| stevebaker[m] | JayF: I've got a -1 on this but it would be nice to address them without blocking on TheJulia . What do you think? As a follow-up change or one of us makes the change and the other lands it? https://review.opendev.org/c/openstack/ironic/+/884783 | 00:56 |
|---|---|---|
| JayF | I'd honestly suggest option c; we land it as-is and one of us do the followup change real quick, none of it is so difficult as to wait and churning the patch makes more ci churn | 00:57 |
| TheJulia | For what it is worth, I started on doc updates in a later patch in the series | 01:07 |
| TheJulia | Didn’t make sense until the ipa change had also landed, at least to me | 01:08 |
| TheJulia | I should be around tomorrow, but I have several meetings on my calendar. | 01:14 |
| stevebaker[m] | Alrighty, +2! | 01:25 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/python-ironicclient master: Firmware Interface support https://review.opendev.org/c/openstack/python-ironicclient/+/891560 | 01:39 |
| iurygregory | JayF, fyi tested in real hardware and fixed the problems I've spoted | 01:40 |
| JayF | \o/ | 01:40 |
| iurygregory | Dmitry had a concern about showing created_at and updated_at without using --long (details) | 01:40 |
| iurygregory | but the current api doesn't have support, so I'm not 100% sure if how we want to handle =), not show at all ? | 01:41 |
| JayF | oh yeah, he's gone a while isn't he | 01:44 |
| iurygregory | yup | 01:44 |
| iurygregory | will be back next week | 01:44 |
| JayF | do we have pattern in other parts of the code of artifically making --long work differently using fields? | 01:44 |
| JayF | (and if we didn't hook up fields, that *is* weird) | 01:45 |
| iurygregory | long is basically used to list details | 01:45 |
| iurygregory | https://paste.opendev.org/show/bCH66JVId7IxaRlym48A/ | 01:45 |
| iurygregory | this is the current output from patch #2 | 01:45 |
| iurygregory | in patch #1 you would have two more columns, created at, updated at | 01:46 |
| JayF | Can you put all the context in the review and I'll take a look with fresh eyes in the morning? | 01:47 |
| JayF | I'm only looking at IRC on a fluke :) | 01:47 |
| iurygregory | or, we can ignore that the api doesn't have details support yet.. and if --long is used, we include created at and updated at | 01:47 |
| iurygregory | sure =) | 01:47 |
| JayF | I think that's not too bad of a way of doing it | 01:47 |
| JayF | because the pattern works for if we add more details to the api later | 01:47 |
| TheJulia | Wait… is the api already returning it? | 01:47 |
| JayF | iurygregory: TheJulia: this is all based on the assumption that the code in question is manipulating fields=blah properly on the request | 01:48 |
| JayF | if we have the data not returning it seems silly | 01:48 |
| JayF | and if we aren't supporting field querying that's sad too but probaly not really relevant in the real world | 01:48 |
| iurygregory | TheJulia, I'm testing the Redfish one in some machines :D | 01:49 |
| JayF | oh I misunderstood the question :) I'm less surprised to see Iury has it working :P | 01:49 |
| iurygregory | it's not no-firmware under test hehe | 01:49 |
| iurygregory | it's a real test :D | 01:50 |
| TheJulia | I guess I’m not I understanding the actual issue, so more details are needed | 01:50 |
| JayF | it was a comment on the review from dmitry; that's the context, can you link it up iurygregory ? | 01:50 |
| JayF | oh here it is, it was just in channel | 01:50 |
| JayF | https://review.opendev.org/c/openstack/python-ironicclient/+/891560/2#message-5c661d3ede3e0af5d63aed5627cf0cf4c2bcf8b5 | 01:50 |
| iurygregory | https://review.opendev.org/c/openstack/python-ironicclient/+/891560/2/ironicclient/tests/unit/osc/v1/test_baremetal_node.py | 01:51 |
| iurygregory | basically if we should include create at, updated at if `--long` is used, (we can have workaround in the code to make this happening, but the right thing is having the API supporting details, currently it doesn't) | 01:52 |
