| vanou | good morning ironic | 01:56 |
|---|---|---|
| opendevreview | Vanou Ishii proposed openstack/ironic master: [iRMC] indentify BMC firmware version https://review.opendev.org/c/openstack/ironic/+/865074 | 03:31 |
| opendevreview | Vanou Ishii proposed openstack/ironic master: [iRMC] Handle IPMI incompatibility in iRMC S6 2.x https://review.opendev.org/c/openstack/ironic/+/865075 | 03:31 |
| opendevreview | Vanou Ishii proposed openstack/ironic master: [iRMC] indentify BMC firmware version https://review.opendev.org/c/openstack/ironic/+/865074 | 03:34 |
| opendevreview | Vanou Ishii proposed openstack/ironic master: [iRMC] Handle IPMI incompatibility in iRMC S6 2.x https://review.opendev.org/c/openstack/ironic/+/865075 | 03:34 |
| rpittau | good morning ironic! o/ | 08:12 |
| opendevreview | Riccardo Pittau proposed openstack/ironic bugfix/20.2: Align iRMC driver with Ironic's default boot_mode https://review.opendev.org/c/openstack/ironic/+/866780 | 09:32 |
| kubajj | Good morning Ironic! | 09:39 |
| *** dmellado_ is now known as dmellado | 10:39 | |
| rpittau | please when someone has a moment vheck https://review.opendev.org/c/openstack/ironic/+/866780 thanks! | 11:17 |
| opendevreview | Riccardo Pittau proposed openstack/sushy stable/zed: Handle proper code_status in unit test https://review.opendev.org/c/openstack/sushy/+/867209 | 11:26 |
| opendevreview | Riccardo Pittau proposed openstack/sushy stable/zed: Handle proper code_status in unit test https://review.opendev.org/c/openstack/sushy/+/867209 | 11:27 |
| TheJulia | good morning | 13:32 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Add doc for lsblk: /dev/nvme0n1p2: not a block device https://review.opendev.org/c/openstack/ironic/+/867228 | 14:05 |
| opendevreview | Kirill proposed openstack/ironic-specs master: new spec: support of vnc console. https://review.opendev.org/c/openstack/ironic-specs/+/866537 | 14:29 |
| TheJulia | Kirill_: thanks for the formatting fix | 14:31 |
| TheJulia | Kirill_: +2'ed, a couple nit's. At a high level I grok what your doing, I don't think there is any impact since your functionally going to be adding another service which supports the novnc interface so the nova-compute service can return a working proxy url | 14:37 |
| Kirill_ | so if +2, which steps are next - should we move it to unimplemented and i can start do it?) | 14:40 |
| TheJulia | you'll need another +2 to merge it | 14:55 |
| JayF | Good morning | 15:01 |
| JayF | #startmeeting ironic | 15:01 |
| opendevmeet | Meeting started Mon Dec 12 15:01:52 2022 UTC and is due to finish in 60 minutes. The chair is JayF. Information about MeetBot at http://wiki.debian.org/MeetBot. | 15:01 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 15:01 |
| opendevmeet | The meeting name has been set to 'ironic' | 15:01 |
| TheJulia | o/ | 15:01 |
| iurygregory | o/ | 15:01 |
| dtantsur | o/ | 15:02 |
| matfechner | o/ | 15:02 |
| hjensas | o/ | 15:02 |
| JayF | #topic Announcements/Reminder | 15:02 |
| ajya | o/ | 15:02 |
| rloo | o/ | 15:02 |
| JayF | Standing reminder to tag your patches that are ready-for-review with hashtag ironic-week-prio | 15:02 |
| JayF | and to review hashtagged changes as well | 15:02 |
| JayF | Reminder also that meetings are cancelled 12/26 and 1/2 to allow folks to celebrate the holiday | 15:03 |
| JayF | #topic Action items from previous meeting | 15:03 |
| rpittau | p | 15:04 |
| rpittau | p/ | 15:04 |
| rpittau | o/ | 15:04 |
| rpittau | argh... | 15:04 |
| JayF | I did push that MR up to add an ironic-releases group; will be getting that populated with folks and getting things in place to retire old bugfix branches. | 15:04 |
