| opendevreview | Jay Faulkner proposed openstack/ironic master: DB & Object layer for node.shard https://review.opendev.org/c/openstack/ironic/+/864236 | 00:04 |
|---|---|---|
| opendevreview | Jay Faulkner proposed openstack/ironic master: API support for CRUD node.shard https://review.opendev.org/c/openstack/ironic/+/866235 | 00:05 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Allow port queries by shard list https://review.opendev.org/c/openstack/ironic/+/872361 | 00:05 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Add support for filtering for sharded nodes https://review.opendev.org/c/openstack/ironic/+/872472 | 00:05 |
| opendevreview | Merged openstack/ironic master: Fix API docs to document port filtering behavior https://review.opendev.org/c/openstack/ironic/+/872229 | 01:20 |
| opendevreview | Merged openstack/ironic stable/zed: Fix "'NoneType' object is not iterable" in RAID https://review.opendev.org/c/openstack/ironic/+/869935 | 02:25 |
| arne_wiebalck | Good morning, Ironic! | 07:20 |
| kubajj | Good morning Ironic | 09:24 |
| rpittau | good morning ironic! o/ | 09:42 |
| opendevreview | Riccardo Pittau proposed openstack/python-ironic-inspector-client master: CI fixes https://review.opendev.org/c/openstack/python-ironic-inspector-client/+/867842 | 10:57 |
| iurygregory | morning Ironic | 11:56 |
| *** priteau_ is now known as priteau | 12:25 | |
| *** dmellado_ is now known as dmellado | 12:58 | |
| kubajj | dtantsur: sorry to bother you again, I think this one is ready for review and should fix the changes you suggested on Friday https://review.opendev.org/c/openstack/ironic/+/871394 | 13:52 |
| TheJulia | good morning | 13:52 |
| iurygregory | good morning TheJulia | 14:04 |
| * TheJulia glares at the work laptop which has apparently decided it cannot use the wifi adapter anymore | 14:22 | |
| JayF | I'm here this morning, iurygregory is going to run the meeting since my weekend was (and sleeping schedules are) shot since my wife has covid | 14:56 |
| JayF | nothing like sleeping in the guest bedroom of your own house for a few days lol | 14:56 |
| iurygregory | #startmeeting ironic | 15:00 |
| opendevmeet | Meeting started Mon Feb 13 15:00:05 2023 UTC and is due to finish in 60 minutes. The chair is iurygregory. Information about MeetBot at http://wiki.debian.org/MeetBot. | 15:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 15:00 |
| opendevmeet | The meeting name has been set to 'ironic' | 15:00 |
| rpittau | JayF: sorry about that! | 15:00 |
| iurygregory | Hello everyone, welcome to our weekly meeting o/ | 15:00 |
| rpittau | o/ | 15:00 |
| iurygregory | who is here today? o/ | 15:01 |
| matfechner | o/ | 15:01 |
| JayF | o/ | 15:01 |
| iurygregory | #topic announcements/reminder | 15:01 |
| iurygregory | #info feel free to tag patches that need reviews with the hashtag ironic-week-prio | 15:02 |
| iurygregory | #info OIS Vancouver 2023 Schedule is out! | 15:03 |
| iurygregory | #link https://vancouver2023.openinfra.dev/a/schedule | 15:03 |
| iurygregory | No results about the Forum sessions yet | 15:03 |
| iurygregory | this is all I have for announcements/reminder, does anyone have something that would like to add to it? | 15:04 |
| iurygregory | ok, moving on | 15:05 |
| iurygregory | no action items from previous meeting, skipping | 15:05 |
| iurygregory | #topic Review Ironic CI status & update whiteboard if needed | 15:05 |
| iurygregory | Did anyone notice something regarding CI that would like to provide an update? | 15:06 |
| JayF | I think it's been smooth | 15:07 |
| JayF | I've had several clean runs for my API changes on master | 15:07 |
| TheJulia | o/ | 15:07 |
| JayF | I think there might be a few older branches still flakey, but things are pretty good | 15:07 |
| rpittau | nothing to report | 15:08 |
| iurygregory | ack, tks JayF rpittau | 15:08 |
| iurygregory | #topic 2023.1 Work in progress | 15:08 |
| iurygregory | #link https://etherpad.opendev.org/p/IronicWorkstreams2023.1 | 15:09 |
| TheJulia | I need to rev the metrics stuff today | 15:09 |
| JayF | sharding is close, I'll get another revision in this morn | 15:09 |
| TheJulia | reviewing it now | 15:10 |
| rpittau | metal3 ci integration is slowly moving forward, but nothing exceptional :/ | 15:10 |