| TheJulia | Ugh | 01:53 |
| TheJulia | This has to micro-version, basically | 01:53 |
| iurygregory | yeah | 01:54 |
| JayF | microversions are cheap | 01:54 |
| TheJulia | Sorry :( | 01:54 |
| JayF | the time isn't though, sorry | 01:54 |
| iurygregory | true | 01:55 |
| JayF | if you can get it up I can get quick reviews? send me a direct email about it if you get one up before I'm awake | 01:55 |
| JayF | I'm disconnecting from irc now though o/ | 01:55 |
| iurygregory | JayF, I think I can work tomorrow morning on adding a new microversion for details | 01:55 |
| iurygregory | will try changing in the current setup I have ironic running | 01:56 |
| JayF | the real race is against the ironicclient deadline too | 01:56 |
| iurygregory | yeah | 01:56 |
| ashinclouds[m] | Perfection is impossible | 01:56 |
| JayF | so is there a path forward where we land the current support, add details as a fast follow up in the api and the client? | 01:57 |
| JayF | like, they aren't inherently incompatible? | 01:57 |
| TheJulia | Cut the fields out of the client, add them back in with proper micro version handling later | 01:57 |
| iurygregory | they are out the client already | 01:57 |
| iurygregory | in #2 | 01:57 |
| TheJulia | Technically it is a breaking change on the api, but there is no good option | 01:57 |
| TheJulia | I guess | 01:58 |
| TheJulia | The challenge is the real world, this is a feature, the delineation that is | 01:58 |
| JayF | it's never going to be released | 01:58 |
| JayF | that's the race for the follow up on the ironic side, before we cut a release | 01:58 |
| TheJulia | Do we really ever need that delineation | 01:59 |
| iurygregory | well, currently we do return created_at, updated at via api already | 01:59 |
| JayF | a minor api break, only in master for what, two weeks, on a brand new feature is not that bad | 01:59 |
| JayF | yeah so adding ?details and dropping them as default | 01:59 |
| JayF | is breaky | 01:59 |
| JayF | in theory | 01:59 |
| iurygregory | yeah | 01:59 |
| iurygregory | right | 01:59 |
| JayF | I say in theory because if someone is consuming that feature a week after we landed it they'd be in here talking to us now lol | 02:00 |
| TheJulia | Unreleased feature, in theory | 02:00 |
| JayF | as long as we document it in the microversion as a bugfix | 02:00 |
| TheJulia | Yeah | 02:00 |
| JayF | I don't care, we just need to make the space between now and then as short as possible | 02:00 |
| iurygregory | working on it | 02:05 |
| iurygregory | i.e microversion first, and will re-install and re-test the client | 02:06 |
| opendevreview | Merged openstack/ironic master: Adds service steps https://review.opendev.org/c/openstack/ironic/+/884783 | 02:56 |
| opendevreview | Merged openstack/ironic master: Update iRMC driver vendor library version https://review.opendev.org/c/openstack/ironic/+/892790 | 02:56 |
| vanou | Thanks JayF and stevebaker[m] for review | 03:10 |
| stevebaker[m] | you're welcome | 03:13 |
| iurygregory | humm interesting, for port we have created at and updated at (you don't need to use --long) | 03:32 |
| opendevreview | Merged openstack/ironic master: Introduce default kernel/ramdisks by arch https://review.opendev.org/c/openstack/ironic/+/890819 | 04:11 |
| opendevreview | Merged openstack/ironic master: Add service steps call to agent logic https://review.opendev.org/c/openstack/ironic/+/891229 | 04:11 |
| opendevreview | Merged openstack/ironic master: Revamp Ironic dev-quickstart documentation https://review.opendev.org/c/openstack/ironic/+/891912 | 04:11 |
| kubajj | Good morning everyone! o/ | 06:49 |
| rpittau | good morning ironic! o/ | 06:52 |
