| *** redrobot4 is now known as redrobot | 04:52 | |
| iurygregory | good morning janders and Ironic o/ | 05:26 |
|---|---|---|
| janders | hey iurygregory o/ | 05:38 |
| arne_wiebalck | Good morning janders iurygregory and Ironic! | 06:59 |
| janders | hey arne_wiebalck | 06:59 |
| *** rpittau|afk is now known as rpittau | 06:59 | |
| rpittau | good morning ironic! o/ | 07:00 |
| iurygregory | morning arne_wiebalck rpittau o/ | 07:00 |
| rpittau | hey iurygregory :) | 07:00 |
| janders | hey rpittau | 07:01 |
| rpittau | hey janders :) | 07:01 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: [WIP] Force selinux presence when testing https://review.opendev.org/c/openstack/bifrost/+/799965 | 07:19 |
| iurygregory | selinux strikes again? :D | 07:24 |
| rpittau | nothing bad this time, it's just we expect to have that enforced but we don't really force to have it :) | 07:54 |
| rpittau | also a way to test the CI | 07:54 |
| iurygregory | nice | 07:56 |
| arne_wiebalck | For long running cleaning steps (like network burn-in, 6 hours), I am seeing SSL timeouts in eventlet, nothing fatal, though. It seems that there are known issues with certain combinations of Python and eventlet ... do we have any knowledge about working combinations? From what I see we do not explicitly set the version of eventlet. (This is all on Victoria.) | 08:11 |
| iurygregory | 6hrs? | 08:16 |
| iurygregory | what? | 08:16 |
| iurygregory | <O> | 08:16 |
| iurygregory | I'm not aware of combinations that would work =( | 08:17 |
| arne_wiebalck | iurygregory: yep, 6 hrs read, 6 hrs write is the default we use, see also https://review.opendev.org/c/openstack/ironic-python-agent/+/799687 | 08:19 |
| iurygregory | fio <3 | 08:20 |
| * iurygregory goes back to 2015 - Master's Degree | 08:20 | |
| arne_wiebalck | I let it run first time tonight ... 12 hrs of sustained 10Gbit traffic between my test nodes, they held up :) | 08:21 |
| arne_wiebalck | I guess the cable got a little warm .. | 08:22 |
| iurygregory | IT will contact you in a few hours :D | 08:22 |
| arne_wiebalck | probably :-D | 08:23 |
| *** sshnaidm_ is now known as sshnaidm | 08:25 | |
| opendevreview | Verification of a change to openstack/bifrost failed: Fix Redeploy Playbook https://review.opendev.org/c/openstack/bifrost/+/798079 | 08:41 |
| rpittau | ok, same error, but I can't reproduce it locally | 08:45 |
| rpittau | maybe the CI host is not upgraded with latest pkgs | 08:46 |
| dtantsur | mmmmm...morning? | 09:00 |
| janders | hey dtantsur o/ | 09:01 |
| rpittau | morning dtantsur :) | 09:02 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: [WIP] Force selinux presence when testing https://review.opendev.org/c/openstack/bifrost/+/799965 | 09:04 |
| rpittau | I already have a headache for the vaccine, the bifrost CI issue is not helping | 09:05 |
| janders | rpittau sorry to hear - and take care! | 09:09 |
| iurygregory | morning dtantsur | 09:11 |
| iurygregory | rpittau, take care =) | 09:11 |
| rpittau | thanks! let's see how long I last | 09:13 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: Collect explicit list of installed packages https://review.opendev.org/c/openstack/bifrost/+/799975 | 09:15 |
| opendevreview | Jacob Anders proposed openstack/ironic master: [WIP] Add support for verify steps https://review.opendev.org/c/openstack/ironic/+/800001 | 09:55 |
| janders | ^ dammit, off-by-one! | 09:56 |
| janders | not sure who got the ID of 800000 but it doesn't look like it was Ironic | 09:58 |
| rpittau | janders: that was me :) | 10:06 |
| rpittau | but not in ironic | 10:06 |
| janders | rpittau nice! :) | 10:06 |
| opendevreview | Jacob Anders proposed openstack/ironic master: [WIP] Add support for verify steps https://review.opendev.org/c/openstack/ironic/+/800001 | 10:54 |
