| *** matt__ is now known as matfechner | 03:06 | |
| opendevreview | Adam Rozman proposed openstack/ironic-python-agent master: prioritize lsblk as a source of device serials https://review.opendev.org/c/openstack/ironic-python-agent/+/855866 | 05:22 |
|---|---|---|
| opendevreview | Adam Rozman proposed openstack/ironic-python-agent master: prioritize lsblk as a source of device serials https://review.opendev.org/c/openstack/ironic-python-agent/+/855866 | 05:48 |
| kamleshChauvhan | Good Morning Ironic! | 06:52 |
| rpittau | good morning ironic! o/ | 07:32 |
| jm1 | dtantsur: hello :) looks like your register_machine patches for openstacksdk have been merged https://review.opendev.org/q/project:openstack/openstacksdk+owner:dtantsur%2540protonmail.com | 09:40 |
| jm1 | dtantsur: thanks for that! :) | 09:40 |
| jm1 | dtantsur: somehow this code is timing out in wait_for_provision_state https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_d6c/854113/13/check/bifrost-collections-src/d6cf784/job-output.txt | 09:41 |
| jm1 | dtantsur: in the old code the provisioning took a minute, now its timing out after 3 minutes. so i guess there is an issue with the code. maybe i am calling register_machine wrong? https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854113/14/plugins/modules/baremetal_node.py#665 | 09:45 |
| kubajj | dtantsur: just a thought, how do we deal with the software RAID skip list - inspector problem? | 10:05 |
| iurygregory | good morning ironic | 11:09 |
| kamleshChauvhan | iurygregory, Is it ok to add priority tag for https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/855710 | 11:28 |
| iurygregory | kamleshChauvhan, it's ok | 11:29 |
| iurygregory | ironic-cores, double checking ipa patches https://review.opendev.org/c/openstack/ironic-python-agent/+/855866 https://review.opendev.org/c/openstack/ironic-python-agent/+/566544 do you think we should try to include in zed ? | 12:33 |
| rpittau | iurygregory: not sure we have time this week? | 12:47 |
| iurygregory | rpittau, yeah, one had a +2 but they updated the patch (would also be good a releasenote on it...) | 12:48 |
| iurygregory | XD | 12:48 |
| TheJulia | I think it would be good to get the nvidia/mellenox thing in, not really a fan of the step name though | 12:52 |
| TheJulia | Also, good morning! | 12:52 |
| iurygregory | good morning TheJulia | 12:57 |
| iurygregory | yeah, I was reading the comments now | 12:58 |
| iurygregory | I'm checking releasenotes for inspector, no open patches \o/ | 13:00 |
| rpittau | TheJulia: I agree on the step name, at least adding a reno to explain that | 13:11 |
| TheJulia | yeah, I suspect this would be an argument to support name aliases | 13:11 |
| kamleshChauvhan | TheJulia, Hi. | 13:19 |
| TheJulia | Hi, I'm on a board meeting right now, whats up | 13:19 |
| kamleshChauvhan | I think you missed to vote post review on patch https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/848709. | 13:19 |
| kamleshChauvhan | Can you please help to share vote on this patch once free.. | 13:19 |
| iurygregory | I was planing in ping Julia about this one (was reviewing the patch) | 13:41 |
| TheJulia | you guys realize the plugin is independently released right? | 13:43 |
| TheJulia | My plate is overflowing, fwiw. | 13:44 |
| iurygregory | yeah, I know =) | 13:45 |
| TheJulia | fwiw, I've been sporatically seeing gerrit not saving some things, I think it async http calls which don't complete from the browser | 13:48 |
| TheJulia | Had the same issue with a conference review tool as well :( Going too fast == stuff not getting saved | 13:51 |
| iurygregory | gerrit was a bit slow for me this days, and I had to re-login almost everyday | 14:02 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: Move bifrost integration tests to Ubuntu Jammy https://review.opendev.org/c/openstack/bifrost/+/848828 | 14:09 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: Move bifrost integration tests to Ubuntu Jammy https://review.opendev.org/c/openstack/bifrost/+/848828 | 14:12 |
| rpittau | this ^ should be ready now | 14:13 |
| TheJulia | I've also had internet connectivity blips where things cut out for a couple seconds or I get a spike of packet loss | 14:18 |
| iurygregory | rpittau, this would also mean we have Jammy as supported right? | 14:18 |
| rpittau | iurygregory: well, partially | 14:19 |
| iurygregory | we would need to update docs I think + releasenote | 14:19 |
| rpittau | need to convert the other jobs, but if that goes well should be quick | 14:19 |
| rpittau | yeah | 14:19 |
