*** acormier has joined #openstack-chef | 00:09 | |
sc` | the good news is that only four cookbooks need to change | 00:11 |
---|---|---|
sc` | it works against master, so as soon as they are deemed okay, we should backport to stable/newton | 00:12 |
sc` | once i can reconnect to gerrit :| | 00:12 |
*** acormier has quit IRC | 00:13 | |
*** acormier has joined #openstack-chef | 00:53 | |
*** acormier has quit IRC | 00:53 | |
*** acormier has joined #openstack-chef | 00:54 | |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-common master: Corrects rabbitmq user, yum cache cleaning https://review.openstack.org/442225 | 02:02 |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-dashboard master: Corrects SELinux enablement, lbaasv2 dashboard installation https://review.openstack.org/442227 | 02:04 |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-identity master: Corrects SELinux enablement, add retries for Keystone packages https://review.openstack.org/442232 | 02:07 |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-telemetry master: Corrects Ceilometer installation, fixes typo https://review.openstack.org/442234 | 02:09 |
sc` | ^^^^ those four allowed for multi and allinone on botn centos and ubuntu to converge | 02:10 |
sc` | i have some repo changes that go with the multinode poc that i've yet to propose, mainly cosmetic changes to better align with the resources we have available | 02:11 |
sc` | the rmq user one was found on superuser. with 'guest', it fails on centos. with 'openstack', it converges and openstack works as it should | 02:12 |
sc` | please, please let me know if they don't work. my poor laptop is still fussing at me for running it as hot as i did | 02:13 |
*** acormier_ has joined #openstack-chef | 02:25 | |
*** acormier has quit IRC | 02:29 | |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-common master: Corrects rabbitmq user, yum cache cleaning https://review.openstack.org/442225 | 02:34 |
*** calbers_ has joined #openstack-chef | 03:18 | |
*** acormier_ has quit IRC | 03:18 | |
*** calbers has quit IRC | 03:21 | |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-identity master: Corrects SELinux enablement, add retries for Keystone packages https://review.openstack.org/442232 | 03:50 |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-dashboard master: Corrects SELinux enablement, lbaasv2 dashboard installation https://review.openstack.org/442227 | 03:53 |
*** emagana_ has joined #openstack-chef | 04:10 | |
*** emagana has quit IRC | 04:11 | |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-identity master: Corrects SELinux enablement, add retries for Keystone packages https://review.openstack.org/442232 | 06:25 |
sc` | the external ci seems to be happy so far | 06:31 |
*** nkrinner_afk is now known as nkrinner | 06:40 | |
*** os-chef-bot has quit IRC | 08:33 | |
*** os-chef-bot has joined #openstack-chef | 08:33 | |
*** mutin-sa has quit IRC | 08:35 | |
*** s-mutin has joined #openstack-chef | 08:36 | |
*** calbers_ has quit IRC | 08:36 | |
*** calbers has joined #openstack-chef | 08:37 | |
*** calbers has quit IRC | 08:41 | |
*** calbers has joined #openstack-chef | 09:08 | |
*** jklare has quit IRC | 09:10 | |
*** jklare has joined #openstack-chef | 09:10 | |
frickler | sc`: looks like it is only and specifically ceilometer-api that it broken, it needs to either not touch ports.conf or drop the Listen from the wsgi template. I'd prefer the first option, making it in line with how we do wsgi for nova-placement and gnocchi | 10:14 |
*** openstackgerrit has quit IRC | 10:33 | |
*** emagana has joined #openstack-chef | 10:50 | |
*** emagana_ has quit IRC | 10:51 | |
*** calbers_ has joined #openstack-chef | 12:34 | |
*** calbers has quit IRC | 12:38 | |
*** chlong has joined #openstack-chef | 13:00 | |
*** acormier has joined #openstack-chef | 13:10 | |
*** eroux has joined #openstack-chef | 13:13 | |
*** acormier has quit IRC | 13:20 | |
*** eroux has quit IRC | 13:20 | |
sc` | frickler: that's a side effect of the apache2 cookbook, unfortunately | 13:21 |
sc` | ports.conf is a debian-ism that gets splatted on rhel machines with that cookbook | 13:22 |
sc` | which means that there's a conflict with putting Listen in virtual host declarations, using that cookbook | 13:22 |
sc` | btw, the yum.conf one is incomplete. i had a run list change that i didn't push up | 13:23 |
sc` | removing yum and apt from the run lists with the changes makes yum/apt fire appropriately | 13:23 |
sc` | probably should have split that out to a different commit, natch | 13:23 |
sc` | so long as we're using apache2, Listen can't go in virtual host templates | 13:31 |
