| *** acormier has joined #openstack-chef | 01:08 | |
| *** acormier has quit IRC | 01:10 | |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-common master: Corrects rabbitmq user, yum cache cleaning https://review.openstack.org/454906 | 01:39 |
|---|---|---|
| openstackgerrit | Jens Rosenboom proposed openstack/cookbook-openstack-compute master: Add default hosts discovery interval https://review.openstack.org/455169 | 08:57 |
| *** acormier has joined #openstack-chef | 12:12 | |
| *** acormier has quit IRC | 12:21 | |
| sc` | frickler: did my comment re: libvirt make sense? if only it were an arbitrary change | 12:52 |
| sc` | i started picking the vm apart by hand to get to the bottom of it, and the root cause for convergence to fail on my dev box was the lack of a group | 12:53 |
| sc` | said dev box does nothing fancy. it runs centos 7 with the 4.10 kernel | 12:54 |
| *** jmb_ has quit IRC | 12:54 | |
| frickler | sc`: hmm, I'm confused, once you talk about Ubuntu and now Centos. If it doesn't work on Ubuntu, I'd rather file an upstream bug and get it fixed, people in #ubuntu-server have been pretty good at that recently. Better than masking it in our cookbook until it explodes in a different fashion | 13:07 |
| frickler | and my local deployment of Ocata on a real HW cluster seems to work fine as it stands | 13:08 |
| sc` | sorry, left out a key detail | 13:08 |
| sc` | the *host* machine is centos, but i'm running it from dk using the provisioning cookbook to pop centos and ubuntu vms | 13:08 |
| sc` | the ubuntu vms fail to work as advertised without removing that guard | 13:09 |
| sc` | it's barely awake and i'm early. not enough caffeine | 13:09 |
| sc` | it seems the requirement of a group was made a requirement (haven't dug through changelogs yet) in recent releases | 13:10 |
| sc` | another detail i omitted due to the time of day: i added the group manually and libvirt started | 13:11 |
| frickler | the group name has changed for Ocata, from libvirtd to libvirt (or the other way round?) | 13:11 |
| frickler | I've looked into the postinst code for libvirt packages and they seem to do the right thing for me | 13:12 |
| sc` | it seems to be 'libvirt' | 13:12 |
| frickler | did you have my patch for using the correct group? | 13:12 |
| sc` | i'll have to spin both platforms to clarify that it doesn't need a case statement | 13:13 |
| sc` | but that's no problem | 13:13 |
| sc` | i did not see a patch. is it in gerrit? | 13:14 |
| sc` | i need to run to take care of someting. will be back in a bit | 13:15 |
| frickler | sc`: it was merged already, but it may have caused issues if you ran your tests without it https://review.openstack.org/443694 | 13:16 |
| *** chlong has joined #openstack-chef | 13:31 | |
| *** openstackgerrit has quit IRC | 13:33 | |
| *** jmb_ has joined #openstack-chef | 13:41 | |
| sc` | oh, that one! | 14:31 |
| sc` | i checked my working copy and it had that change merged. i tend to rebase against master before i submit a review against it | 14:32 |
| sc` | so it was definitely broken since that was merged on my feature branch | 14:32 |
| sc` | i rebooted the host since then to uptake a newer kernel, but that shouldn't impact chef converging in vms | 14:33 |
| sc` | quasi-related, i've started exploring the rabbit hole that is test-kitchen. if kitchen-dokken worked, it would resolve a great deal of uncertainty | 14:39 |
| *** acormier has joined #openstack-chef | 15:47 | |
| *** acormier has quit IRC | 16:01 | |
| *** siva has joined #openstack-chef | 16:04 | |
| *** siva is now known as Guest42734 | 16:04 | |
| *** Guest42734 has quit IRC | 16:13 | |
| *** acormier has joined #openstack-chef | 18:23 | |
| *** acormier has quit IRC | 19:09 | |
| *** acormier has joined #openstack-chef | 19:14 | |
| *** acormier has quit IRC | 19:46 | |
| *** jmb_ has quit IRC | 22:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!