| iurygregory | I don't think kubajj or dtantsur are around for us to get an update regarding Merging inspector into ironic, but I know there was good progress on it, and saw some updates in patches today | 15:12 |
| iurygregory | so we should be fine, moving on? | 15:12 |
| kubajj | o/ | 15:12 |
| kubajj | The first step (adding the inventory API) is almost done. There are just minor follow ups to finish. | 15:13 |
| rpittau | awesome | 15:13 |
| iurygregory | cool! | 15:13 |
| iurygregory | tks for the update kubajj o/ | 15:13 |
| iurygregory | no items in RFE review, so skipping | 15:14 |
| iurygregory | #topic VirtualPDU update | 15:14 |
| iurygregory | rpittau, any updates you would like to share? | 15:15 |
| rpittau | not good ones | 15:15 |
| * iurygregory hides | 15:15 | |
| TheJulia | rutro | 15:15 |
| rpittau | still no news for the patch or the adoption yet | 15:16 |
| rpittau | I sent a mail one week ago | 15:16 |
| rpittau | going to bump that duringthis week | 15:16 |
| iurygregory | =( | 15:16 |
| iurygregory | let us know if there is something we can help with | 15:17 |
| rpittau | yesh, thanks, will do | 15:18 |
| iurygregory | #topic Open discussion | 15:18 |
| iurygregory | anything we would like to discuss today? | 15:18 |
| JayF | So, we're really, really close to the release. | 15:19 |
| TheJulia | So I'm curious if anyone would be interested in being able to disable or select focus areas of metrics | 15:19 |
| JayF | If we were an RC-using project, we'd be cutting that soon | 15:19 |
| iurygregory | ok we have two topics lol | 15:20 |
| JayF | TheJulia: like "application" vs "hardware" metrics? Or something more generic e.g. "all outgoing metric keys must match this regexp" | 15:20 |
| TheJulia | JayF: I was thinking hardware specifically | 15:20 |
| TheJulia | Do we know what week is final releases -1 ? | 15:20 |
| JayF | release date is 3/22 | 15:21 |
| JayF | R-1 3/13-17 | 15:21 |
| JayF | So we have almost exactly a month | 15:21 |
| iurygregory | Mar 13 - Mar 17 R-1 Final RCs and intermediary releases | 15:21 |
| TheJulia | 3-ish weeks, we will need to release early that week at the latest | 15:22 |
| iurygregory | Feb 27 - Mar 03 - RC1 | 15:22 |
| JayF | We generally don't cut RCs | 15:22 |
| TheJulia | since at the end of the week they won't actually push the button | 15:22 |
| JayF | right? | 15:22 |
| JayF | let me say, our releases-profile (cycle-with-intermediary) says we don't use RCs | 15:22 |
| iurygregory | JayF, no we don't cut RC, but this two are normally when we release the stable/<name> | 15:23 |
| TheJulia | it is when the angry emails start | 15:23 |
| JayF | ah | 15:23 |
| TheJulia | and the attempts to force us into doing RC releases if it plays out like it typically does | 15:23 |
| iurygregory | and when we will get emails with "hey you haven't done a release for x y z" | 15:23 |
| JayF | So I'll say, next meeting, lets make some time and see what's getting in/what isn't | 15:23 |
| iurygregory | :D | 15:23 |
| JayF | and everyone push to complete their items | 15:23 |
| JayF | I'm especially thinking about the metal3 CI job + autocommit revert | 15:24 |
| JayF | if we miss that this cycle it'll be a huge disappointment IMO | 15:24 |
| JayF | given how far ahead of the curve we already are on sqla2 | 15:24 |
| rpittau | JayF: you can't imagine how disappointing it is for me :) | 15:24 |
| JayF | rpittau: idk, my imagination is super good at imagining horrible things :| | 15:24 |
| TheJulia | yeah, which reminds me, I never got an answer from the oslo folk on the error we see :( | 15:25 |
| JayF | there aren't many oslo folk :( | 15:25 |
| TheJulia | true | 15:25 |
| JayF | might wanna take it to the list? | 15:25 |
| JayF | TheJulia: your metrics question; I think it's useful and interesting | 15:26 |
| JayF | TheJulia: you thinking of trying to sneak that in with the rest of the prom support this cycle? | 15:26 |
| TheJulia | dunno | 15:26 |
| TheJulia | my plate feels super full | 15:26 |
| JayF | Makes sense; was just curious | 15:27 |
| iurygregory | it would be interesting to select if you want metrics of given nodes (like a flag that would provide or not the sensor data) even if is enabled in the conf | 15:27 |
| TheJulia | That actually makes snese | 15:27 |
| TheJulia | to just add in the conditional | 15:28 |
| JayF | my main question is, more or less | 15:28 |
