| opendevreview | Takashi Kajinami proposed openstack/python-ironic-inspector-client master: setup.cfg: Replace dashes by underscores https://review.opendev.org/c/openstack/python-ironic-inspector-client/+/828081 | 07:48 |
|---|---|---|
| arne_wiebalck | Good morning, Ironic! | 07:48 |
| rpittau | good morning ironic! o/ | 08:22 |
| janders | good morning arne_wiebalck rpittau and Ironic o/ | 09:14 |
| rpittau | hey janders :) | 09:15 |
| arne_wiebalck | hey janders o/ | 09:50 |
| iurygregory | good morning Ironic o/ | 11:06 |
| arne_wiebalck | hey iurygregory o/ | 11:29 |
| iurygregory | arne_wiebalck, o// | 11:39 |
| janders | see you tomorrow Ironic o/ | 12:59 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: CI: use a custom debian partition image instead of cirros https://review.opendev.org/c/openstack/ironic/+/827882 | 13:12 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [DNM] Testing the CI https://review.opendev.org/c/openstack/ironic/+/827500 | 13:12 |
| * TheJulia wipes sleep from her eyes | 13:37 | |
| arne_wiebalck | Hi TheJulia o/ | 13:47 |
| arne_wiebalck | TheJulia: Once you have had your first coffee, let me know: I tried your inspector patch. | 13:47 |
| parasiti1 | hi, i have a question during the very first enrollment of a host, when it's pxe netif mac address is not even known: is there a way to setup ironic in a way that if a pxe boot presents an unknown mac address, it receives the IPA by default so we can inspect it ? | 13:53 |
| TheJulia | arne_wiebalck: First coffee, and starting to wake up | 14:03 |
| arne_wiebalck | TheJulia: sure, no rush :) | 14:07 |
| arne_wiebalck | parasiti1: your question is if it is feasible to configure Ironic's PXE to give the IPA image to unknown nodes or if you can inspect nodes which are not previously known to Ironic (or both) ? | 14:08 |
| TheJulia | arne_wiebalck: ironic... not *entire* possible for *ironic* | 14:12 |
| TheJulia | but, your base ipxe *can* be modified | 14:12 |
| TheJulia | so it falls back | 14:12 |
| TheJulia | Bifrost and I believe metal3 do this for fallback to discovery/inspection | 14:13 |
| TheJulia | inspecting unknown nodes == discovery | 14:13 |
| arne_wiebalck | TheJulia: we do this in our PXE infra as well | 14:13 |
| parasiti1 | arne_wiebalck, TheJulia : i can tweak the ipxe template to fall back to some special "IPA in discovery mode" image correct ? that could be nice | 14:42 |
| parasiti1 | could that image "auto enroll hosts" or is this a generally bad idea ? | 14:44 |
| TheJulia | parasiti1: inspector has configuration for this, a node_not_found hook | 14:45 |
| TheJulia | which *can* enroll the node | 14:45 |
| arne_wiebalck | parasiti1: you need to configure the ironic/inspector to allow for the discovery of new nodes: https://docs.openstack.org/ironic-inspector/latest/user/usage.html#discovery | 14:45 |
| TheJulia | the conundrum is the BMC password | 14:45 |
| TheJulia | well, credentials in general | 14:45 |
| iurygregory | good morning TheJulia =) | 14:46 |
| TheJulia | and not all BMC's can be discovered via in-band discovery, especially if IPMI is locked down. So that may add some complication, but most operators I know have done this either have a pre-baked singular credentials they load in with introspection rules, OR they discover the machine, and back-fill the credentials once the node is known by Ironic. | 14:46 |
| *** Guest1862 is now known as dansmith | 14:47 | |
| rpittau | test with tinycore 13.x went quite well -> https://review.opendev.org/c/openstack/ironic-python-agent/+/827664 | 14:48 |
| parasiti1 | TheJulia: ok thanks, that's what i was looking for | 14:48 |