| opendevreview | Jacob Anders proposed openstack/ironic master: [WIP] Add support for verify steps https://review.opendev.org/c/openstack/ironic/+/800001 | 10:59 |
| opendevreview | Merged openstack/ironic stable/wallaby: Avoid double file removal in create_boot_iso https://review.opendev.org/c/openstack/ironic/+/799774 | 11:27 |
| dtantsur | cenne: hi! not sure if you've noticed that: https://review.opendev.org/c/openstack/ironic/+/797055 ended up in a merge conflict, unfortunately, you need to rebase it | 11:27 |
| janders | see you tomorow Ironic o/ | 11:56 |
| rpittau | bye everyone, see you tomorrow o/ | 12:45 |
| *** rpittau is now known as rpittau|afk | 12:45 | |
| cenne | dtantsur: doing rebase now again. | 12:59 |
| TheJulia | Good morning | 13:18 |
| opendevreview | cenne proposed openstack/ironic master: Add `boot_mode` and `secure_boot` to node object and expose in api https://review.opendev.org/c/openstack/ironic/+/797055 | 13:21 |
| cenne | Good morning TheJulia | 13:21 |
| * TheJulia feels like she didn't get any sleep | 13:22 | |
| iurygregory | good morning TheJulia | 13:25 |
| dtantsur | TheJulia: morning. same :( | 13:26 |
| iurygregory | dtantsur, I was checking the schemas https://redfish.dmtf.org/schemas/v1/EventDestination.v1_5_0.json https://redfish.dmtf.org/schemas/v1/EventDestination.v1_0_6.json Protocol is only Redfish, I will remove from the spec the requirement to require this information in the API request | 13:33 |
| dtantsur | iurygregory++ | 13:33 |
| dtantsur | it's easier to add something later than to remove it | 13:33 |
| iurygregory | about extra I have no idea if we should support | 13:33 |
| iurygregory | do we have any use cases? .-. | 13:34 |
| * iurygregory is also thinking about not includding http_headers tbh =X | 13:34 | |
| dtantsur | extra is for operator-attached information. I think all our top-level resources support it? | 13:34 |
| dtantsur | I only suggest it for consistency | 13:34 |
| iurygregory | ack | 13:34 |
| dtantsur | re http_headers: we need to research how well they are supported | 13:34 |
| dtantsur | maybe add them as a later addition? | 13:34 |
| iurygregory | yeah | 13:34 |
| dtantsur | somebody has told me that their hardware ignores the provided headers | 13:35 |
| iurygregory | I wouldn't be surprised =) | 13:35 |
| iurygregory | for the Context I will just add the description provided by the redfish schema | 13:36 |
| iurygregory | funny that can be null but it's required :D | 13:36 |
| dtantsur | Oo | 13:36 |
| dtantsur | what's it even for? do we need it? | 13:37 |
| dtantsur | is it somehow supplied as part of notifications? | 13:37 |
| dtantsur | re EventTypes: do we need to support EventFormatType instead? (probably also as a follow-up) | 13:37 |
| iurygregory | well we will consider default to Alert | 13:38 |
| iurygregory | since is supported by most hardware vendors | 13:38 |
| iurygregory | because in 1.5.0 they deprecated EventTypes | 13:39 |
| dtantsur | can we just skip it? | 13:39 |
| iurygregory | not if BMC is using an old version | 13:39 |
| dtantsur | and what about SubscriptionType? just use RedfishEvent | 13:39 |
| dtantsur | omg | 13:39 |
| iurygregory | we need to send becuase is required | 13:39 |
| * TheJulia drinks more coffee in hopes of waking up | 13:40 | |
| dtantsur | oooh, so wait. EventTypes were required in old versions. SubscriptionType is required now. | 13:40 |
| dtantsur | how are we even going to manage that? (this should be a chapter in the spec IMO) | 13:40 |
| iurygregory | all hardware I've tested doesn't support the newer versions... | 13:41 |
| * dtantsur likes how these objects have different "required" and "requiredOnCreate"... | 13:41 | |
| iurygregory | yeah I also like this ^ | 13:41 |
| dtantsur | btw Context is required, not requiredOnCreate :) | 13:41 |