*** acormier has joined #openstack-chef | 13:41 | |
*** acormier has quit IRC | 13:45 | |
sc` | frickler: if not changing the default rabbit user, where might that change go in? it's a necessity for convergence. guest will cause centos to not spin up computes in multinode. from the compute.log it just looks like rabbit is down, but it's clearly up and functional | 13:45 |
sc` | repo can be a place for that to live, but it feels anti-pattern to move that to an environment file | 13:46 |
sc` | when testing this stuff, my laptop tried taking off more than once. i also ensnared my mac pro to speed up the build times | 13:49 |
sc` | but it all converged. openstack seemed to work from the perspective of horizon and nova | 13:49 |
frickler | sc`: I still don't understand why the default user should be working everywhere except with centos multinode | 13:49 |
sc` | nor do i. i'm still looking for that | 13:49 |
sc` | the upgrade path is still build up a new deployment, forklift, tear down the old | 13:50 |
sc` | which effectively means the cookbooks are forklifted the same way | 13:50 |
frickler | hmm, we are hoping that we can do better for mitaka -> newton currently | 13:51 |
sc` | yeah, i know | 13:51 |
sc` | i'm almost psyched up to rebase and propose the multinode poc | 13:51 |
sc` | day job has been keeping me a bit busy, too, so i didn't want to propose it and let it fester | 13:52 |
sc` | once zuulv3 comes along, that'll be way easier, with multinode jobs being first-class citizens | 13:57 |
sc` | now that i'm thinking about it, i did read about the rabbit user on superuser. trying to find the thread now | 14:00 |
*** openstackgerrit has joined #openstack-chef | 14:03 | |
openstackgerrit | Jens Rosenboom proposed openstack/cookbook-openstack-common master: Switch to 'ocata' default release https://review.openstack.org/442506 | 14:03 |
sc` | jklare: so are we skipping boston? :( | 14:10 |
jklare | sc` i am :) | 14:10 |
sc` | day job may or may not fund me to go. i'm still neck deep in openstack there | 14:11 |
jklare | sc` i would have loved to come to boston, but not only the hotels and the travel is expensive, but i also have to pay an additional 300$ for the ticket | 14:11 |
sc` | yeah. i saw that and was someone amazed that they would do that | 14:11 |
sc` | if i do wind up going, i'll rock the kangaroo | 14:13 |
*** laurent\ has quit IRC | 14:20 | |
sc` | frickler: here's the 'why' on changing the default from 'guest' to 'openstack' https://github.com/openstack/puppet-openstack-integration/blob/master/manifests/rabbitmq.pp#L21 | 14:20 |
sc` | not using guest is more secure, as it turns out | 14:23 |
sc` | who knew ;) | 14:23 |
sc` | thinking about it with a security hat on, it makes perfect sense. i was approaching it from the work/no work angle and thought it to be a quirk | 14:24 |
*** calbers_ has quit IRC | 14:26 | |
*** calbers has joined #openstack-chef | 14:26 | |
openstackgerrit | Samuel Cassiba proposed openstack/openstack-chef-repo master: Removes hardcoded apt/yum entries from run lists https://review.openstack.org/442521 | 14:29 |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-common master: Corrects rabbitmq user, yum cache cleaning https://review.openstack.org/442225 | 14:34 |
openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-identity master: Corrects SELinux enablement https://review.openstack.org/442232 | 14:41 |
sc` | the more i'm reading in to nodepool, zuul (v2 and v3), the more it seems that the use of 'multi-node' needs to change to 'multinode' to more align with the overarching naming scheme that has developed for these things | 14:50 |
*** emagana has quit IRC | 15:08 | |
*** acormier has joined #openstack-chef | 15:19 | |
*** emagana has joined #openstack-chef | 15:31 | |
*** emagana has quit IRC | 15:33 | |
*** emagana has joined #openstack-chef | 15:35 | |
*** jmb_ has quit IRC | 15:36 | |
*** emagana has quit IRC | 15:38 | |
*** emagana has joined #openstack-chef | 15:39 | |
*** emagana has quit IRC | 15:39 | |
*** emagana has joined #openstack-chef | 15:40 | |
openstackgerrit | Samuel Cassiba proposed openstack/openstack-chef-repo master: Removes hardcoded apt/yum entries from run lists https://review.openstack.org/442521 | 15:42 |
*** nkrinner is now known as nkrinner_afk | 15:53 | |
*** jmb_ has joined #openstack-chef | 16:28 | |
*** calbers has quit IRC | 16:55 | |
*** jmb_ has quit IRC | 17:18 | |
*** jmb_ has joined #openstack-chef | 17:27 | |
*** eroux has joined #openstack-chef | 19:32 | |
*** eroux has quit IRC | 20:03 | |
*** eroux has joined #openstack-chef | 20:09 | |
*** jmb_ has quit IRC | 20:31 | |
*** jmb_ has joined #openstack-chef | 20:47 | |
*** eroux has quit IRC | 21:31 | |
*** jmb_ has quit IRC | 21:36 | |
*** laurent\ has joined #openstack-chef | 22:23 | |
*** jmb_ has joined #openstack-chef | 22:30 | |
*** emagana has quit IRC | 22:52 | |
*** acormier has quit IRC | 22:55 | |
*** calbers has joined #openstack-chef | 23:30 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!