| opendevreview | Merged openstack/python-ironic-inspector-client master: setup.cfg: Replace dashes by underscores https://review.opendev.org/c/openstack/python-ironic-inspector-client/+/828081 | 15:00 |
| TheJulia | arne_wiebalck: so out of curiosity, did you have luck with your load balancers? | 15:01 |
| arne_wiebalck | TheJulia: no, same issue as before | 15:01 |
| TheJulia | :( | 15:01 |
| TheJulia | what sort of load balancer is this? | 15:02 |
| TheJulia | any exceptions? | 15:02 |
| arne_wiebalck | ha-proxy | 15:21 |
| arne_wiebalck | I have not checked yet, will do when I get a moment and let you know! | 15:21 |
| TheJulia | Interesting, becaues that resolved out haproxy issues... | 15:22 |
| arne_wiebalck | wasn't there also a managed vs non-managed distinction in the mix? | 15:28 |
| *** dansmith is now known as Guest2102 | 15:29 | |
| TheJulia | yes | 15:30 |
| TheJulia | there is a lock in the non-managed code path | 15:31 |
| TheJulia | .. or is it managed | 15:31 |
| TheJulia | I don't remember | 15:31 |
| *** Guest2102 is now known as dansmith | 15:37 | |
| *** akahat|rover is now known as akahat|dinner | 15:43 | |
| arne_wiebalck | I will try to have a look. | 15:46 |
| * iurygregory almost lost track of time while having lunch | 15:56 | |
| iurygregory | #startmeeting ironic | 15:59 |
| opendevmeet | Meeting started Mon Feb 7 15:59:59 2022 UTC and is due to finish in 60 minutes. The chair is iurygregory. Information about MeetBot at http://wiki.debian.org/MeetBot. | 15:59 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 15:59 |
| opendevmeet | The meeting name has been set to 'ironic' | 15:59 |
| iurygregory | o/ | 16:00 |
| ajya | o/ | 16:00 |
| iurygregory | Hello everyone, welcome to our weekly meeting! | 16:00 |
| rpioso | o/ | 16:00 |
| rpittau | o/ | 16:00 |
| rloo | o/ | 16:00 |
| arne_wiebalck | o/ | 16:00 |
| ameya49 | o/ | 16:00 |
| iurygregory | seems like we have quorum =) | 16:01 |
| iurygregory | you can find the agenda for our meeting in the wiki | 16:01 |
| iurygregory | #link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting | 16:01 |
| TheJulia | o/ | 16:02 |
| iurygregory | #topic Announcements / Reminders | 16:02 |
| iurygregory | #info CFP for OIS Berlin 2022 closes in 2 days (Feb 09)! | 16:02 |
| iurygregory | #link https://cfp.openinfra.dev/app/berlin-2022/ | 16:02 |
| erbarr | o/ | 16:02 |
| iurygregory | Submit all the talks \o/ | 16:02 |
| TheJulia | Including the crazy idea talks! | 16:03 |
| TheJulia | Just no pet rocks. :) | 16:03 |
| iurygregory | crazy ideas ++ :D | 16:03 |
| iurygregory | #info Baremetal SIG - Scaling Ironic by TheJulia Feb 8, 2022, at 2pm UTC on zoom. | 16:03 |
| iurygregory | #link https://etherpad.opendev.org/p/bare-metal-sig | 16:03 |
| iurygregory | Tomorrow is our monthly presentation from the Baremetal SIG =D | 16:04 |
| iurygregory | #info PTG April 4th-8th | 16:05 |
| iurygregory | #link http://lists.openstack.org/pipermail/openstack-discuss/2022-February/027051.html | 16:05 |
| iurygregory | I've started looking at the slots for the PTG, I will share the idea for the slots in the ML during this week | 16:06 |
| iurygregory | #info Combined PTL/TC Election Feb-Mar 2022 Season (Z Cycle) | 16:06 |
| iurygregory | #link http://lists.openstack.org/pipermail/openstack-discuss/2022-February/027070.html | 16:06 |
| iurygregory | Seems like is already this time of the year ... elections :D | 16:07 |
| TheJulia | Where did the last three months go?!? | 16:08 |
| iurygregory | time flies :D | 16:08 |
| iurygregory | #info Second release for the projects this week (Yoga) | 16:09 |
| iurygregory | we had some problems last week =) | 16:10 |