| iurygregory | let's hold a bit get some open patches merged and see if we can get this by eod today if goes green =) | 14:19 |
| rpittau | I would put the reno in the final patch though | 14:20 |
| iurygregory | yeah, but CI jobs we can change after release etc | 14:20 |
| rpittau | mmm not this one, bifrost is all on us | 14:21 |
| rpittau | ah nvm, read that wrong | 14:21 |
| TheJulia | "perfection is the enemy of good" | 14:21 |
| rpittau | looks like jammy will be officially proposed to replace focal in CI/CD during Antelope, so we're right in time | 14:42 |
| * dtantsur remembers to open IRC | 14:45 | |
| dtantsur | jm1: looking.. | 14:45 |
| dtantsur | jm1: I sense a bad timeout value somewhere, 3 minutes may not be enough | 14:46 |
| dtantsur | jm1: https://opendev.org/openstack/ansible-collections-openstack/src/branch/master/plugins/module_utils/openstack.py#L98 somehow takes priority over the value in module? | 14:52 |
| jm1 | dtantsur: reversed it to previous timeout of 30 minutes and indeed it passed after 5 minutes. now it fails in the next step, which is expected because we changed the return values https://f8bf2d4a845608449589-8f1f171af5507b30533733754e7f1ad8.ssl.cf5.rackcdn.com/854113/15/check/bifrost-collections-src/7187f5a/job-output.txt | 14:52 |
| dtantsur | jm1: yeah, 5 minutes is more realistic for cleaning on VMs (even more for bare metal) | 14:53 |
| jm1 | dtantsur: so your sense was good ;) | 14:53 |
| dtantsur | :) | 14:53 |
| jm1 | dtantsur: maybe you could change the role to check for both the old and the new return value name? | 14:54 |
| jm1 | dtantsur: then it should work with both the old and new aoc/sdk versions | 14:54 |
| dtantsur | jm1: what was the reason for rename? | 14:55 |
| dtantsur | this is not coming from the SDK, is it? | 14:55 |
| dtantsur | is there any sort of a convention? I'd expect baremetal_node_info to return baremetal_nodes? | 14:57 |
| jm1 | dtantsur: we are trying to enforce a consistent name scheme across all modules. in most modules this means dropping the prefixes, so that we are left with "nodes" or "servers" or whatever. a side effect is that users will get errors which will notify them about changed return values. which is good because module results of nearly all modules have changed in one or another way due to the sdk's cloud layer rewiring | 15:00 |
| jm1 | dtantsur: the idea is better warn users about changes than to let their code silently fail | 15:01 |
| dtantsur | I guess I'm a bit uneasy about result names not matching module names.. | 15:01 |
| dtantsur | from the bifrost's perspective, we'll need to change a dozen of occurrences. possible, but not ideal. | 15:02 |
| opendevreview | Merged openstack/ironic-tempest-plugin master: Swap maintenance fix to cleanup step https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/850524 | 15:04 |
| jm1 | dtantsur: module names is another thing we have to think about, but just had no time for. some modules have short names like server, others have longer names such as baremetal_node, then again others have the projectname inside their names such as keystone_* and then we have completely crazy names. | 15:04 |
| dtantsur | jm1: I'd keep baremetal_ prefix, at least because we have a conflict about ports | 15:04 |
| dtantsur | I'd definitely drop the usage of project codenames | 15:05 |
| TheJulia | Is anyone doing a blog post for Zed? | 15:05 |
| * TheJulia has word writing cap on thanks to board meeting | 15:05 | |
| jm1 | dtantsur: actually i agree with your approach of using longer names everywhere. for now i will rename the return value back to your previous name and we can rethink the module names (and its return value names) later | 15:06 |
| dtantsur | jm1: yeah, let's start with a formal consensus first | 15:06 |
| jm1 | dtantsur: you might have to fix bifrost anyway and without the name change you will have to discover the locations your self ;) | 15:08 |
| dtantsur | grep to the rescue \o/ | 15:08 |
| TheJulia | grep is always to the rescue | 15:11 |
| TheJulia | unfortunately it lacks an AI to self determine what needs to be found in *all* cases without input | 15:11 |
| * TheJulia assumes this statement launches the AI which eliminates humanity | 15:12 | |
| opendevreview | Verification of a change to openstack/ironic-tempest-plugin master failed: Add iDRAC Redfish sync boot mode test https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/848709 | 15:13 |
| opendevreview | Merged openstack/bifrost master: Rename include_dhcp_server -> enable_dhcp for consistency https://review.opendev.org/c/openstack/bifrost/+/858090 | 15:24 |