| JayF | with the changes we have in place now, prom metrics users get app metrics, right? | 15:28 |
| iurygregory | regarding selecting "focus areas" are you talking about "I just want fan metrics" | 15:28 |
| iurygregory | or just "power" ? | 15:28 |
| JayF | can people using prom for hardware metrics opt-out of the new app support or not? | 15:28 |
| iurygregory | the operator could just choose? | 15:28 |
| TheJulia | iurygregory: power, there is a push to expose power to end users in some threads of discussion, albeit somewhat abstracted away from the hardware. | 15:28 |
| TheJulia | JayF: it is all together right now, which means making it selectable/controllable at a high level JustMakesSense^TM | 15:29 |
| iurygregory | ++ agree | 15:29 |
| JayF | Yeah; I'm mainly wondering if there's an easy "app metrics on/off" toggle we could slip in this cycle | 15:29 |
| JayF | as to not annoy those opers who want hardware but not app metrics | 15:30 |
| JayF | but I mean, metrics are cheap, they can deal with it? lol | 15:30 |
| TheJulia | likely, since I need to rev the code as it is still wip | 15:30 |
| * TheJulia glares at the specs repo | 15:30 | |
| TheJulia | dtantsur: reason for no test on https://review.opendev.org/c/openstack/ironic/+/872658 is there is no existing testing of it, it is all mocked out so it would be entirely net new testing, and it wasn't looking like a easy lift when I was looking at it. I can try to look again later this week. | 15:34 |
| iurygregory | metrics are cheap <- depends on the config they are using XD | 15:34 |
| TheJulia | dtantsur: btw, I think your right on your second comment, the first I don't think makes much sense given there is no other way for it to transition | 15:34 |
| JayF | iurygregory: I assume that any sane metrics system should make it easyish to filter metrics downstream :) | 15:34 |
| TheJulia | We're in the age of not understanding the cost or substrate | 15:35 |
| JayF | JUST THROW SOME HARDWARE AT IT /s :D | 15:35 |
| TheJulia | so..... our definition of sane might not be the same as others | 15:35 |
| * TheJulia suddenly sees MetricsCloud deployed | 15:35 | |
| * TheJulia sees racks and racks get wheeled into the cluster | 15:35 | |
| TheJulia | Suddenly, it becomes conscious through sheet local gravity | 15:36 |
| iurygregory | are we good with open discussion? | 15:37 |
| JayF | seems like it :D | 15:37 |
| iurygregory | ok | 15:38 |
| iurygregory | thanks everyone! o/ | 15:38 |
| iurygregory | #endmeeting | 15:38 |
| opendevmeet | Meeting ended Mon Feb 13 15:38:20 2023 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 15:38 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/ironic/2023/ironic.2023-02-13-15.00.html | 15:38 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/ironic/2023/ironic.2023-02-13-15.00.txt | 15:38 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/ironic/2023/ironic.2023-02-13-15.00.log.html | 15:38 |
| TheJulia | oh heh, there is a "send_sensor_data_types" parametera lready | 15:43 |
| JayF | you just implemented it, right ;) | 15:43 |
| JayF | wow, such fast! | 15:43 |
| dtantsur | TheJulia: first comment = including both INSPECTING and INSPECTWAIT? I thought, OOB implementations transition from INSPECTING, while the in-band one transitions from INSPECTWAIT? | 15:50 |
| TheJulia | oh, hmm | 15:51 |
| TheJulia | good point, I'll check because I didn't think oob | 15:51 |
| TheJulia | if so, then yeah, it needs to be both | 15:52 |
| opendevreview | Mohammed Boukhalfa proposed openstack/sushy-tools master: Add mac address to the fakedriver node https://review.opendev.org/c/openstack/sushy-tools/+/873606 | 16:30 |
| opendevreview | Mohammed Boukhalfa proposed openstack/sushy-tools master: Add mac address to the fakedriver node https://review.opendev.org/c/openstack/sushy-tools/+/873606 | 16:32 |
| rpittau | good night! o/ | 16:51 |
| *** jamesdenton_ is now known as jamesdenton | 16:56 | |
| JayF | Reminder: Openinfra summit early bird pricing is only until 2/15 | 16:59 |
| JayF | so if you know you're going in person, make sure to sign up now | 16:59 |
| JayF | ...what hotel are folks booking? there doesn't seem to be a conference hotel | 17:01 |
| jrosser | does anyone use ironic+horizon plugin on zed? | 17:12 |