| rpittau | JayF: I'll be out for almost a month starting this Friday | 06:55 |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: Revert "Move metal3 integration job to non-voting temporarily" https://review.opendev.org/c/openstack/ironic/+/892797 | 08:40 |
| rpittau | yet another rebase, please merge this when you have a moment ^ | 08:41 |
| kubajj | rpittau: I don't want to recheck something that I don't understand, but it seems that neutron did not start again for https://review.opendev.org/c/openstack/networking-generic-switch/+/888051 the change seems relatively simple however and I doubt that it should influence neutron in any way | 09:51 |
| rpittau | kubajj: if you look at the neutron logs you'll see the error, it's quite explicit, looks like it's failing to load the netmiko plugin | 09:59 |
| rpittau | https://746119455d7fca2d93eb-21abb6d2b9f578210dfe07e5ee1d658a.ssl.cf2.rackcdn.com/888051/2/check/networking-generic-switch-tempest-dlm/ac74c9f/controller/logs/screen-q-svc.txt | 09:59 |
| kubajj | rpittau: Oh, I see. So the pipeline is failing? For future reference, can I get to these logs from the zuul site, or is this somewhere else? | 10:02 |
| kubajj | Found it here https://zuul.opendev.org/t/openstack/build/ac74c9f73cda4f0985e8e3044b1fc68d/log/controller/logs/screen-q-svc.txt | 10:03 |
| rpittau | kubajj: I believe something wrong in the enabled plugins, so not your patch fault, you can always access the logs of each job just clicking on the job link itself | 10:03 |
| rpittau | the zuul page is not the best place, the patch itself has everything you need | 10:03 |
| kubajj | rpittau: sorry if I am blind but which link do you mean? I can see only the zuul links on gerrit | 10:07 |
| rpittau | kubajj: if you link on the zuul message in your patch you'll see a list of jobs, each one with a link that you can open and where you will find the logs | 10:16 |
| kubajj | rpittau: oh, I see, so it is on the zuul page | 10:17 |
| kubajj | the link you've sent is just when you inspect logs raw | 10:17 |
| kubajj | thanks | 10:18 |
| rpittau | yeah | 10:18 |
| rpittau | np sorry for the confusion | 10:18 |
| iurygregory | good morning Ironic | 11:41 |
| iurygregory | rpittau, funny thing is that metal3 failed in https://review.opendev.org/c/openstack/ironic/+/892797 .-. | 11:42 |
| iurygregory | fatal: https://storage.googleapis.com/kubernetes-release/release/v1.28.1/bin/linux/amd64/kubectl binary checksum '\''e7a7d6f9d06fab38b4128785aa80f65c54f6675a0d2abef655259ddd852274e1'\'' differs from expected checksum '\''4685bfcf732260f72fce58379e812e091557ef1dfc1bc8084226c7891dd6028f' | 11:42 |
| iurygregory | oh god | 11:42 |
| iurygregory | .-. | 11:42 |
| opendevreview | Merged openstack/ironic master: Remove instance image after deployment https://review.opendev.org/c/openstack/ironic/+/878680 | 11:55 |
| rpittau | ah ok that's version issue | 12:06 |
| rpittau | let me check again on metal3 | 12:06 |
| iurygregory | TheJulia, JayF , so I was thinking a little bit, maybe we don't need to bump a microversion.. we already include all the information in the response see https://paste.opendev.org/show/bzY1vHVkCcsZ6xvDox1a/ , so maybe we just update the client (without --long we don't show created_at and updated_at, with --long we give all information), wdyt? | 12:09 |
| iurygregory | maybe is just me, but a microversion bump would be too much .-. | 12:16 |
| rpittau | iurygregory: issue on metal3 integration job is being fixed as we write :) | 12:30 |
| iurygregory | rpittau, nice =) | 12:30 |
| opendevreview | Merged openstack/ironic master: Permit Ironic to notify IPA it can support MD5 https://review.opendev.org/c/openstack/ironic/+/882168 | 12:32 |
| TheJulia | good morning | 13:06 |
| iurygregory | good morning TheJulia | 13:07 |
| TheJulia | Does anyone remember what I was working on last week?! | 13:12 |
| TheJulia | :) | 13:12 |
| iurygregory | service steps? | 13:14 |
| iurygregory | but I think the patch is merged | 13:14 |
| iurygregory | and I would suggest to take slow today =) | 13:14 |