| iurygregory | I'm going to check the releasenotes and see if need any urgent fix to get merged before pushing the release | 16:10 |
| iurygregory | #action iurygregory check releasenotes and push release patches | 16:11 |
| iurygregory | Does anyone have anything to announce or remind us of this week? | 16:11 |
| TheJulia | iurygregory: feel free to /msg me with reno patches to review | 16:11 |
| iurygregory | TheJulia, ack | 16:11 |
| rpittau | iurygregory: I can help with the releases if needed | 16:11 |
| iurygregory | rpittau, ack (I hope I will be able to fix my gerrit problem today...) | 16:12 |
| iurygregory | tks :D | 16:12 |
| iurygregory | #topic Review Action Items for last week | 16:12 |
| iurygregory | no action items from last week | 16:12 |
| iurygregory | #topic Review subteam status reports | 16:13 |
| iurygregory | #link https://etherpad.opendev.org/p/IronicWhiteBoard | 16:13 |
| iurygregory | starting around L62 | 16:13 |
| iurygregory | I think we can move on | 16:19 |
| iurygregory | #topic Deciding on priorities for the coming week | 16:20 |
| iurygregory | #link https://review.opendev.org/q/status:open+hashtag:ironic-week-prio | 16:20 |
| ameya49 | Can I please have feedback on : https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/826646 | 16:21 |
| ameya49 | Zuul is passing and have added priority tag as well | 16:21 |
| ajya | can these 2 be added https://review.opendev.org/c/openstack/ironic/+/825489 https://review.opendev.org/c/openstack/ironic/+/825753 ? | 16:22 |
| iurygregory | ameya49, sure since you added the tag it will show in the list and we will review =) | 16:22 |
| ameya49 | iurygregory: Thanks :) | 16:23 |
| iurygregory | ajya, I'm ok with adding them, quick question since they are features, you don't need newer version for dracclient ? | 16:24 |
| rpittau | I will add the tinycore 13.x upgrade, seems to work quite well | 16:24 |
| iurygregory | rpittau, nice! | 16:24 |
| ajya | iurygregory: no, the feature does not rely on dracclient | 16:25 |
| iurygregory | ajya, ack I just wanted to double check :D | 16:25 |
| iurygregory | feel free to add the hashtag | 16:25 |
| ajya | ok, thank you | 16:25 |
| iurygregory | anything else? =) | 16:28 |
| TheJulia | nothing from me at this moment | 16:28 |
| rpittau | one more in sushy-tools, I will add the hash | 16:28 |
| dtantsur | my CI rework is not ready, I think | 16:28 |
| iurygregory | rpittau, ack np | 16:29 |
| dtantsur | and we probably don't want to start merging it before the release :) | 16:29 |
| iurygregory | woot dtantsur is around | 16:29 |
| dtantsur | yeah, I'm lurking :) | 16:29 |
| rpittau | done, thanks | 16:29 |
| iurygregory | dtantsur, yeah I agree | 16:29 |
| iurygregory | ok, moving on | 16:29 |
| iurygregory | #topic Discussion | 16:29 |
| iurygregory | we have one topic to discuss \o/ | 16:30 |
| iurygregory | #info Running InspectorBasicTest in a 3rd Party CI | 16:30 |
| iurygregory | rpioso or ajya do you want to go and give context? | 16:30 |
| rpioso | Dell is working on improving the test coverage in its third-party CI. | 16:31 |
| rpioso | We are developing Tempest tests to automate much of the testing we have performed manually. | 16:31 |
| rpioso | Those will benefit our third-party CI, regression testing of new server firmware versions, qualification of new server models, and early regression of pre-release firmware. | 16:32 |
| TheJulia | Awesome! | 16:33 |
| rpioso | We are aiming to leverage as much of the existing tests as possible. | 16:33 |
| rpioso | The team is new to Tempest :-) | 16:33 |
| TheJulia | Is the plan to contribute this back to ironic-tempest-plugin? | 16:33 |
| TheJulia | tempest is.. unique. :) | 16:34 |