| jrosser | i am able to launch baremetal instances from the CLI but not via horizon, and i think this worked before we did Y->Z upgrade | 17:13 |
| jrosser | the error is like this https://paste.opendev.org/show/bv0JKcLCkZBWSxqwCnpg/ | 17:14 |
| JayF | I don't have your answer jrosser, sorry, hopefully someone else will see and respond. If you don't hear back; please post it to the list | 17:15 |
| jrosser | sure | 17:15 |
| TheJulia | jrosser: so vif detachment is not a hard failure | 17:15 |
| TheJulia | there is surely other errors? | 17:16 |
| jrosser | let me double check whats happening | 17:16 |
| TheJulia | horizon via nova? | 17:17 |
| jrosser | oh thats interesting - just digging a bit deeper below "vif detachment is not a hard failure" and i think i know whats happening | 17:27 |
| jrosser | it looks like it's deploying this time, the different being i paid more attention to the "launch instance -> source" to make sure "create new volume" is No. | 17:29 |
| JayF | \o/ | 17:30 |
| JayF | Those are the easiest bugs to fix! | 17:30 |
| jrosser | which otherwise is default Yes for a BFV ceph setup for vm | 17:30 |
| TheJulia | ahh yes | 17:32 |
| TheJulia | that would do it | 17:32 |
| opendevreview | Mohammed Boukhalfa proposed openstack/sushy-tools master: Add mac address to the fakedriver node https://review.opendev.org/c/openstack/sushy-tools/+/873606 | 17:39 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: DB & Object layer for node.shard https://review.opendev.org/c/openstack/ironic/+/864236 | 17:40 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: API support for CRUD node.shard https://review.opendev.org/c/openstack/ironic/+/866235 | 17:40 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Allow port queries by shard list https://review.opendev.org/c/openstack/ironic/+/872361 | 17:40 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Add support for filtering for sharded nodes https://review.opendev.org/c/openstack/ironic/+/872472 | 17:40 |
| *** jgwentworth is now known as melwitt | 18:33 | |
| TheJulia | stevebaker[m]: JayF made the requested changes on https://review.opendev.org/c/openstack/ironic/+/866235/24..25 with one minor exception which I think is good as a follow-up | 19:22 |
| JayF | I have extremely strong feelings against forcing fixing existing-bugs in new feature things | 19:23 |
| JayF | it should be fixed tho | 19:23 |
| TheJulia | agree 100% | 19:23 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: DB & Object layer for node.shard https://review.opendev.org/c/openstack/ironic/+/864236 | 19:46 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: API support for CRUD node.shard https://review.opendev.org/c/openstack/ironic/+/866235 | 19:46 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Allow port queries by shard list https://review.opendev.org/c/openstack/ironic/+/872361 | 19:46 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Add support for filtering for sharded nodes https://review.opendev.org/c/openstack/ironic/+/872472 | 19:46 |
| JayF | I missed dropping an import when clearing the schemas out :| | 19:46 |
| TheJulia | surprising you had to update all or them to just change one | 19:48 |
| JayF | well, two thigns: 1) it was the 2nd patch in the series, whcih meant 2/3/4 had to update | 19:50 |
| JayF | 2) I use a rebase -i workflow, so usually I end up rebasing all my patches, but gerrit is smart enough to retain votes even though it re-runs CI (if it's only a rebase) | 19:51 |
| TheJulia | interactively editing during rebase? | 19:51 |
| JayF | so for instance, with this change | 19:51 |
| JayF | I start at tip of the chain | 19:51 |
| JayF | make the desired edit | 19:51 |
| TheJulia | oh, I use -i, but only from the change I edit | 19:51 |
| JayF | git add; git commit -m 'REBASE INTO CHANGE X' | 19:52 |
| JayF | git rebase -i gerrit/master; then sort the commits and mark as fixup | 19:52 |
| JayF | I have tried other workflows (even yours), and it usually ends up with me orphaning a patch ... this way may get more patches updated than needed, but it works pretty reliably and honestly it's just waht I know at this poitn | 19:52 |
| stevebaker[m] | JayF, TheJulia : Oh I should have clarified I wasn't expecting the headings issue to be fixed in this series | 19:53 |
| TheJulia | https://review.opendev.org/c/openstack/ironic/+/866235/25..26 +2'ed | 19:53 |
