15:00:34 #startmeeting ironic 15:00:35 Meeting started Mon Apr 26 15:00:34 2021 UTC and is due to finish in 60 minutes. The chair is TheJulia. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:36 o/ 15:00:37 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:37 o/ 15:00:40 The meeting name has been set to 'ironic' 15:00:42 o/ 15:00:42 o/ 15:00:46 \o 15:00:50 o/ 15:00:50 o/ 15:00:57 o/ 15:00:57 o/ 15:01:20 Our growing agenda today can be found on the wiki. 15:01:21 #link https://wiki.openstack.org/wiki/Meetings/Ironic 15:01:27 To get started! 15:01:42 #topic Announcements / Reminders 15:01:57 o/ 15:02:00 #info JayF reminds us that the new ironic whiteboard is updated and in place. 15:02:09 #link https://etherpad.opendev.org/p/IronicWhiteBoard 15:02:35 JayF encourages people to continue to improve it and in particular check the review dashboard links. 15:03:15 #info The Foundation is attemptin got schedule an OpenInfra Life webcast on the 29th for a PTG recap. If anyone is interested in joining that and representing Ironic, please ping TheJulia. 15:03:16 o/ 15:03:42 #info Additionally, anyone interested in contributing content to ironicbaremetal.org should see TheJulia 15:03:45 Finally 15:04:11 yay 15:04:42 It has been pointed out to me that referring to vaccination as 5G may make some people feel uncomfortable. We *do* have people who work for several telecoms in this IRC channel, so moving forward please avoid the nanite enabled microchip vaccination jokes. 15:05:15 Does anyone have anything else to announce or remind us of? 15:05:57 I don't think we need to review action items, but I'm just checking in case 15:06:28 Merged openstack/bifrost stable/wallaby: [CI] Make Kolla job running again https://review.opendev.org/c/openstack/bifrost/+/787605 15:06:39 No action items so I *think* we can skip ahead 15:06:45 ++ 15:06:47 yup 15:07:05 Do we feel the need to review subteam status reports, given this is the week after the PTG? 15:07:36 not sure, probably need to define the new priorities first ? 15:07:49 Yeah, and I think that is all rooted in the change which needs attention on the repository 15:07:56 yep 15:07:59 so I think we can skip both and move forward to discussion for today? 15:08:05 Any objections? 15:08:08 makes sense to me 15:08:09 +1 15:08:12 yep 15:08:46 #link https://review.opendev.org/c/openstack/ironic-specs/+/784143 15:08:57 #topic Discussion 15:09:36 JayF brings us a discussion today of extreme old patches and asks if we should mass-abandon anything 6 months or older? 15:09:53 I have mixed feelings on this subject, but I wonder how others feel. 15:09:59 Yeah, it just became obvious how many old patches we have when I brought up the dashboards. 15:10:06 mmm 6 months feels still too close 15:10:09 probably? I have some patches that I intend to eventually continue 15:10:16 I think there are some patches that 100% should be mass abaandoned. I'm not sure about where we should set the line. 15:10:22 likewise though 15:10:26 12 months? 18? 15:10:36 18 month is what I'd use 15:10:41 18 probably 15:10:43 Maybe 1-ish year, but I still think we should skim the backlog to see if there is anything we want to keep 15:10:43 after all, abandoned patches can be resurrected 15:10:43 abandon, IMO, is more about removing from review queue. You can always unabandon a patch 15:10:45 actually 15:11:09 if you haven't updated your patch in, say, 12 months -- what's the value in it *not* being abandoned? 15:11:11 we can abandon everything 6 months old and ask the author to revive if they think it's useful 15:11:25 18 ++ 15:11:29 that's pretty much what I was thinking dtantsur 15:11:31 the value is that I won't lose it in hundreds of my patches 15:11:36 or a core spots it, I still think should skim through the items out there 15:11:41 but I can instant-resurrect patches that I find useful 15:11:52 if someone submits a spec and they are waiting for comments, isn't it somewhat mean to just abandon and ask them to resubmit? 15:11:57 I have no objection to 6 months really if that is the approach we take 15:12:07 mmm, I assumed we're talking about patches with negative feedback or merge conflict 15:12:07 what about abandon automatically what's 12 months old or older and discussing 6 months to 12 months ? 15:12:19 rpittau: that is a good compromise 15:12:26 dtantsur: good point 15:12:34 Bluntly, if I do this, I'm not going to discuss items beyond the line. I'm going to abandon stuff that needs it with a copy+paste message. 15:12:44 I don't think we should abandon patches that are all green (still) 15:12:49 but I don't want to dedicate the time to IRC and get consensus on a small set of patches 15:12:50 I agree 15:13:00 yeah, I don't think it's realistic to review all patches 15:13:03 so if we have 12 months as the line, then 12 months is the line 15:13:15 it's easier to resurrect those few we still find valuable 15:13:17 and if I see something that's all green along the way... maybe I'll just approve it (or review it properly) and see if it can get in 15:13:26 I think 12 months is a reasonable line 15:13:34 i think we've had discussions in the past about how to encourage folks to contribute upstream, etc, etc, so... lets make sure we do this in a human-friendly way... 15:13:52 rloo: I'll make a kind message to put in there :) You can trust me to be human friendly ;) 15:13:59 ++ 15:14:03 :) 15:14:14 sounds like 12 months with a friendly message is consensus? 15:14:19 alright 15:14:56 #agreed We will abandon patches not updated in the past 12 months to remove them from the review queue with a kind message to authors to resurrect them. 15:15:10 Next item I added a little before the meeting. 