| rpioso | iurygregory: Thank you so much for giving me the opportunity to provide context. | 16:34 |
| iurygregory | rpioso, np | 16:34 |
| rpioso | TheJulia: Yes, we plan to contribute it upstream :-) | 16:34 |
| TheJulia | awesome! You *might* want to have a specific class for these, fwiw. | 16:35 |
| iurygregory | at the moment they are having issues running the inspector tests because we try to assert the amount of ram the server has | 16:35 |
| rpioso | TheJulia: ameya49 's contribution is the first of several changes we plan to contribute: https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/826646. | 16:36 |
| iurygregory | https://github.com/openstack/ironic-tempest-plugin/blob/master/ironic_tempest_plugin/tests/scenario/test_introspection_basic.py#L27-L29 | 16:36 |
| iurygregory | would it make sense to have a specific class for the introspection test also? I think we only run on CI and we have some infor that is *hardcoded* like the flavor we use in the tests | 16:37 |
| TheJulia | So, anyone know of a page covering copyright stamping off-hand in OpenStack, because stampign for a couple line change has generaly been frowned upon in the past... also since it is pointless and redundant given all rights get assigned and are not actually retained AIUI | 16:39 |
| iurygregory | I've talked with Dell folks last week, they will have X machines and they want tempest to run the tests on all the machines, the introspection will fail because it won't match the amount of memory , how we can help them do be able to run this tests (I don't think that for Introspection we would need to create a new class) | 16:40 |
| iurygregory | I see a few options on how we can help them | 16:41 |
| TheJulia | a specific test would likely need to understand it's role in consuming all of the nodes. Ultimately working all the way through nova and scheduling may not be ideal given that is a number of layers of abstraction when the goal sounds like it is regression and qualification of new hardware models. | 16:42 |
| rpioso | Do those assertions against the flavor ^^^ apply, since, AFAIU, they are not used by nova? See the 2d paragraph on https://docs.openstack.org/ironic/latest/install/configure-nova-flavors.html. Perhaps they could be relaxed to check that they are simply greater than zero (0). | 16:42 |
| TheJulia | at which point, memory flavor matching is pointless | 16:42 |
| TheJulia | They really arn't used for scheduling anymore | 16:44 |
| TheJulia | so relaxed could work | 16:44 |
| iurygregory | I like the idea | 16:45 |
| *** dansmith is now known as Guest2108 | 16:45 | |
| *** Guest2108 is now known as dansmith | 16:46 | |
| rpioso | TheJulia: That would be cool! The existing test otherwise seems to fit our needs. We may derive classes from it to ensure iDRAC-specific interfaces are configured on the baremetal node. | 16:46 |
| rpioso | *nodes :-) | 16:46 |
| TheJulia | derived class would be highly preferred, fwiw | 16:46 |
| rpioso | Speaking of which, why are the inspector tests the only ones that deal with all of the available baremetal nodes? | 16:47 |
| ameya49 | TheJulia : Ack (noted) | 16:48 |
| TheJulia | rpioso: because it is about api contract testing for the most part, integration scenario tests get piggy backed on that | 16:49 |
| rpioso | With real physical servers, a benefit of that approach is that the test can be completed against a fleet of different server models, along with their different firmware, in 1/N the time, where N represents the number of servers. Running it serially could take most of a day with reboots, etc. | 16:49 |
| TheJulia | so nothing is really designed to test every server the same exact way | 16:49 |
