16:01:02 #startmeeting openstack_chef 16:01:03 Meeting started Mon May 2 16:01:02 2016 UTC and is due to finish in 60 minutes. The chair is sc`. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:04 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:01:07 The meeting name has been set to 'openstack_chef' 16:01:53 good morning/afternoon 16:02:21 i trust everyone is still recovering from the summit, aches and pains fully propagating 16:05:16 hey 16:05:27 hey jklare 16:05:34 but yes, still recovering :D 16:05:45 :D 16:06:31 let's go right into it then 16:06:49 #topic tempest status 16:07:38 Mark was taking this on. he's on vacation, so I will be taking that on and working through what's left 16:08:23 so far, a package being installed as a dependency is causing problems. python-setuptools-0.9.8 gets installed from yum on the centos side, which conflicts with the pip installed version and fails the build 16:09:23 this isn't a problem with the ubuntu side 16:10:26 #topic things outstanding for stable/mitaka 16:11:25 in our sessions, we went over what we want to ship with stable/mitaka. that came down to finishing the docs, cleaning up lp, doing some other housekeeping with recipe names, and getting tempest working 16:12:57 i still have outstanding the updates to the bare metal aio doc from stable/liberty. i'll be getting that up asap 16:15:46 i'm going to be digging into openstack-model-t to cannibalize it for our reference deployments 16:17:04 the multi_node example in this patch https://review.openstack.org/#/c/311197/ is nearly working :) 16:17:39 nice 16:18:00 there is an outstanding bug for the libvirtd package (seems to be on ubuntu and rhel https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1577455) 16:18:01 Launchpad bug 1577455 in libvirt (Ubuntu) "virtlogd not started automatically with libvirt-bin 1.3.1-1ubuntu10~cloud0" [Undecided,New] 16:18:19 after that is solved, this should be good to go 16:21:09 cool. i'll be chasing down this problem with setuptools. i found a newer version on cbs.centos.org but it isn't included anywhere that we're consuming 16:21:28 mmh... ok 16:21:48 and it's an infra specific problem 16:22:19 is it? 16:23:10 it isn't an issue outside of openstack ci. it's not an infra thing to fix, though 16:24:02 if a newer version of setuptools were included with the openstack packages, this might not be an issue. unknown how it would coincide with the base python, though 16:24:22 i see 16:24:26 it's a multi-faceted problem full of wtf 16:24:53 openstack-integration-test::setup installs a handful of dependencies, which pulls in python-setuptools 16:25:32 removing packages from there means that downstream consumers have to rely on a pip installed version and not a packaged instance of tempest 16:25:59 though that might be for the best anyway, since most people i know end up installing from tempest master 16:26:20 which also might be the only option here 16:27:09 yeah 16:28:02 though i'd like to use packages all the way across for both distros, tempest (and rally when it happens) necessitate deviation from that mindset 16:30:02 did you check how puppet or ansible do this? 16:30:11 not yet, that's my next step 16:31:52 at first glance, looks like they do just that 16:32:14 so that makes the decision easy enough. there's only one path forward 16:32:43 :) 16:39:00 since there's only one way forward with tempest, i'll be looking into refactoring the tempest cookbook to not use distro packages 16:39:14 sounds good 16:39:34 maybe we can just copy some code from puppet there 16:39:41 or ansible 16:39:44 or whatever 16:39:45 perhaps 16:40:05 i only looked at puppet's implementation briefly, to get an idea as to how 16:42:09 #topic open floor 16:43:53 i've got nothing else for this one. anything else jklare? 16:44:12 nothing from my side 16:44:36 alright. going to call it 16:44:37 thanks! 16:44:46 #endmeeting