19:01:05 #startmeeting operators_ops_tools_monitoring 19:01:06 Meeting started Wed Apr 6 19:01:05 2016 UTC and is due to finish in 60 minutes. The chair is raginbajin. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:01:07 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:01:10 The meeting name has been set to 'operators_ops_tools_monitoring' 19:01:13 o/ 19:01:17 o/ 19:01:18 o/ 19:01:43 Agenda - Can be found here - https://etherpad.openstack.org/p/osops-irc-meeting-20160406 19:01:46 #link https://etherpad.openstack.org/p/osops-irc-meeting-20160406 19:02:23 Only real set of topics is the last email I sent out looking for ideas and help about how does OSOps help collect all the meetup and other data points. 19:02:40 and I may need to run early depending on a few things. 19:03:16 #topic OSOps and Collecting Issues/Comments/User Stories 19:03:32 i didn't see much response in the ops thread on that 19:03:39 :( 19:03:41 some guy mentioned wanting scripts to migrate nova-network to neutron 19:03:46 which is kind of not related to that thread 19:04:18 Yeah and another about coming to the Product Working Group to add user stories to their project. 19:04:27 Which is the end result of what I think we can do for Operators. 19:04:45 raginbajin: i don't think i've ever actually seen something come from the product working group into nova as a spec or a requirements 19:05:00 maybe sideways via cross-project specs, which has it's own efficiency problems 19:05:30 we need someone that can take the specs from the customer and bring them to the developers 19:05:36 Exactly 19:05:37 lol 19:05:42 someone with people skills 19:06:07 hehe.. Yeah That's exactly what I am thinking. 19:06:18 We create all these etherpads during our meetups 19:06:27 how do we capture them and get them to you and other projects. 19:06:45 I can manually do it, but that's not the best way or any sort of efficient. 19:06:51 this is why i was hoping to get an ops cross-project liaison for nova to attend ops things and funnel that back to the dev team 19:07:01 melwitt and sdague and others have done that before, 19:07:13 or at least by opening bugs in nova or relaying highlights (pain points) in the dev ML 19:07:26 I asked Tom F. for some some thoughts and we are going to catch-up in a day or two. 19:07:43 ok. really just a rundown of highlights from ops meetups in the dev list is a good start 19:07:51 Hmm Ok.. That's an idea as well. See if I can get an Nova Ops liason going. 19:08:50 #idea Creating a Liason that would be the ops representative to particular projects and feed information back and forth. 19:09:08 #idea Populating the highlights of the Ops Meetup Etherpads to the Dev List as a start. 19:09:12 to be clear, it doesn't need to be an ops person, it could be a nova dev 19:09:28 just needs to be someone that can funnel the info and prioritize it 19:09:51 gotcha.. 19:10:48 was there going to be a monday summit session about some of this? 19:10:56 Yes. There is 19:11:15 Basically the same thing we are discussing now. I'm hoping to have more ideas maybe with help from Tom a few others to have more to talk about then 19:11:39 I'll be running that session then and then also the OSOps one as well, but that is hopefully going to be about more that OSOps and Operators can do. 19:12:38 I think there isn't much to discuss about this more since we are the only two talking about it this week and last. 19:12:40 :) 19:12:54 I'll keep on the operators to see if I get any more feedback from them 19:12:56 yeah, i just want to make sure that markus_z is at the right session 19:13:32 I'm going to create an etherpad once I get more info about when and where. I know the times for some sessions moved around. 19:14:08 You have anything else for this one. 19:14:09 ? 19:14:11 nope 19:14:15 cool 19:14:17 #topic Nova has removed the default flavors for new deployments. 19:14:31 just an fyi really 19:14:37 there was the favorable ops thread 19:14:43 and one in the dev list which is a bit more colorful 19:14:54 Yeah.. This one got some decent conversation on our list. 19:14:56 #link default flavor removal ops thread http://lists.openstack.org/pipermail/openstack-operators/2016-March/010045.html 19:15:17 #link default flavor removal dev thread http://lists.openstack.org/pipermail/openstack-dev/2016-April/091434.html 19:15:24 I'm still in the camp that we should have an option that people could run to add those back in during an install, but I see what everyone else is saying. 