| TheJulia | it is all about running x scenario to determine the expected result | 16:50 |
| TheJulia | so to do so, it requires a separate class as it is less an integration/contract test, but a driver/hardware behavior/capability validation test. | 16:50 |
| rpioso | Would standing up N DevStacks, each with one (1) available baremetal node, be the preferred approach for introspection and our other Tempest tests? | 16:51 |
| iurygregory | that would ensure that you run the scenario in all tests | 16:52 |
| TheJulia | That would be really quite inefficent, tbh | 16:52 |
| TheJulia | one test could drive multiple | 16:52 |
| TheJulia | but the other thing to keep in mind is tempest is designed to also be able to be run by operators in the field or customers to validate clouds. So these sorts of thigns typically end up needing to be opt-in as well. | 16:53 |
| rpioso | TheJulia: Is there an example of one test driving multiple tests you could point us at? | 16:53 |
| TheJulia | rpioso: several, but not offhand' | 16:54 |
| iurygregory | I think standalone is one | 16:54 |
| TheJulia | Yeah, most of them are instandalone | 16:54 |
| rpioso | iurygregory: Doesn't that run against just one node? | 16:55 |
| TheJulia | aside from basic ops, but that is contract behavior validation + integration testing between multiple services | 16:55 |
| TheJulia | rpioso: there is a networking test which uses two nodes | 16:55 |
| TheJulia | and optionally a VM or a barmetal node in addition to it | 16:55 |
| iurygregory | rpioso, standalone is just because is ironic without nova =) | 16:55 |
| iurygregory | it's in the basic_ops | 16:56 |
| rpioso | iurygregory: Yep :-) | 16:56 |
| rpioso | For us, each node is associated with a specific physical server. | 16:56 |
| iurygregory | rpioso, I think we can end the meeting and continue the discussion in the irc (we only have 3min left) | 16:57 |
| iurygregory | can be? | 16:58 |
| rpioso | iurygregory: +1 | 16:58 |
| iurygregory | ty =) | 16:58 |
| iurygregory | #topic Baremetal SIG | 16:58 |
| iurygregory | #link https://etherpad.opendev.org/p/bare-metal-sig | 16:58 |
| arne_wiebalck | you said it all already during the announcements, tell all your friends :) | 16:58 |
| iurygregory | yeah :D | 16:59 |
| iurygregory | tweet! | 16:59 |
| arne_wiebalck | ML and twitter sent out | 16:59 |
| iurygregory | tks arne_wiebalck ! | 16:59 |
| iurygregory | #topic RFE review | 16:59 |
| iurygregory | skipping =) | 16:59 |
| iurygregory | #topic Open discussion | 16:59 |
| iurygregory | skipping | 16:59 |
| iurygregory | #topic Who is going to run the next meeting? | 16:59 |
| iurygregory | me o/ | 16:59 |
| iurygregory | Tks everyone! | 17:00 |
| iurygregory | #endmeeting | 17:00 |
| opendevmeet | Meeting ended Mon Feb 7 17:00:08 2022 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 17:00 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/ironic/2022/ironic.2022-02-07-15.59.html | 17:00 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/ironic/2022/ironic.2022-02-07-15.59.txt | 17:00 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/ironic/2022/ironic.2022-02-07-15.59.log.html | 17:00 |
| arne_wiebalck | thanks iurygregory ! | 17:00 |
| rpittau | thanks! | 17:00 |
| ameya49 | thanks | 17:00 |
| rpioso | Thank you! | 17:00 |
| dtantsur | Stderr: 'mount: /tmp/tmp8tyh6bah/dev: mount point does not exist.\n' | 17:01 |
| dtantsur | is partition deploy even working? with any image? | 17:01 |
| dtantsur | hmm, it's still cirros despite being not cirros.... lovely | 17:01 |
| TheJulia | dtantsur: code expects the folder to be present | 17:01 |
| TheJulia | and cirros's partition deployment is an empty filesystem. | 17:01 |