| TheJulia | OVN, I think | 13:15 |
| TheJulia | yeah | 13:15 |
| TheJulia | .... I feel really tired | 13:15 |
| TheJulia | but I have energy, if that makes sense | 13:16 |
| TheJulia | coffee, then I'll re-test | 13:16 |
| TheJulia | so, looks like OVN doesn't work on the pxe-grub job for some reason | 13:19 |
| TheJulia | still positive /o\ | 13:26 |
| iurygregory | we want to move to OVN, but OVN doesn't like us lol | 13:35 |
| iurygregory | =( normally takes 4-5 days I think (last time I heard) | 13:37 |
| TheJulia | I suspect it is still lingering, unfortuntely | 13:52 |
| iurygregory | =( yeah, just take slow and get some rest | 13:54 |
| TheJulia | so the ovn grub failure appears to be it never seeks a config file | 13:56 |
| TheJulia | .... | 13:56 |
| iurygregory | <flip table> | 13:59 |
| TheJulia | it appears to have enough config, I'll have to pcap the vm to see what it is doing | 14:05 |
| TheJulia | baed upon that I can update docs on the ovn change | 14:05 |
| TheJulia | based | 14:07 |
| iurygregory | ++ | 14:35 |
| opendevreview | Mahnoor Asghar proposed openstack/ironic master: Add inspection hooks https://review.opendev.org/c/openstack/ironic/+/890817 | 14:58 |
| rpittau | good night! o/ | 16:12 |
| iurygregory | bye rpittau o/ | 16:17 |
| samcat116 | Is there any docs or a guide for using Bifrost without dhcp and relying on only redfish virtual media? | 16:22 |
| TheJulia | samcat116: afaik, no. I think you'd have to populate the network_info field on the node, which would get injeccted into the ramdisk | 16:24 |
| opendevreview | Merged openstack/ironic master: Add inspection (processing) hooks https://review.opendev.org/c/openstack/ironic/+/887554 | 16:45 |
| gmann | JayF: need your +1 on this, please check https://review.opendev.org/c/openstack/releases/+/885713 | 17:23 |
| JayF | ack | 17:23 |
| JayF | +1 | 17:23 |
| gmann | thanks | 17:24 |
| TheJulia | so! grub is not working with ovn because it thinks 10.1.0.34 is on the same broadcast domain as 10.208.225.67 (i.e.) the host | 17:29 |
| iurygregory | O.o | 17:32 |
| iurygregory | wth | 17:32 |
| clarkb | is it a /8? | 17:35 |
| clarkb | probably not but that would explain it at least | 17:35 |
| TheJulia | it is /28 | 17:38 |
| TheJulia | I think I need to open a bug and prod grub maintainers | 17:39 |
| TheJulia | "dude, this makes no sense... can I paypal you beer money" | 17:39 |
| JayF | it makes a lot of sense if it's an intentional lack of support for routed networks ... it's a lot of code to write to support routed networks and have a gateway | 17:43 |
| JayF | https://www.gnu.org/software/grub/manual/grub/html_node/Network.html note no settings for gateway or subnet | 17:44 |
| TheJulia | it worked previously.... | 17:44 |
| JayF | So it worked previously on a routed network | 17:45 |
| TheJulia | yes | 17:45 |
| JayF | then scratch all that, but it is suspicious that they don't even have the config language to talk about routing | 17:45 |
| JayF | is it possible it was doing the same wrong thing before and somehow (ovs magic?) it got to it's destination anyway? No way that's possible | 17:45 |
| JayF | because it'd be breaky on real hardware | 17:46 |
| JayF | yeah idk | 17:46 |
| JayF | TheJulia: this is interesting https://bugzilla.redhat.com/show_bug.cgi?id=1878612 | 17:46 |
| TheJulia | hmm | 17:48 |
| TheJulia | I am actually going to have to change the VM to be able to run some commands on it, but that was after I started typing out a bug | 17:48 |
| TheJulia | so yeah, somehow grub believes there is 10.0.0.0/8 as the route | 18:22 |
| *** melwitt_ is now known as melwitt | 18:34 | |
| iurygregory | TheJulia, JayF not sure if you saw the ping I left earlier regarding microversion bump to add details... currently this is what the API returns https://paste.opendev.org/show/bzY1vHVkCcsZ6xvDox1a/ I think we could just handle things in the client itself, https://paste.opendev.org/show/bcv7SjC4rOMO5EwFtzW3/ wdyt? | 19:53 |