| dtantsur | iurygregory: could you also write some words on how you're going to handle it for Redfish specifically? | 13:41 |
| iurygregory | yeah and it can be null =) | 13:41 |
| dtantsur | I understand that the spec is generic, but I think it's useful to discuss this aspect | 13:42 |
| iurygregory | not sure if I "how we are going to handle for Redfish specifically" .. we have the support in sushy (is considering 1.0.6 version) | 13:42 |
| dtantsur | some implementation details. what exactly you're going to send given that required fields are so different. | 13:43 |
| dtantsur | pretty much a detailed version of this conversation | 13:43 |
| iurygregory | In what section should I provide this information? | 13:44 |
| iurygregory | I don't see where it would fit .-. sorry | 13:44 |
| dtantsur | as a new section in "Proposed changes"? something like "Redfish implementation details" | 13:45 |
| dtantsur | you cannot remove sections, but you can add more :) | 13:45 |
| iurygregory | ohhhhhhhhhhhh | 13:45 |
| iurygregory | I wasn't aware of "you can add more" =) | 13:45 |
| dtantsur | :) | 13:45 |
| iurygregory | tks for the discussion dtantsur =) | 13:45 |
| dtantsur | sure, you're welcome. thank you for diving into this topic. | 13:46 |
| * dtantsur didn't expect it to turn out so problematic when he wrote the first version of the RFE... | 13:46 | |
| iurygregory | quick question... since we have changes in ManagementInterface, the endpoint can still be /v1/nodes/<node_id>/subscriptions or has to be /v1/nodes/{node_ident}/management/subscriptions ? | 13:47 |
| TheJulia | joy :( | 13:57 |
| iurygregory | just wondering if we need to keep this pattern or not :D hehe | 13:57 |
| iurygregory | like we have indicators API under management (sad they don't need to store info in the DB hehe) | 13:58 |
| TheJulia | eh, db objects are kind of easy when you have a correlating node_id :) | 13:59 |
| dtantsur | I wonder if we want to make it top-level | 14:01 |
| dtantsur | (I came to this idea after reviewing the CLI design in the spec) | 14:01 |
| dtantsur | and I'd personally kill the /management/ part if I could | 14:02 |
| * TheJulia shrugs | 14:03 | |
| TheJulia | pretty much requires the same db interaction, and not having it in nodes makes it easier to code | 14:03 |
| TheJulia | at least, I think | 14:03 |
| TheJulia | the nodes controller is a bit narly as-is | 14:03 |
| iurygregory | top level would be the one I have in the spec or you mean /v1/subscriptions? .-. | 14:03 |
| TheJulia | yes | 14:04 |
| dtantsur | thinking /v1/subscriptions | 14:04 |
| iurygregory | hummm | 14:04 |
| dtantsur | similar to ports which also don't exist outside of nodes really | 14:04 |
| TheJulia | I was thinking exactly that | 14:04 |
| iurygregory | so we can require the node_uuid to match things | 14:04 |
| dtantsur | to me GET /v1/subscriptions also makes sense | 14:04 |
| TheJulia | iurygregory: yes | 14:05 |
| iurygregory | I like the idea | 14:05 |
| dtantsur | "list all subscriptions in the system" | 14:05 |
| dtantsur | similar to how you can list all ports, not only ports of a node | 14:05 |
| TheJulia | that *does* complicate it, I'll just ask for a few simple rbac tests so we know a non-owner/lessee can't touch subscriptions | 14:05 |
| TheJulia | at least, for a node | 14:05 |
| dtantsur | I suspect it can be cargo-culted from ports | 14:05 |
| TheJulia | yeah, it should all be able to be | 14:06 |
| iurygregory | TheJulia, cool cool cool =) | 14:06 |
| TheJulia | good morning kkillsfirst | 14:28 |
| kkillsfirst | Good Morning TheJulia. | 14:29 |
| trandles | Hello ironic o/ | 14:51 |
| TheJulia | good morning trandles | 15:01 |
| opendevreview | Aija Jauntēva proposed x/sushy-oem-idrac master: Add RAID mode conversion methods https://review.opendev.org/c/x/sushy-oem-idrac/+/799506 | 15:20 |