| JayF | I'm thinking it might be wise to wait until Jan 2023 to activate on that? Does anyone have a strong opinion? | 15:04 |
| Kirill_ | o/ | 15:04 |
| JayF | I guess nobody cares about timing on that then? | 15:06 |
| JayF | rpittau: I also had you listed as having an action to remove centos8 jobs from bifrost/IPE wallaby | 15:06 |
| JayF | I think I saw that patch come through? | 15:06 |
| rpittau | JayF: that's done | 15:07 |
| iurygregory | I don't think we are in a hurry to retire them, or are we? | 15:07 |
| TheJulia | jayf: no opinion from my point of view | 15:07 |
| opendevreview | Mark Goddard proposed openstack/ironic-python-agent stable/yoga: Warn when smartctl not found https://review.opendev.org/c/openstack/ironic-python-agent/+/867159 | 15:07 |
| opendevreview | Mark Goddard proposed openstack/ironic-python-agent stable/xena: Warn when smartctl not found https://review.opendev.org/c/openstack/ironic-python-agent/+/867160 | 15:07 |
| opendevreview | Mark Goddard proposed openstack/ironic-python-agent stable/wallaby: Warn when smartctl not found https://review.opendev.org/c/openstack/ironic-python-agent/+/867161 | 15:07 |
| iurygregory | we can have a deadline but it can probably wait till begin of 2023 | 15:07 |
| JayF | iurygregory: the only real driving factors are that it's confusing when backporting stuff, and that the olld bugfix branches are all showing up in zuul-config-errors | 15:07 |
| iurygregory | JayF, I see | 15:07 |
| JayF | Yeah; I'm thinking I'll get all the commands lined up, sent to the list to review | 15:07 |
| JayF | and then execute after holiday | 15:08 |
| JayF | I just don't want someone to get broken christmas week :) | 15:08 |
| TheJulia | ++ | 15:08 |
| JayF | aight it looks like we successfully did the things | 15:08 |
| JayF | #action JayF to document steps needed to retire bugfix branches; and schedule it to happen sometimes post-holiday (2023) | 15:08 |
| JayF | Next up | 15:08 |
| JayF | #topic 2023.1 WIP | 15:08 |
| JayF | Shards workflow is actively blocked on reviews at this point | 15:09 |
| JayF | please help us get the spec landed in ironic side so we can start landing db /api code | 15:09 |
| JayF | I'll make sure to put a review on 860731 for SQLA 2.0+inspector | 15:10 |
| JayF | #link https://etherpad.opendev.org/p/IronicWorkstreams2023.1 | 15:10 |
| TheJulia | ++ | 15:10 |
| JayF | Moving on | 15:11 |
| JayF | there are no RFEs for review, skipping that | 15:11 |
| JayF | #topic Open Discussion | 15:11 |
| JayF | Anything for open discussion today? | 15:11 |
| TheJulia | I'm happy to get on a call and walk through ether the shard spec or the inspector sqlalchemy stuffs | 15:11 |
| JayF | I'd be happy to participate in such a review jam | 15:11 |
| JayF | such jam, wow | 15:11 |
| rpittau | is anyone looking at the ironic weekly priorities recently? | 15:12 |
| rpittau | https://review.opendev.org/q/hashtag:ironic-week-prio+status:open | 15:12 |
| rpittau | we have a lot still under review since november | 15:12 |
| dtantsur | TheJulia: I'm looking through the sqla stuffs right now | 15:12 |
| rloo | (I hope to go through the shard spec one more time today, if i have outstanding questions will ping TheJulia and/or JayF) | 15:12 |
| JayF | I've been trying to get the metrics bits done for reviews rpittau, I think they are incredibly down across the board | 15:13 |
| JayF | rpittau: I'll personally make sure to make a pass; but I am concerned overall review capacity in the project is diminished right now for reaosns I can't see/understand | 15:13 |
| JayF | I think we might have fallen into the trap where people are reviewing things they have comments for but not putting a lot of approval +1/+2s on stuff too, but I am not sure | 15:13 |