| dtantsur | yep, except that my patch does not use cirros any more | 17:01 |
| TheJulia | hmm | 17:01 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: CI: use a custom debian partition image instead of cirros https://review.opendev.org/c/openstack/ironic/+/827882 | 17:03 |
| dtantsur | or maybe I need a vacation again :( | 17:03 |
| opendevreview | Ruby Loo proposed openstack/ironic master: Anaconda deploy handles configdrive correctly https://review.opendev.org/c/openstack/ironic/+/827933 | 17:03 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [DNM] Testing the CI https://review.opendev.org/c/openstack/ironic/+/827500 | 17:03 |
| rloo | TheJulia: (and arne_wiebalck, dtantsur) I am happy to say that we've now gotten anaconda interface to work (for our use case). I have two PRs up with fixes for upstream. ^^ is one of them. https://review.opendev.org/c/openstack/ironic/+/827924 is the other. | 17:05 |
| arne_wiebalck | rloo: awesome! | 17:06 |
| *** akahat|dinner is now known as akaha|rover | 17:08 | |
| TheJulia | rloo: so, speaking of which, I've had a back-channel discussion with folks at rh. I think we're going to try and see if a centos-stream stage2 image can get posted publically as a build artifact which would enable CI... | 17:08 |
| rloo | TheJulia: that'd be awesome. I recall discussion with jay and dmitry? about CI but I haven't paid any attention to it (sorry) | 17:09 |
| dtantsur | rloo: \o/ | 17:09 |
| iurygregory | dtantsur, vacation++ | 17:09 |
| dtantsur | end of March, if the stars align.. | 17:09 |
| * iurygregory will probably have 2 more days in Feb.. | 17:09 | |
| TheJulia | I'm really feeling the need for a vacation... | 17:10 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: More fixes for anaconda deploy interface https://review.opendev.org/c/openstack/ironic/+/827924 | 17:12 |
| rloo | heh, dtantsur is SO FAST! I added a comment to that PR cuz something was bothering me... | 17:14 |
| dtantsur | rloo: should I hold approving? | 17:14 |
| dtantsur | I've been dealing with instance_info code.. a lot recently | 17:14 |
| rloo | not sure. take a quick look | 17:14 |
| rloo | (I don't really know anaconda that well) | 17:14 |
| dtantsur | rloo: mmm, I don't see your comment | 17:14 |
| rloo | that's cuz i haven't done upstream stuff for so long, i've forgotten how to. ok, should be there now! | 17:15 |
| dtantsur | rloo: a comment on https://review.opendev.org/c/openstack/ironic/+/827933: a configdrive *can* be a dict | 17:15 |
| dtantsur | I don't remember if we convert it specially for kickstart, have you checked it? | 17:15 |
| dtantsur | it will come as an ISO from now, but a standalone user can provide a dict | 17:16 |
| rloo | dtantsur: i traced the code and didn't see configdrive as a dict. or i forgot, was a while ago. will look again now... | 17:16 |
| dtantsur | I know that because we rely on this property in our downstream deploy step :) | 17:16 |
| rloo | oh. hmm. ok, i better look then... | 17:17 |
| dtantsur | maybe you only need to keep the 'if isinstance(configdrive, dict)` logic | 17:17 |
| dtantsur | rloo: re your question: yes, I think I agree. | 17:18 |
| dtantsur | I can imagine a case of a stage2 coming from an image (it's an inherent image property, after all) and a template coming from instance_info (it's something an operator is planning) | 17:19 |
| dtantsur | I don't quite agree that you cannot change it in a follow-up though | 17:19 |
| dtantsur | allowing something is usually fine, unlike disallowing :) | 17:19 |
| rloo | dtantsur: yes, that is what i was thinking. | 17:19 |
| rloo | ok, if a followup is ok, that's fine. i was worried about versioning... | 17:20 |