| rpittau | bye everyone, see you tomorrow! o/ | 15:25 |
| JayF | dtantsur: ty for the detailed information on use of bugfix/* branches on the list; it's nice to know the work backporting to them isn't going into a vacuum :D | 15:27 |
| dtantsur | sure thing :) | 15:31 |
| JayF | dtantsur: I will likely (in the next week or two) propose a small spec to cleanup our docs on our supported releases -- basically trying to make our docs line up with reality | 15:34 |
| JayF | I bet there are a ton of Ironic users who assume, wrongly, that we don't regularly backport stuff to ussuri and even at some points train | 15:34 |
| dtantsur | I'm one of them :D | 15:36 |
| JayF | Basically these days, if there's a backportable patch, I take it back to ussuri minimum unless the rebase starts getting too hairy | 15:37 |
| JayF | I got all our nova driver fixes from the last 2-3 years backported thru to train in nova as well | 15:37 |
| JayF | (FWIW this also means if I miss one, lmk and I'll put it in my queue) | 15:37 |
| JayF | This is why I wanted to make sure we actually use the stuff -- if we support it, I wanna do an A++ job of it, if we don't, I don't wanna do the job at all lol | 15:38 |
| TheJulia | So I just realized I forgot to mention I won't really be around Friday | 15:47 |
| TheJulia | . o O (not that most open source contributors *really* are most Fridays...) | 15:47 |
| JayF | Same here; it's my birthday so I took the day off. But as usual will probably have IRC open anyway :P | 15:48 |
| TheJulia | heh | 15:49 |
| * TheJulia is going to go meet up with some NC friends in LA on Friday | 15:50 | |
| dtantsur | JayF: note that we (= OpenShift) are not typical OpenStack consumers. We tend to consume newer releases than average. | 15:50 |
| JayF | dtantsur: good :D | 15:50 |
| TheJulia | dtantsur: fwiw, this is a pattern we hope to consume on my side of the house in the future (yay for strong compatibility contracts!) | 15:50 |
| dtantsur | \o/ | 15:51 |
| JayF | I was about to say, we've worked so hard on upgrade compatability | 15:51 |
| JayF | upgrades shouldn't be so scary anymore | 15:51 |
| TheJulia | no, yet there is still a huge stigma | 15:51 |
| TheJulia | and I think everyone who has been bitten over the past few years just flat out ignored release notes | 15:51 |
| TheJulia | at least, for those that I've talked to | 15:51 |
| * dtantsur -> driving school, see you tomorrow | 15:52 | |
| TheJulia | dtantsur: good luck! | 15:53 |
| * TheJulia needs to learn to drive the 45' mobile command center soon | 15:53 | |
| JayF | TheJulia: I told our downstream deployers here that I'm nearly 100% sure you could take an openstack installation of nearly any modern release, cut everything off, run migrations, upgrade code, cut everything on, and it would work *as long as* you didn't miss anything in the release/deprecation notes | 15:54 |
| JayF | TheJulia: (then told them not to do that because the statement after the "as long as" is impossible) | 15:54 |
| TheJulia | honestly, I think the vast majority of time I spend on release to release stuff is either things break in CI because a branch is x age, or folks have y code downstream with conditionals and weird logic which interacts with things in a way they want to have that we wouldn't advise | 15:55 |
| JayF | The upgrade pain at places I've worked have always been 1) business-process related (e.g. they won't eat the downtime or accept the risk) or 2) patches making upgrading be a bunch of rebase hell and refactoring workflows to "fit" new upstream features that replaced downstream patches | 15:57 |
| TheJulia | or features which people have created downstream and never bothered to upstream :( | 15:58 |
| TheJulia | I'm *really* surprised how much that happens | 15:58 |
| TheJulia | but... not eveyrone groks how to navigate Open Source | 15:58 |
| JayF | TheJulia: ...I have personally helped upstream downstream features that I'm nearly 100000% sure when they were written would've been rejected due to overly-strict project scoping or not-enough-review-time | 16:00 |
| TheJulia | And sometimes they are, and sometimes we would have been "cool" or "cool, but maybe not at this moment" | 16:01 |
| TheJulia | all a dialog which is needed | 16:01 |
| TheJulia | which relates to the current thing I'm writing actually | 16:01 |
| opendevreview | Merged openstack/python-ironicclient master: Fix wrong assertion methods https://review.opendev.org/c/openstack/python-ironicclient/+/858240 | 16:19 |