| JayF | Either way I noticed that too rpittau and would appreciate if cores across the project worked to get things landed | 15:14 |
| rpittau | JayF: I'll have a look this week | 15:14 |
| JayF | thanks for bringing it up :D | 15:15 |
| rpittau | :) | 15:15 |
| JayF | Alright, is there anything else for Open Discussion? | 15:16 |
| JayF | If not we can give folks a nice 40 minutes+ back :D | 15:16 |
| TheJulia | sounds good to me | 15:17 |
| JayF | #endmeeting ironic | 15:18 |
| opendevmeet | Meeting ended Mon Dec 12 15:18:00 2022 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 15:18 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/ironic/2022/ironic.2022-12-12-15.01.html | 15:18 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/ironic/2022/ironic.2022-12-12-15.01.txt | 15:18 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/ironic/2022/ironic.2022-12-12-15.01.log.html | 15:18 |
| Kirill_ | Hi, can someone help to review: https://review.opendev.org/c/openstack/ironic-specs/+/866537. TheJulia has already done it, need someone more)) | 15:18 |
| TheJulia | I was just going to mention that! | 15:18 |
| Kirill_ | oh)))thanks | 15:18 |
| rloo | TheJulia: qq, does this need a reno? https://review.opendev.org/c/openstack/ironic/+/866933 | 15:18 |
| TheJulia | It is a fairly quick read, semi-vague because it is primarily forklifting a service from nova and implementing an ironic version, the existing console specs should cover the API interaction with ironic, fwiw. | 15:19 |
| TheJulia | rloo: likely does, it is not explicitly required to backport, fwiw | 15:19 |
| TheJulia | I can add one real quick | 15:19 |
| rloo | TheJulia: ok, ping me when it is ready. going to look at shard spec now... | 15:19 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Catch any exception for Cleaning https://review.opendev.org/c/openstack/ironic/+/866933 | 15:22 |
| TheJulia | rloo: ^^^ done | 15:22 |
| rloo | thx TheJulia! JayF if you want to approve ^^ | 15:23 |
| JayF | I hate it when they update under me LOL | 15:23 |
| JayF | +2a | 15:23 |
| TheJulia | Anyone object if I single core workflow a networking-generic-switch submitted driver | 15:34 |
| * dtantsur does not | 15:36 | |
| dtantsur | anyone knows why this failed? https://zuul.opendev.org/t/openstack/build/3eb1359f5b5d422c91b77c8615c57c25 | 15:36 |
| JayF | that was the tox version update | 15:39 |
| JayF | it broke a lot of stuff all the way around | 15:39 |
| JayF | if that's not a recent result; recheck and it should pass | 15:39 |
| JayF | if it's a recent result then they probably didn't land the fix yet :| | 15:40 |
| rpittau | I think all the fixes landed | 15:40 |
| rpittau | https://review.opendev.org/q/I9a138af94dedc0d8ce5a0d519d75779415d3c30b | 15:41 |
| JayF | nice | 15:41 |
| * TheJulia needs a very large cup of coffee | 15:52 | |
| rloo | JayF, TheJulia I've commented on https://review.opendev.org/c/openstack/ironic-specs/+/861803/19. Over to you. Is anyone else going to review? | 15:59 |
| JayF | arne_wiebalck had comments on it at some point; I hope someone else will review too | 15:59 |
| JayF | as we said in the meeting; there's just not a lot of folks reviewing | 16:00 |
| rloo | the overall 'shard' idea is fine. So I'm good with a MVP, V1 that provides the basic plumbing. It just isn't clear to me how to bootstrap this with the consumer. the consumer cannot 'self delineate' if it doesn't have any input so I am scratching my head there. | 16:01 |
| TheJulia | I'm on a quick call at the moment | 16:02 |
| rloo | LMK if you would like to meet to discuss. | 16:02 |
| rloo | (am avail most of today) | 16:02 |