| rloo | and backporting -- whether that is a 'feature' or a bug. (but we don't need that backported so...) | 17:20 |
| dtantsur | backporting... | 17:20 |
| dtantsur | may be an argument to changing this patch | 17:20 |
| dtantsur | rloo: I've removed the approval. If you still prefer to follow-up, feel free to re-add it. | 17:21 |
| rloo | ha ha. i'm happy to update that patch. although looking at the configdrive thing now. so i'll update that patch later today. thx for your feedback! | 17:22 |
| rloo | well, maybe i should update that patch now, since that is prob the more urgent one... | 17:22 |
| rpittau | good night o/ | 17:28 |
| rloo | dtantsur: does it make sense for instance_info to have stage2 but not ks_template? | 17:29 |
| dtantsur | rloo: standalone ironic, but using the default template? | 17:29 |
| rloo | eh, i'm going to allow it. for now anyway. | 17:29 |
| dtantsur | I think there is a narrow but valid case | 17:29 |
| rloo | always some use case, thx dtantsur! | 17:29 |
| TheJulia | So has everyone at least asked one person to submit something to the summit? | 17:37 |
| TheJulia | I think the valid usecase is largely that the cfg aiui is largely the contents/definition of the $thing to do. In our narrow scope, maybe that is fairly standardized, but maybe elsewhere they need to heavily customize at times | 17:38 |
| opendevreview | Ruby Loo proposed openstack/ironic master: More fixes for anaconda deploy interface https://review.opendev.org/c/openstack/ironic/+/827924 | 17:58 |
| dtantsur | TheJulia: iurygregory and I have submitted something :) | 18:02 |
| opendevreview | Merged openstack/sushy-tools master: Accept non-string types for BIOS settings https://review.opendev.org/c/openstack/sushy-tools/+/827769 | 18:04 |
| iurygregory | =) | 18:05 |
| iurygregory | janders and I will submit something also =) | 18:05 |
| dtantsur | could we merge https://review.opendev.org/c/openstack/ironic/+/826467/ and https://review.opendev.org/c/openstack/ironic/+/826470 before the release please? | 18:11 |
| dtantsur | ideally https://review.opendev.org/c/openstack/ironic/+/826927 and https://review.opendev.org/c/openstack/ironic/+/825305 but they can wait if nobody has cycles | 18:12 |
| iurygregory | looking now | 18:18 |
| dtantsur | oh, and https://review.opendev.org/c/openstack/bifrost/+/827648 please, this is actually the most urgent | 18:24 |
| * dtantsur tries a xena backport | 18:25 | |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost stable/xena: Revert "Install libvirt-python from source instead of a wheel" https://review.opendev.org/c/openstack/bifrost/+/828178 | 18:26 |
| dtantsur | why on bloody earth do some metalsmith jobs end up with IPv6 only addresses? | 18:34 |
| dtantsur | Guru Meditation in neutron, mmmm, nice | 18:38 |
| dtantsur | https://www.dropbox.com/s/qce2u1nc8h7zqb6/cat-aaaaaaaa.jpg?dl=0 | 18:38 |
| arne_wiebalck | :-D | 18:40 |
| iurygregory | probably some sort of magic... | 18:41 |
| dtantsur | well, I wonder if it actually just does nothing | 18:42 |
| iurygregory | this is interesting, the OpenStack Ops Meetup just tweeted about the cinder quota form https://twitter.com/osopsmeetup/status/1490757015844569094?t=BxVVYnrCX-8NRbbTpqFe_w&s=19 :D | 18:42 |
| iurygregory | maybe so maybe if we want to do some sort of survey we can ask their help to post on twitter XD | 18:43 |
| dtantsur | Failed to bind port 47a33344-d3dc-4248-8c67-be5ea644f9f7 on host 0f9251e1-00c3-437f-947b-ec46f08be421 for vnic_type baremetal using segments [{'id': '32eb1d82-d533-4fe7-93ac-77ff5d9678c4', 'network_type': 'vxlan', 'physical_network': None, 'segmentation_id': 270, 'network_id': '09b6d9fc-dc9c-4d0a-a4a3-0ce04691dfa8'}] | 18:43 |