| opendevreview | Merged openstack/ironic master: Concurrent Distructive/Intensive ops limits https://review.opendev.org/c/openstack/ironic/+/840006 | 16:38 |
| * TheJulia dances | 16:39 | |
| JayF | nice | 16:40 |
| TheJulia | I'm really surprised how many of my "eh, I'll do this because it makes a lot of sense to just do since I'm in that area for other reasons" stuff landed this cycle | 16:55 |
| TheJulia | *sigh* https://github.com/OpenStackweb/ironic-website/pull/50 | 16:57 |
| TheJulia | JayF: what is your gh username? we need to ensure you have rights on that repo | 16:58 |
| JayF | TheJulia: jayofdoom | 16:59 |
| JayF | https://github.com/jayofdoom | 16:59 |
| JayF | I need to update my personal readme there (and on my blog) to reflect the last week of news | 16:59 |
| TheJulia | gah, I need to email someone | 17:03 |
| opendevreview | Merged openstack/ironic-tempest-plugin master: Add iDRAC Redfish sync boot mode test https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/848709 | 18:54 |
| iurygregory | TheJulia, do you mind if push the clean-up for ironic release notes in the same patch you pushed to add the prelude? | 19:07 |
| TheJulia | iurygregory: not at all | 19:08 |
| iurygregory | TheJulia, ack =) | 19:08 |
| TheJulia | le sign, netlify | 19:47 |
| TheJulia | err, le sigh | 19:47 |
| arne_wiebalck | JayF: was there any feedback on your bare metal SIG mail? I did not see anything on the list, I think. | 19:49 |
| JayF | arne_wiebalck: zero response whatsoever :( | 20:08 |
| arne_wiebalck | JayF: not unexpected | 20:09 |
| JayF | Eh, you're right but it's still sad | 20:09 |
| arne_wiebalck | JayF: we videos we did have a couple 100 views each so someone is watching them :) | 20:09 |
| arne_wiebalck | *the videos | 20:10 |
| JayF | getting folks to actively participate is hard | 20:10 |
| arne_wiebalck | yes | 20:10 |
| JayF | I'm going to start broadcasting some of my work via streams either in twitch or yt | 20:10 |
| JayF | we can see if ... less interactive ironic-adjacent videos do well | 20:10 |
| arne_wiebalck | ok, great | 20:10 |
| JayF | but I don't know what to do about the recurring sig. I'd suggest maybe keep the time, and do it even if we don't have a talk? | 20:10 |
| arne_wiebalck | the talk was the trigger to make people join | 20:11 |
| JayF | Yeah, you're likely right. | 20:11 |
| arne_wiebalck | there was very little else happening, but a talk sometimes triggered a discussion | 20:11 |
| JayF | We can stoop to bribery, I can give free stickers to attendees lol | 20:11 |
| arne_wiebalck | and the ops feedback sessions were good | 20:11 |
| arne_wiebalck | we had to add sessions | 20:11 |
| JayF | It's sorta an inherently difficult concept, right? | 20:11 |
| arne_wiebalck | or metal pins to speakers | 20:12 |
| JayF | operators are, almost everywhere, going to be overworked and understaffed | 20:12 |
| arne_wiebalck | yep | 20:12 |
| JayF | so getting them to participate is double-hard | 20:12 |
| JayF | I wonder if like, we could pitch it as a live helpdesk | 20:12 |
| arne_wiebalck | I mean it kind of worked to have one talk per month for a year or so | 20:12 |
| JayF | I mean, I'd be happy to spend that hour fielding questions about how to use ironic, how to solve problems, etc | 20:12 |
| JayF | arne_wiebalck: that's almost more talks/year than we have active contributors, it's just hard to sustain | 20:13 |
| JayF | maybe a quarterly or something? IDK | 20:13 |
| arne_wiebalck | yeah, good idea: "come and bring your question, no matter which level" | 20:13 |
| * JayF has to bail, he's got a meeting in 2 minutes | 20:13 | |
| JayF | exactly | 20:13 |
| JayF | I'd enjoy that as someone on the "helping" side of the fence, too | 20:13 |
| arne_wiebalck | yes ... ok, to be continued :) | 20:13 |
| JayF | I mean, I log IRC and am always watching | 20:14 |
| JayF | feel free to dump thoughts here or maybe even in a future-of-bm-sig etherpad | 20:14 |
| arne_wiebalck | ok! | 20:14 |
| stevebaker[m] | good morning | 20:41 |
| TheJulia | good morning stevebaker[m] | 21:19 |
| TheJulia | JayF: I wrapped the blog post and fixed the images.... so it should be good now. now to just get a volunteer for a zed post... :) | 21:42 |
| *** scottsol[m] is now known as scottsol | 21:57 | |
| TheJulia | oh my inspector's sqlalchemy stuffs is going to be "fun" | 22:26 |
| TheJulia | only 15 unit tests pass right now | 22:26 |
| TheJulia | mostly the query/interaction model, so hopefully not horrible | 22:27 |
| TheJulia | maybe... this *is* the time to migrate inspector into ironic | 22:46 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!