| opendevreview | Aija Jauntēva proposed x/sushy-oem-idrac master: Add RAID clear foreign config method https://review.opendev.org/c/x/sushy-oem-idrac/+/800071 | 15:20 |
| opendevreview | Aija Jauntēva proposed openstack/ironic master: Update iDRAC virtual media boot known issue https://review.opendev.org/c/openstack/ironic/+/800073 | 16:00 |
| JayF | this has 3x +2 and no +A, cna someone land it? https://review.opendev.org/c/openstack/ironic/+/799880 | 16:15 |
| cenne | dtantsur: JayF: can you just check and re-approve https://review.opendev.org/c/openstack/ironic/+/797055 | 16:46 |
| JayF | lgtm, +2 | 16:47 |
| cenne | (new, small) Changes made are to resolve the merge conflict. | 16:47 |
| JayF | yeah; looks like you had to rebase against julia's changes | 16:48 |
| cenne | yup. | 16:49 |
| arne_wiebalck | bye everyone o/ | 16:53 |
| cenne | bye arne_wiebalck o/ | 16:53 |
| cenne | thanks JayF | 16:53 |
| JayF | o/ have a good ome arne_wiebalck | 16:53 |
| opendevreview | cenne proposed openstack/ironic master: [WIP] Add endpoint for setting node boot_mode on demand https://review.opendev.org/c/openstack/ironic/+/800084 | 16:58 |
| opendevreview | cenne proposed openstack/ironic master: [WIP] Add endpoint for setting node boot_mode on demand https://review.opendev.org/c/openstack/ironic/+/800084 | 18:05 |
| opendevreview | Merged openstack/ironic stable/wallaby: Only return the requested fields from the DB https://review.opendev.org/c/openstack/ironic/+/799880 | 18:41 |
| opendevreview | Merged openstack/ironic stable/wallaby: Set stage for objects to handle selected field lists. https://review.opendev.org/c/openstack/ironic/+/799904 | 18:41 |
| opendevreview | Merged openstack/ironic stable/wallaby: API to pass fields to node object list https://review.opendev.org/c/openstack/ironic/+/799905 | 18:41 |
| opendevreview | Merged openstack/ironic stable/wallaby: Allow node_sanitize function to be provided overrides https://review.opendev.org/c/openstack/ironic/+/799906 | 18:46 |
| opendevreview | Merged openstack/ironic stable/wallaby: Use selectinload for all list queries https://review.opendev.org/c/openstack/ironic/+/799907 | 18:46 |
| JayF | Hmm. No way those are going to apply cleanly to Victoria, because it's missing RBAC, right? Wondering if Wallaby is as far back as those backports should go | 19:22 |
| iurygregory | yeah | 19:41 |
| iurygregory | TheJulia, for the subscriptions api I need to create new policy Rules, so it will only work if using RBAC and not the legacy one? | 19:45 |
| iurygregory | something like http://paste.openstack.org/show/807303/ | 19:51 |
| TheJulia | JayF: uhh... smight be a little more difficult but logically it should work. The policy/sanitization stuff need some changes | 20:50 |
| *** mgoddard- is now known as mgoddard | 20:50 | |
| TheJulia | iurygregory: would there any information in the subscription which would be sensitive if reviewed? | 20:51 |
| opendevreview | Merged openstack/ironic master: Add `boot_mode` and `secure_boot` to node object and expose in api https://review.opendev.org/c/openstack/ironic/+/797055 | 20:59 |
| stevebaker | good morning | 21:02 |
| iurygregory | TheJulia, to me destination would be a sensitive information (but should be ok for admin/operators to look at that in RBAC this would be system scope only) | 21:37 |
| TheJulia | iurygregory: so all project readers would be a bad idea then | 21:49 |
| TheJulia | I think there is a SYSTEM_READER_OR_OWNER_READER alias | 21:50 |
| iurygregory | TheJulia, ack I will look at it o/ | 22:09 |
| * iurygregory time to sleep zzz | 22:09 | |
| TheJulia | iurygregory: or make a secondary policy which is admin readable/otherwise redacted. | 22:10 |
| janders | good morning Ironic o/ | 23:27 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!