| JayF | Your comment about port not supporting details was telling | 19:54 |
| JayF | if it mimics what we're already doing somewhere else, I'm OK with it | 19:54 |
| iurygregory | I don't think returning all of them without long would be a problem (since is not a lot of information regarding number of columns) | 19:57 |
| iurygregory | if we had more things stored in DB this would probably make more sense to support I would say | 19:58 |
| iurygregory | I just want to get the consensus so we can get the patch merged XD | 19:58 |
| JayF | The part I struggle with | 19:59 |
| JayF | and honestly I kinda hate being in this spot | 19:59 |
| JayF | is negative feedback without the person being here to clear it | 19:59 |
| JayF | I was OK with it the whole time | 19:59 |
| JayF | so it's hard for me to be like "yeah this satisfies dmitry's concern" | 19:59 |
| iurygregory | yeah =( | 19:59 |
| JayF | TheJulia: you have any opinion? | 19:59 |
| JayF | I sorta feel like if we get all three of us onboard it'll probably have to be fine | 19:59 |
| iurygregory | ++ | 19:59 |
| * iurygregory goes to grab more coffee brb | 20:01 | |
| TheJulia | can we talk thorough it for a few minutes? | 20:05 |
| TheJulia | it woule be easier for me to think through outloud | 20:05 |
| JayF | I'm happy to jump in a sync chat if that will help | 20:06 |
| * JayF writing up docs for his MLH fellow starting in a month-ish | 20:07 | |
| TheJulia | coffee is a good ida | 20:13 |
| JayF | I will turn into a pumpkin at some point, I've not been feeling great today and I'm reaching my end :) | 20:14 |
| JayF | so if we wanna chat lets do it in the next 30? | 20:14 |
| TheJulia | ++ | 20:14 |
| TheJulia | I need to self pumpkinize soon myself | 20:14 |
| TheJulia | ~1 hr and I'm a pumpkin pie | 20:15 |
| TheJulia | mmmm pie | 20:16 |
| JayF | iurygregory: you ready? | 20:18 |
| iurygregory | yeah | 20:24 |
| * JayF starts a zoom | 20:25 | |
| JayF | TheJulia: iurygregory: https://us06web.zoom.us/j/85464940784?pwd=bkVFcHg1bkZ6MmRXd1NmWTFEdTUzUT09 | 20:25 |
| JayF | anyone else interested in API semantics of node firmware please join :D | 20:25 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Add service steps and initial docs https://review.opendev.org/c/openstack/ironic/+/891630 | 20:25 |
| iurygregory | JayF, https://review.opendev.org/c/openstack/python-ironicclient/+/891560/2/ironicclient/tests/unit/osc/v1/test_baremetal_node.py | 20:34 |
| TheJulia | next! | 20:34 |
| iurygregory | you can probably add using reply and mark things as resolved | 20:34 |
| * iurygregory updates the patch | 20:34 | |
| JayF | I put a comment there | 20:36 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Enable OVN CI https://review.opendev.org/c/openstack/ironic/+/885087 | 20:43 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/python-ironicclient master: Firmware Interface support https://review.opendev.org/c/openstack/python-ironicclient/+/891560 | 20:44 |
| iurygregory | JayF, done =) | 20:44 |
| TheJulia | well, two integration jobs working with ovn is still not bad | 20:45 |
| iurygregory | better than none | 20:45 |
| TheJulia | and it is technically a feature I guess | 20:47 |
| TheJulia | albeit one where bugs were opened... many bugs | 20:47 |
| JayF | two jobs and a lot of bugs > 0 jobs and 0 bugs | 20:48 |
| TheJulia | yup | 20:51 |
| * iurygregory goes back to inspector client | 20:54 | |
| opendevreview | Julia Kreger proposed openstack/ironic master: CI: Fix our internal MTU settings https://review.opendev.org/c/openstack/ironic/+/893112 | 21:14 |
| opendevreview | Merged openstack/ironic master: Revert "Move metal3 integration job to non-voting temporarily" https://review.opendev.org/c/openstack/ironic/+/892797 | 21:16 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!