursinha | ||
stgraber | hallyn_: nah, I think adding more hooks makes sense, it's just that we need to expose the arguments to the hooks and we should do that in a generic way for all hooks (as in, I should be able to get the arguments of lxc-start ... -- ARG in some environment variable in our current hooks) | 18:35 |
stgraber | that way we could also have pre-start/mount/... hooks that vary based on init arguments (for example) | 18:36 |
stgraber | so it sounds like a thing we can generically apply to all hooks with no special casing for the clone usecase | 18:36 |
oDiafanos | Hello! I run ubuntu server x86 12.04.1 and mysql refuses to start. On reboot it hanged so i tried 2 different backups. Mysql refuse to start.error seems to be dpkg: dependency problems prevent configuration of mysql-server: | 18:38 |
oDiafanos | <oDiafanos> mysql-server depends on mysql-server-5.5; however: | 18:38 |
oDiafanos | <oDiafanos> Package mysql-server-5.5 is not configured yet. | 18:38 |
hallyn_ | to be clear i misspoke before, the lxc.clonehook entry in /var/lib/lxc/$c/config would not list the extra arguemtns, they'd be listed on the lxc-clone command line. | 18:38 |
hallyn_ | but yeah | 18:38 |
sarnold | oDiafanos: try sudo dpkg-reconfigure -a | 18:39 |
stgraber | hallyn_: right, that's how I understood it | 18:39 |
hallyn_ | stgraber: btw i haven't yet pushed Dwight's last 6-patch set to staging. | 18:39 |
stgraber | hallyn_: I did | 18:39 |
hallyn_ | thx | 18:39 |
hallyn_ | can't believe i did some of those... <blush> | 18:39 |
stgraber | hallyn_: I'm a bit lost with the lxc-devel backlog, so if there's anything that I should be reviewing, let me know, I completely lost track... | 18:40 |
hallyn_ | stgraber: i think i pushed everything up to those 6 | 18:40 |
stgraber | ok good. I'll probably send an e-mail post-vUDS asking for anyone who submitted patches that didn't get applied to speak up, there are always a few we miss... | 18:41 |
hallyn_ | yeah | 18:43 |
oDiafanos | sarnold, thanks but no..same error | 18:44 |
sarnold | oDiafanos: try sudo dpkg-reconfigure --force mysql-server mysql-server-5.5 ? | 18:44 |
oDiafanos | 1 sec I'm on ssh so i trying instantly | 18:45 |
oDiafanos | nope..I even tried to del and install again the pkg | 18:46 |
oDiafanos | any other suggestions? | 18:51 |
=== mikehale_ is now known as mikehale | ||
pmatulis | does kexec/kdump/linux-crashdump work on 12.04? | 20:33 |
hallyn_ | stgraber: so i've updated lxc-snap, it now snapshots under $lxcpath/snapshots, but as a result it supports lvm, btrfs and zfs, not just overlayfs. (I figured I"d test the limits before considering putting it in the api) | 22:03 |
stgraber | neat | 22:03 |
bcessa | hi there, any recommendations for a CLI tool to manage a custom CA, yes I know OpenSSL is just enough but I'm just curious if there's already a tinyCA0-like CLI alternative | 22:24 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!