| JayF | I think the answer to that will just be documentation | 16:03 |
| JayF | e.g. "run a lot of node update calls to set a shard" | 16:03 |
| JayF | we don't touch on migration much in the ironic spec though, that is fair | 16:04 |
| JayF | it is mentioned in the nova spec; but not the pieces you're asking about I think | 16:04 |
| rloo | i can envision manual stuff to do this. but not automated stuff. | 16:04 |
| JayF | Yeah; I am not sold on automated shard setting | 16:04 |
| JayF | which is part of why I wanted it scoped out of the first spec | 16:04 |
| rloo | i think automated might make sense, but need more details to work the kinks out. | 16:05 |
| JayF | yeah; I'm not even pointing brainpower at that | 16:05 |
| JayF | until the whole more-manual story is done | 16:05 |
| rloo | and yes, it would be easier if this spec did not attempt to do anything automated. | 16:05 |
| JayF | it doesn't | 16:05 |
| JayF | if it says it does it's accidental and leftover :) | 16:05 |
| rloo | there is a stretch goal or something. | 16:06 |
| JayF | yeah, just as sorta a visioning thing "this may be a next step" | 16:06 |
| JayF | I can tell you I'm not taking that step until ironic+nova pieces are done and talking :D | 16:06 |
| rloo | that isn't a stretch goal... so it should be reworded... anyway, that's why we're discussing this :) | 16:07 |
| JayF | I'd be OK with completely removing it from the spec if it's a point of contention | 16:07 |
| JayF | and just leave it for future() | 16:07 |
| rloo | i have to read/review the spac as it is worded... | 16:07 |
| *** hjensas is now known as hjensas|out | 16:11 | |
| *** jamesdenton_ is now known as jamesdenton | 16:12 | |
| TheJulia | okay, off that call | 16:14 |
| TheJulia | Ahh, yeah, so the idea is it becomes opt-in, and they can through node-update calls | 16:15 |
| TheJulia | I *think* one thing that might make migration "easy" moving forward, which is not clearly defined in Nova yet, but should likely come about likely in the next cycle a concept to possibly just "evacuate the entire compute node", but would be much automated, but slower in a sense. :\ | 16:19 |
| JayF | I'm trying to push off that higher level logic for now | 16:19 |
| JayF | I want us to have node-update working, GET /v1/shard, and that about it for this spec | 16:20 |
| TheJulia | Yeah, I think it will come somehow | 16:20 |
| JayF | then plug in the nova stuff | 16:20 |
| JayF | then after that, we can evaluate if we want ironic magic, nova magic, or some combinatino | 16:20 |
| JayF | but for now, I just want foundations in place for nova driver to jump off from | 16:20 |
| TheJulia | ++ | 16:20 |
| TheJulia | build the foundation, let the next level evolve from there | 16:21 |
| TheJulia | and those next levels might be so stupidly simple we won't need much in the way of process to go "oh yeah, that makes sense" | 16:21 |
| TheJulia | the challenge is... the foundation is a bit the Doctor's Tardis. | 16:21 |
| JayF | so if I read rloo's comments here correctly, it might be useful to just remove that stretch commentary altogether from the spec | 16:22 |
| JayF | and just be direct about what we're doing and let the magic follow later | 16:22 |
| TheJulia | Maybe change the heading to "future ideas out of immediate scope" | 16:22 |
| rloo | i can comment on the spec itself if that would help you. | 16:22 |
| JayF | I question the value of having possible ideas in a doc that will live even if those ideas don't get implemented? idk | 16:22 |
| JayF | I see both sides | 16:22 |
| JayF | it's good to have it written down but it's also good to only document that which is getting changed | 16:23 |