| *** tzumainn_ is now known as tzumainn | 20:12 | |
| opendevreview | Julia Kreger proposed openstack/ironic master: Get conductor metric data https://review.opendev.org/c/openstack/ironic/+/865447 | 20:15 |
| TheJulia | stevebaker[m]: do you remember if we cache the hash ring in the api at all? | 20:15 |
| *** bodgix9 is now known as bodgix | 20:19 | |
| *** shadower9 is now known as shadower | 20:19 | |
| TheJulia | hash ring does it itself | 20:21 |
| opendevreview | Merged openstack/ironic master: [DOC] Set cleaning requirement with retirement https://review.opendev.org/c/openstack/ironic/+/871829 | 20:46 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Get conductor metric data https://review.opendev.org/c/openstack/ironic/+/865447 | 20:50 |
| TheJulia | JayF: clarified ^, it is a noted issue in the reno as well | 20:51 |
| JayF | Is that some kind of, technical limitation? | 20:51 |
| JayF | It seems like an extremely weird hair to split | 20:51 |
| TheJulia | yeah, it is a technical limitation until we figure out *how* to do it (if we can like with a wsgi app) | 20:52 |
| TheJulia | the wsgi side is a little weird since it is not a long lived process potentially, it could be slayed and started at will and there could be numerous processes | 20:52 |
| TheJulia | also means the data might get *stupid* weird | 20:53 |
| JayF | I find myself wondering why statsd is different? | 20:53 |
| JayF | because it just pushes, in the existing process, and needs no aggreggation step? | 20:54 |
| TheJulia | it triggers a socket immediately | 20:54 |
| JayF | yeah ok, I understand now | 20:54 |
| TheJulia | so it generates the overhead then and there, and is not cached/batched up | 20:54 |
| JayF | you'd almost need to emit it to the conductor for collection | 20:54 |
| JayF | which is a whole ball of wax that makes the value prop not as good | 20:55 |
| TheJulia | .... yeah | 20:55 |
| TheJulia | but... hmm | 20:55 |
| TheJulia | hmmmmmm | 20:55 |
| TheJulia | *hmmmmmmmm* | 20:55 |
| JayF | send_request_metrics() RPC call incoming | 20:55 |
| TheJulia | not today, or likely this month | 20:55 |
| JayF | which all requests in API end with if metrics collection enabled | 20:55 |
| TheJulia | the scatter/gather "go do something idea" seems more powerful to me | 20:55 |
| JayF | I'm just curious if you were thinking something more 4d cess than that lol | 20:55 |
| JayF | *chess | 20:55 |
| TheJulia | nah, not right now | 20:56 |
| TheJulia | we would likely need to teach the conductor to keep a second cache, and then there websever would need to stream it to a specific conductor | 20:57 |
| TheJulia | which is doable ... | 20:58 |
| TheJulia | kind of | 20:58 |
| TheJulia | I guess the data resolution starts to shift when you can run hundreds of api services | 20:58 |
| *** iurygregory_ is now known as iurygregory | 21:22 | |
| stevebaker[m] | JayF: the shard series looks good now, I suppose we'll need a release note though | 21:41 |
| JayF | lmao did I seriously not have a release note on that whole chain? | 21:41 |
| JayF | that sounds about right | 21:41 |
| JayF | can I just push a bonus-patch with a changelog covering it all, so I don't have to roll CI dice again? | 21:42 |
| JayF | If this was going to be backported, it'd matter, but it's not... | 21:42 |
| JayF | stevebaker[m]: ^ ? | 21:53 |
| stevebaker[m] | fine with me :) | 21:53 |
| *** dmellado_ is now known as dmellado | 22:05 | |
| TheJulia | I'd just toss a reno on the end | 22:07 |
| TheJulia | I mean... it is not like we're going to backport it | 22:07 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Add release note for node sharding https://review.opendev.org/c/openstack/ironic/+/873652 | 22:12 |
| JayF | plus it means we can bikeshed the release notes relatively painlessly lol | 22:13 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Add release note for node sharding https://review.opendev.org/c/openstack/ironic/+/873652 | 22:13 |
| TheJulia | lacy, gently wafting curtains, yes? | 22:39 |
| *** dmellado_ is now known as dmellado | 22:39 | |
| *** dmellado_ is now known as dmellado | 23:02 | |
| opendevreview | Julia Kreger proposed openstack/ironic-specs master: Add modify steps framework https://review.opendev.org/c/openstack/ironic-specs/+/872349 | 23:52 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!