[00:14] <igoryonya> jhobbs, http://paste.ubuntu.com/8859114/
[09:13] <lordievader> Good morning.
[11:29] <ayr-ton> Some knows if is possible to text LXD without nova-compute-flex, like without install all the openstack?
[12:34] <_ruben> argh ... bitten by the no root over ssh by default again .. now i gotta walk down to the server room again .. hooray for pxe booting new servers :p
[12:38] <lordievader> _ruben: Hehe, I allways have that problem with debootstrapping. Forgetting to set a password for root or making new users that actually have a password...
[12:42] <_ruben> lordievader: my preseed file sets a root pw, which worked fine upto 12.04, just doesn't cut it anymore on 14.04 ;)
[12:42] <_ruben> Time to upgrade our install process I guess (automate the post-install tasks further)
[12:56] <RoyK> anyone that knows a magic trick on how to get performance statistics per database with mysql?
[12:56]  * RoyK is really trying to think of something good about mysql
[12:59] <MacroMan> RoyK, If you had a unique user for each db, then you could run a performance script to check just one but only using those credentials
[13:00] <MacroMan> I use http://mysqltuner.com/ which seems to work quite well.
[13:00] <RoyK> MacroMan: ah - what sort of script would that be?
[13:00] <MacroMan> It's a perl script that you run and it gives some useful info. It's not the be all and end all, but it's certainly a good indication of performance
[13:02] <RoyK> MacroMan: thanks - lemme take a look
[13:02] <RoyK> http://karlsbakk.net/fun/USA/american_people_2014.jpg :P
[13:04] <RoyK> MacroMan: the solution to this is really s/MySQL/PostgreSQL/g
[13:17] <RoyK> MacroMan: thanks for the advice, but really, this doesn't show very useful performance counters :P
[13:17]  * RoyK forces developers to move to postgresql :P
[14:09] <MacroMan> exit
[18:32] <blkperl> anyone know how to get past Hash Sum mismatch errors on apt-get update when rm'ing /var/lib/apt/lists/* doesn't work?
[18:33] <sarnold> blkperl: check dmesg to make sure you don't have dying hardware
[18:34] <blkperl> sarnold: its a fresh vagrant image
[18:34] <sarnold> blkperl: aha :)
[18:35] <blkperl> I've added a Third party mirror and am getting the hash mismatch errors
[18:37] <blkperl> is their a way to verify if the third party repo is actually broken?
[18:38] <sarnold> blkperl: it gets difficult; download the files by hand, verify the hashes by hand. it's not fun.
[18:43] <blkperl> ugh after apt-get upgrade, apt-get update works...
[18:44] <sarnold> that can happen too, if the mirror was in the middle of a sync...
[18:46] <blkperl> I don't think thats the issue, I think some combination of apt-get upgrade or adding the gpg key first and then apt-get update, "fixes" the hash mismatch error for this particular case
[18:47] <blkperl> maybe TravisCI is needs to upgrade their precise image
[18:47] <blkperl> since their apt is out of date
[20:16] <hallyn> jdstrand: hey, are you around to sanity check a libvirt-apparmor debdiff?
[20:16] <hallyn> jdstrand: http://paste.ubuntu.com/8872448/  (i'm goig to test a bit more and then probably push to v and sru to u and t)
[20:20]  * jdstrand looks
[20:23] <jdstrand> hallyn: I think you also want something for 'network netlink,' for 14.10 only
[20:24] <jdstrand> other than that, without actually running it, it looks good
[20:25] <jdstrand> hallyn: also, even though libvirt-qemu and libvirt-lxc don't have unix rules, it is probably good to run the sed on them too to future proof us in case we add something there
[20:26] <jdstrand> (that goes for the netlink rule too)
[20:36] <hallyn> jdstrand: gah.  so I should pull 'network netlink' out for anything earlier than 14.10?  but keep it in for 14.10 and 15.04?
[20:37] <hallyn> (so same as the unix rules?)
[20:38] <jdstrand> hallyn: exactly
[20:38] <hallyn> jdstrand: cool, thanks
[20:39] <jdstrand> hallyn: so, I would phrase that as '14.10 and later' :)
[20:39] <jdstrand> since I seem to have my pedantic hat on
[20:40] <hallyn> jdstrand: but the rules are stripping it out for 14.04 and earlier
[20:40] <hallyn> theyr'e doing nothing for 14.10 and later
[20:40] <hallyn> hopefully i've got this (*$&%$ right :)
[20:40] <hallyn> took me days to get something that seems to work
[20:41] <jdstrand> hallyn: I just mean you said '14.10 and 15.04', but that should be '14.10 and later'
[20:41] <hallyn> oh, right :)
[20:42] <jdstrand> hallyn: the code is doing that. I was just adding confusion to the conversation
[20:42] <hallyn> ok updated debdiff i'm going to build real quick - http://paste.ubuntu.com/8872671/
[20:42] <jdstrand> hallyn: what you came up with seems very clean
[20:43] <hallyn> excellent, thx.  hopefully it all builds fine and i'll push to v tonight.  (dunno about the SRUs - i'm out quite a bit next week)
[20:43] <hallyn> anyway - gnight
[20:43] <jdstrand> hallyn: certainly you have people for that :)
[20:43] <jdstrand> hallyn: g'night
[21:52] <hallyn> jdstrand: gah.  so 9 days ago libvirt built+tested ok.  today, libvirt fails the qemuxml2argvtest testcase.  i'm not quite sure whether i should just push it to vivid and see if the buildds do a better job, or if this is more likely to be due to a recent package change
[21:55] <hallyn> jdstrand: d'oh, wasn't there a libxml security update.  wonder whether that did it
[21:56] <sarnold> hallyn: http://www.ubuntu.com/usn/usn-2389-1/
[21:56] <sarnold> hallyn: there were also kvm changes in kernel security updates: http://www.ubuntu.com/usn/usn-2395-1/
[21:57] <hallyn> sarnold: yeah but i don't *think* qemu-* is involved
[21:57] <sarnold> ahh
[21:57] <hallyn> more like "AAAAAARGH"
[21:58] <hallyn> as i recall the error was something like it expected '//V' but got 'V'
[21:59] <hallyn> sarnold: sigh.  i wanted to go t osleep, but i guess i'll try a build iwth the old libxml
[22:00] <sarnold> hallyn: go to sleep :)
[22:01] <hallyn> well if libxml is broken i wont *fix* it now :)  i'll just push the libvirt and post a bug
[22:03] <sarnold> hallyn: well, okay, if knowing if libxml is to blame will help you get some sleep .. :)
[22:09] <hallyn> when you say it out loud it doesn't sound likely does it
[22:10] <sarnold> I think I've heard complaints about libxml updates breaking other package builds before, apparently people come to rely upon the bugs :)
[22:20] <hallyn> sarnold: feh.  confirmed.  after a dpkg -i libxml2_2.9.1+dfsg1-3ubuntu4.4_amd64.deb  libvirt builds
[22:21] <sarnold> hallyn: sigh
[22:25] <hallyn> sarnold: i don' tknow why i'm telling you this, as i don't hold you responsible :)  but i opened bug 1390637
[22:25] <hallyn> i'll be working 2 or 3 days next week, can take a look if noone else has then (T or W)
[22:25]  * hallyn out
[22:26] <sarnold> hallyn: have a good weekend :)