| rloo | i'm ok with possible future ideas or whatever. | 16:23 |
| TheJulia | we're good at loosing ideas and oinly half implementing others at times | 16:23 |
| TheJulia | so the line is kind of hard to draw | 16:23 |
| rloo | i'm not ok with high level stuff that this spec is supposed to describe ) | 16:23 |
| rloo | we've done this before. had several specs to get some feature done. | 16:24 |
| TheJulia | rloo: what high level stuff specifically? | 16:24 |
| rloo | the automated stuff for example. | 16:24 |
| TheJulia | so we can strip it all out | 16:25 |
| rloo | i'll comment in the spec. give me a few min. | 16:25 |
| TheJulia | one of the really negative feedbacks we got on resource_class is the lack of auto-setting | 16:25 |
| TheJulia | since this is sort of similar, it is entirely for a consumer | 16:26 |
| TheJulia | also.. humans didn't quite know what to set because placement was a mystery for most mortals | 16:26 |
| JayF | the exception here is that small scale users of ironic arguably won't need to set shard | 16:28 |
| JayF | which is separate from resource_class | 16:28 |
| JayF | in fact; part of why I'm slightly skeptical about too much magic is that I think folks need to have an operational understanding of how/why they are sharding, rather than just letting it be magic | 16:29 |
| TheJulia | This is a good point, we just need to make sure we actually write enough details about how theory relates to practice | 16:29 |
| JayF | we have to coordinate between devops (to setup computes to manage the shards) and ironic API; just like with something like networking that requires external state be reflected in our api | 16:29 |
| TheJulia | so they can grok and apply | 16:29 |
| TheJulia | I think the challenge is a number of medium shops run highly automated pipelines these days | 16:30 |
| JayF | like, with my ops hat on | 16:30 |
| TheJulia | i.e. plug in rack and walk away while everything gets added/setup/enrolled/etc | 16:30 |
| JayF | I would've never, ever opted into any automatically-set-shard magic | 16:30 |
| JayF | if anything; I might would've wanted a feature to allow a cap set to count(nodes) per shard, and fail a create if there's "no room" | 16:31 |
| JayF | because I would want the "create a new shard" logic at a higher level, in the automation itself | 16:31 |
| TheJulia | Yeah, I guess that does play into your own business proceses and how you want to do things | 16:31 |
| JayF | yeah, I am so far from sold the automated stuff would be a win but I mostly didn't have this conversation because it was in the "maybe eventually" space | 16:32 |
| JayF | and I'm happy to have those conversations on the maybe-eventually time frame lol | 16:32 |
| TheJulia | I think there are users who would love/appreicate it, but it is really not a complex feature in the end to have it as an opt in thing | 16:32 |
| TheJulia | at some point down the road | 16:32 |
| JayF | yeah, I think it will look like a minor next step | 16:33 |
| TheJulia | yeah | 16:33 |
| JayF | if we did the foundational work -- including operational docs -- well | 16:33 |
| TheJulia | yup | 16:33 |
| JayF | if it doesn't seem like a minor next step, we should reconsider :D | 16:33 |
| TheJulia | ++ | 16:33 |
| TheJulia | Which relates back to the overall effort to make it easier to contribute to OpenStack | 16:34 |
| rloo | ok JayF TheJulia. i commented in the spec, the sections to maybe reword or whatever. i think i had a few? other questions in that spec. I think that we can discuss with arne later if he isn't avail now. the basic sharding won't preclude him using the same value for conductor group as for shard value... | 16:38 |
