16:03:00 #startmeeting blazar 16:03:01 Meeting started Thu Jun 3 16:03:00 2021 UTC and is due to finish in 60 minutes. The chair is priteau. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:03:02 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:03:04 The meeting name has been set to 'blazar' 16:03:24 Hi diurnalist 16:03:33 heyyo 16:03:35 #topic Roll call 16:03:38 Anyone else? 16:03:48 Proposed agenda for today 16:03:51 Xena release 16:03:51 IRC meeting location 16:03:51 AOB 16:04:29 #topic Xena release 16:04:58 opendevmeet is stuck 16:05:31 Nevermind 16:05:55 As a reminder our priorities are listed at https://etherpad.opendev.org/p/xena-ptg-blazar 16:06:23 I'll provide an update on some of the areas we're working on 16:06:51 * Finish merging usage enforcement: just started looking at the remaining patches again, will push updated versions 16:07:27 * Reduce divergence between Chameleon and upstream: resumed code review, starting with https://review.opendev.org/c/openstack/blazar/+/766526 16:07:40 diurnalist: There are some comments for you in this one ^ 16:08:00 yes, i just saw that, i will look into the 500 on double-PUT 16:08:44 * Reservation calendar: my new colleague has started working on this, got a bit blocked by the amount of API changes required to make the calendar work, and then got preempted to another project 16:09:11 He's found a nice looking and better maintained JS library for doing the calendar though 16:09:17 Hmm, what API changes? so far as i know, the allocations endpoint can suffice 16:09:22 THAT i am not surprised by ;) 16:09:36 I think it was extensions to the allocations API 16:09:39 that embedded d3 gantt library is long in the tooth 16:09:52 Like showing some information about leases and hosts that would be admin only 16:10:04 But this is not first-hand knowledge so I am not fully sure 16:10:07 ah right... we must have implemented this actually... 16:10:26 maybe another thing that i can try to dust off 16:11:34 I may have to take a deep dive into Chameleon vs upstream API with him to figure out what we need, he is fairly new to Blazar 16:11:55 And finally, somewhat related to Blazar, I am working on something new 16:12:32 Grafana dashboards for visualization of reservation usage (i.e. are you really using the hosts you reserved) 16:13:01 Do you have something similar in Chameleon? 16:14:12 no we do not. how will you get the data? 16:14:21 prometheus openstack-exporter 16:14:38 oh, by "used" you mean "have launched an instance on" 16:14:45 yes, we do have that 16:14:53 yes, vcpu instance allocation 16:15:01 although CPU usage is next step 16:15:17 Obviously only possible in a virtualised cloud, not bare metal 16:15:26 mmhmm 16:15:47 we ended up scraping DB state for this if i recall, and post-processing and shoving into a secondary DB 16:16:04 so not so much a general turn-key "just turn it on" thing 16:16:08 sounds cool though 16:17:06 I am not sure how well it would work with bare metal since it's vcpu-based, but maybe it could be modified to be full-host based 16:17:08 adding blazar to openstack-exporter is a good addition. which implementation of that are you using? i recall kolla uses one that we didn't pick. we picked the wrong one a few years back, kolla picked the other 16:17:24 two roads diverged, etc. 16:18:07 Actually I am not adding any blazar-specific metrics, yet 16:18:16 It's all based on host aggregates and their metadata 16:18:30 It's this exporter: https://github.com/openstack-exporter/openstack-exporter 16:18:47 priteau: for my part, i missed my goal of getting a first draft of the external plugins blueprint by M1. it's still on my list to do this, at this point i will have to just schedule it with the team 16:19:42 Understood. It's a bit task so I can't blame you :) 16:20:39 Which one is your exporter? 16:21:25 i think we use https://github.com/att-comdev/prometheus-openstack-exporter currently 16:21:48 i guess we did this before the other one existed 16:22:48 i did already do some work to get more ironic metrics in the "new" exporter though. at some point we will transition to it 16:23:12 I see, https://github.com/ChameleonCloud/openstack-exporter vs https://github.com/ChameleonCloud/prometheus-openstack-exporter 16:23:19 The att one seems unmaintained now 16:24:01 Anything else to discuss for Xena? 16:24:06 the fork is from when i was adding ironic metrics to the new exporter. i have a WIP thing for migrating us to that 16:24:18 erm, i don't think so 16:24:33 we have a new guy who started and he may pick up OSC work 16:24:41 but we'll see 16:24:54 i will start inviting him to these meetings too 16:25:10 That's great :) 16:25:33 #topic IRC meeting location 16:26:00 So the OpenStack Technical Committee has asked that we consider moving meetings to our own channel 16:26:04 That's #openstack-blazar 16:26:30 I am not sure what the rationale is for this, since before it was advised *not* to do this and use the shared meeting channels instead 16:26:54 Anyway, if you don't have objections to join #openstack-blazar next time, I will propose the patch to update the irc meeting definition 16:27:30 That would be for the meeting in 2 weeks time 16:27:59 #topic AOB 16:28:04 Anything else for today? 16:29:16 If not we can end early 16:29:25 i don't think so. no objections to moving to the blazar channel. i don't have any history on the thoughts behind the original move 16:30:13 Let's talk over there in two weeks then! 16:30:23 Have a good rest of the day 16:30:26 #endmeeting