15:15:17 Is anyone interested in writing a PTG recap 15:15:17 ? 15:15:20 for Ironic itself 15:15:29 As a note, I don't expect to include specs in these abandons :) 15:15:42 JayF: sounds good to me 15:16:37 * TheJulia hears crickets 15:17:46 Okay, I guess I can string words together 15:17:55 Onward to the Baremetal SIG! 15:17:58 #topic Baremetal SIG 15:18:15 arne_wiebalck brings us an interesting question today. arne_wiebalck would you like to elaborate? 15:18:22 Sure. 15:18:40 We have no topics yet for the next meetings, so I put some proposals up. 15:18:56 And this is ofc a call for suggestions :) 15:19:02 And volunteers :-)) 15:19:24 I can talk about IPA-builder 15:19:29 The first two topics, "Getting started with Ironic: Bifrost" and "Building images for Ironic: The IPAB (not automatic brewer)" sound interesting 15:19:37 nice, thanks dtantsur ! 15:19:38 and sound like they could potentially be very short 15:19:57 Scaling and deploying topics are a bit... more complex 15:20:00 Yeah, and we could maybe reuse some of this for the website :) 15:20:13 exactly, so more thought is definitely required for htose 15:20:26 "Realities of scaling ironic, what to know!" 15:20:50 Any volunteer for a bifrost intro/demo? 15:20:58 I'm personally okay with fairly short clips too, even something just two to three minutes long could be awesome 15:21:09 agree 15:21:31 I guess I could or if someone wants to partner up I can talk about the history of it while the example is running or whatever 15:21:33 Ok, that is mostly it for the SIG, I think :) 15:21:50 thanks, TheJulia ! 15:22:18 Speaking of SIG, I would love to see if there are thoughts on "appropriate metrics" to be collected, but maybe the scaling topic might weigh in on that 15:22:34 I'll ponder scaling since that is such an abstract topic 15:22:46 appropriate metrics for monitoring a deployment? 15:23:32 Would it make sense to maybe have one talking about metal3<->ironic 15:23:45 oh, that would be nice! 15:23:46 * TheJulia wonders if dhellmann could be convinced to talk about it 15:23:49 I was going to write a blog post on ^^^ 15:23:57 or we can ask dhellmann to talk, works for me as well :) 15:24:20 maybe an underlying theme is "spread the load" 15:24:38 arne_wiebalck: metrics as in reporting back to the community, but in-deployment metrics could be a thing 15:24:40 * dhellmann tries to catch up with the scrollback 15:25:13 TheJulia: ah, ok, basically what we discussed at the PTG 15:25:17 and maybe a larger operator discussion is needed to gauge relative pain from the bottleneck points and comfort levels or needs to try and develop a semi-consistent/nice way 15:25:21 arne_wiebalck: yes 15:25:27 TheJulia: what to look at for a deployment may also be interesting 15:25:28 Anyway, that is just an idea. 15:25:36 arne_wiebalck: indeed 15:25:48 Anyway, all ideas. We should revisit this next week. 15:25:58 TheJulia: ++ 15:26:09 We have no listed RFE's today, so I'm going to go ahead and move us forward to RFE review 15:26:10 err 15:26:13 not RFE review 15:26:15 Open Discussion 15:26:19 when is the the next SIG meeting ? 15:26:55 usually 2nd Tue of each month 15:27:07 TheJulia : I imagine we can come up with better people to talk about that than me, these days. dtantsur, zaneb, someone from upstream metal3 perhaps 15:27:09 May 11th 15:27:14 ok, thanks 15:27:33 May 4th would be better :D 15:27:39 hehehe 15:27:48 heh 15:27:52 :-D 15:27:54 iurygregory: that's a sacred day, not sure :) 15:28:01 may the ironic force be with you 15:28:09 cinco de cuatro? 15:28:31 dhellmann: thanks 15:28:42 ironic force, oh my :) 15:28:45 #topic Open Discussion 15:29:02 There was something I wanted to bring up, but it has escaped me. Possibly it saw something, screamed and ran away 15:29:10 A simple question ... During in-band introspection/out-of-band inspection, for which hardware ports should ironic baremetal ports be created? 15:29:30 Ports without link? 15:29:49 What about ports which have been disabled via BIOS settings? 15:29:50 Inspector and ilo oob inspection offers knobs if memory serves 15:30:41 I think it boils down to ports with links, but it has been a very long time since I had last thought of this or looked at it. 15:30:57 And, should in-band and out-of-band behave consistently? 15:31:07 well, there are knobs 15:31:14 so... *shrug* 15:31:33 It is a choose your own adventure path, I'd return the list and have similar decision point capability 15:32:10 "all", "all link active", or "what we boot from" seems reasonable and to be essentially what there is already 15:32:22 of course, the last being a bit harder to determine 15:33:11 * rpioso looks for knobs ... 15:35:24 Well, I guess we can move to the final step of the meeting since no other topics are coming forth 15:35:32 #topic Who is going to run the next meeting? 15:35:59 TheJulia: Thank you :-) 15:36:01 This is something we discussed during the PTG and I'd like to get a volunteer each week to run the next week's meeting. 15:36:22 If there are fears/worries, it is fairly straight forward and the irc bot commands are easy :) 15:36:53 And also, the week of May ?10th? I may be completely MIA due to family stuff, so it will be unlikely for me to be able to run the meeting that week. 15:37:02 TheJulia: I can run next meeting 15:37:12 rpittau: Thanks! 15:37:25 Well, if there is nothing else, Thanks everyone! 15:37:29 o/ 15:37:45 Thanks TheJulia ! 15:37:49 thanks all 15:37:55 #endmeeting