| JayF | I think the biggest piece we failed to comm for arne_wiebalck's original comment was how the scaling is different | 16:38 |
| JayF | Ironic scales higher than all these other tools which are designed to scale to about the size of a hypervisor | 16:39 |
| JayF | so we use a shard so we can emulate that size (hypervisor) without having to sacrifice our own backend scaling freedom | 16:39 |
| TheJulia | That is a good way of putting it | 16:39 |
| rloo | (ping me if you have questions, i'm sticking my head downstream) | 16:39 |
| TheJulia | I'm going to check one thing and I'll crack it back open | 16:39 |
| TheJulia | Apparently someone is putting a telephone pole in my yard.... o O (wut?) | 16:40 |
| JayF | TheJulia: Ironically, we'd be at about the same scale, conductor:compute:networking-agent, if we never made an agent driver and everyhing was still iscsi'd ;) | 16:40 |
| TheJulia | depending on the rollover frequency, yeah | 16:44 |
| opendevreview | Merged openstack/ironic master: Fixes anaconda deploy for PXE boot https://review.opendev.org/c/openstack/ironic/+/860055 | 16:58 |
| rpittau | good night! o/ | 17:11 |
| opendevreview | Julia Kreger proposed openstack/ironic-specs master: Add a shard key https://review.opendev.org/c/openstack/ironic-specs/+/861803 | 17:27 |
| TheJulia | rloo: hopefully that clears things up | 17:37 |
| rloo | thx TheJulia! Will look in the next hour or so. | 17:38 |
| opendevreview | Merged openstack/networking-generic-switch master: Add support for Dell OS10 https://review.opendev.org/c/openstack/networking-generic-switch/+/860067 | 17:44 |
| opendevreview | Merged openstack/sushy stable/zed: Handle a different error code for missing TransferProtocolType https://review.opendev.org/c/openstack/sushy/+/866899 | 17:46 |
| opendevreview | Merged openstack/sushy stable/yoga: Handle a different error code for missing TransferProtocolType https://review.opendev.org/c/openstack/sushy/+/866901 | 17:58 |
| opendevreview | Julia Kreger proposed openstack/ironic-specs master: Add a shard key https://review.opendev.org/c/openstack/ironic-specs/+/861803 | 18:36 |
| opendevreview | Julia Kreger proposed openstack/ironic-specs master: Add a shard key https://review.opendev.org/c/openstack/ironic-specs/+/861803 | 18:41 |
| TheJulia | rloo: okay, there it should go | 18:41 |
| TheJulia | well, be good | 18:41 |
| rloo | +2 TheJulia Thanks for all the updates! | 18:44 |
| JayF | arne_wiebalck: if you have further comments on the shard key spec; please make them. Otherwise I'm going to land it, probably sometime this week | 19:29 |
| -opendevstatus- NOTICE: Gerrit will be unavailable for a short time as it is upgraded to the 3.6 release | 19:59 | |
| JayF | TheJulia: if you wanna update https://review.opendev.org/c/openstack/ironic-inspector/+/860731 I will be a quick re-review; up to you if you wanna follow-up or update inline | 20:00 |
| TheJulia | oh... hmm | 20:09 |
| TheJulia | gerrit upgrade | 20:09 |
| * TheJulia determines gerrit cookies won't work for that | 20:09 | |
| opendevreview | Merged openstack/ironic master: [grenade] Explicitly enable Neutron ML2/OVS services in the CI job https://review.opendev.org/c/openstack/ironic/+/866993 | 21:25 |
| TheJulia | JayF: I can do a follow-up tomorrow | 21:47 |
| JayF | aight | 21:48 |
| opendevreview | Merged openstack/ironic master: Fix unit tests for Python 3.11 https://review.opendev.org/c/openstack/ironic/+/866861 | 21:49 |
| opendevreview | Yusuke Okada proposed openstack/ironic stable/zed: Change boot_interface order of iRMC driver https://review.opendev.org/c/openstack/ironic/+/865998 | 22:41 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!