19:16:00 i think that's a choice for the deployment scripts at this point 19:16:05 and always should have been 19:16:10 we must be in the minority that use slightly modified defaults 19:16:46 but I think for us - puppet is taking care of it - so it should be good. 19:17:39 klindgren: apparently so.. Yeah. We create ours as well, but sometimes you just need an example quickly for something. Going from nothing to something is sometimes hard, which I why its ok and could have just been a file someone ran to add them 19:18:16 raginbajin: you mean like sometimes you just need a sample image to test with? 19:19:17 klindgren: if puppet is taking care of it for you, then you're probably carrying local patches 19:19:27 unfortunately puppet-nova can't manage flavors currently 19:19:41 I think people misunderstand the difference between a source and a deployment 19:19:57 we have some home grown stuff for managing flavors via puppet that I'd love to throw away 19:20:06 removing flavors from source doesn't mean that deployers won't provide them 19:20:37 so, we agreed on providing flavors for devstack, but deployers can still provide the same flavors than devstack too 19:20:46 sure, but it's an extra step that has to be taken now. 19:20:46 like puppet, Fuel or OSP... 19:20:58 that's why we have release notes 19:21:00 it'd be nice if the script for creating the flavors was part of the nova repo in tools/ or something, instead of part of devstack 19:21:10 clayton: but presumably you wouldn't want us to seed a default image right? 19:21:21 no, but you've never done that before either 19:21:25 clayton: what if we seeded a default instance so that `nova show` worked right out of the box? :) 19:21:37 clayton: right, we never should have done this either 19:21:37 I think flavors are inherently less opinionated than images 19:21:45 Setting up an image is much different than flavors. 19:21:52 clayton: agreed. 19:21:56 nova provides APIs, not items 19:21:59 it's data in the database 19:22:00 yeah 19:22:07 to clarify, I'm fine with the change. it doesn't effect us at all, but I understand where other people are coming from 19:22:23 it probably feels gratitutous if you do use them. 19:22:24 so, if people want to use API resources for creating items, that's a deployment case 19:22:31 agreed.. It's already been approved and this was just an FYI more than anything in the meeting. 19:22:57 I'd like to see if there is any new business to talk about 19:23:03 I don't see a burden for deployers adding flavors tbh 19:23:08 #topic New Business 19:23:22 @clayton you are correct apparently uts part of our tools setup that stored in puppet and is just a bash script 19:23:49 I remember that mdorman contributed a patch for doing it that wasn't accepted that I thought should have been :) 19:23:50 and slack has forever damaged me with the use of a @ infront of peoples names 19:24:12 bauzas: clearly it's a burden, it's just a very minor one. 19:24:28 clayton: exactly like any other upgrade change we do 19:24:41 as long as flavor creation works in cells v2 19:24:43 clayton: https://github.com/openstack-dev/devstack/blob/5df868e83d2761fec9cdfba1c04df17c6a766fb4/lib/nova#L947-L951 is exactly what you need 19:24:54 since it was horrible horrible horrible broken in cellsv1 19:25:10 klindgren: it will, unless tenant.name == klindgren :P 19:25:11 Can I ask someone to take over as the admin.. I have to leave but want this conversation to keep going. 19:25:29 klindgren: literally, this change came from making flavors proper cellsv2 citizens, by the way 19:25:40 yup 19:25:45 raginbajin: i can 19:25:46 w00t w00t - +100 19:25:48 then 19:26:22 #addchair mriedem 19:26:29 #chair rather 19:26:37 #chair mriedem 19:26:38 Current chairs: mriedem raginbajin 19:26:43 Thanks! 19:27:03 Sorry about that.. Thanks guys.. our next meeting will be at the Summit.. I"ll send out notes once I get back online 19:27:23 klindgren: clayton: others - did anyone else have anything new to go over? 19:27:31 sdague: again, I don't have a problem with them being removed, but at the bottom of a nearly 1000 line shell script in a repo most operators never look at it's very discoverable :) 19:27:43 isn't very discoverable I mean :) 19:28:07 there is nothing else in https://etherpad.openstack.org/p/osops-irc-meeting-20160406 19:28:27 i say we wrap this up then 19:29:07 not hearing anything, i'm going to end the meeting, thanks everyone 19:29:09 #endmeeting