| dtantsur | WHY | 18:43 |
| * arne_wiebalck has now the image from above in the head whenever dtantsur says 'WHY?' | 18:45 | |
| dtantsur | this is roughly how I look now, yes | 18:45 |
| opendevreview | Ruby Loo proposed openstack/ironic master: Anaconda deploy handles configdrive correctly https://review.opendev.org/c/openstack/ironic/+/827933 | 18:47 |
| dtantsur | okay, a successful job also has a ton of these "failed to bind" messages. don't we use networking-baremetal to avoid it? do I want to hear the answer to my question? | 18:48 |
| TheJulia | dtantsur: yeah, it is supposed to wrap up the end of the stack for vnic_baremetal and prevent neutron from doing its blanket failure status on the port | 18:52 |
| dtantsur | this matches my memory. then I wonder why we're seeing the failures like above | 18:53 |
| opendevreview | Dmitry Tantsur proposed openstack/metalsmith stable/wallaby: Avoid non-Stream CentOS 8, it is EOL; make one job non-voting https://review.opendev.org/c/openstack/metalsmith/+/827234 | 18:58 |
| dtantsur | a bit brutal ^^ but what can O id? | 18:58 |
| dtantsur | * can I do? | 18:58 |
| dtantsur | okay, https://review.opendev.org/c/openstack/bifrost/+/828178/ seems to be passing and can be reviewed as well | 19:01 |
| arne_wiebalck | bye everyone o/ | 19:02 |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost master: DNM test the CI https://review.opendev.org/c/openstack/bifrost/+/828182 | 19:02 |
| dtantsur | hmm, is this enough to declare partition local boot at least partly broken or am I missing something again? https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_f03/827882/5/check/ironic-tempest-ipa-partition-pxe_ipmitool/f0348ae/controller/logs/ironic-bm-logs/node-1_no_ansi_2022-02-07-18:37:52_log.txt | 19:07 |
| iurygregory | grub is broken XD | 19:11 |
| iurygregory | I'm wondering if we are missing something... | 19:12 |
| dtantsur | maybe? https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_f03/827882/5/check/ironic-tempest-ipa-partition-pxe_ipmitool/f0348ae/controller/logs/ironic-bm-logs/node-1_no_ansi_2022-02-07-18:09:17_log.txt shows a completely normal run of install_bootloader | 19:12 |
| dtantsur | at least from my perspective | 19:12 |
| dtantsur | and grub *does*, in fact, start | 19:12 |
| dtantsur | I wish we logged the generated grub.cfg... | 19:12 |
| iurygregory | Stderr: 'readlink: /sys/class/net/dummy0/device/driver/module: No such file or directory\n' _detect_cna_card /usr/local/lib/python3.6/site-packages/ironic_python_agent/hardware_managers/cna.py:42 | 19:17 |
| dtantsur | red herring | 19:18 |
| iurygregory | oh =) | 19:19 |
| dtantsur | More fun. My image_type patch works when you insert assetions in the code but does not otherwise (?). | 19:19 |
| dtantsur | Does anyone have a rational explanation or am I (finally) getting deeply insane? | 19:20 |
| opendevreview | Merged openstack/bifrost master: Revert "Install libvirt-python from source instead of a wheel" https://review.opendev.org/c/openstack/bifrost/+/827648 | 19:24 |
| dtantsur | okay, at least the grub problem does reproduce locally | 19:25 |
| stevebaker[m] | good morning | 19:26 |
| iurygregory | asserts in the code to make it work? what?! O.o | 19:30 |
| iurygregory | since you are able to reproduce locally this is a good thing | 19:31 |
| dtantsur | iurygregory: maybe it just works every other time.. (no, it does not make sense either) | 19:31 |
| dtantsur | morning stevebaker[m] | 19:31 |
| iurygregory | if we try to find it makes sense we will go crazy | 19:33 |
| dtantsur | I can only suspect that debian-minimal images are not suitable for us? which is weird. | 19:41 |
| dtantsur | an ubuntu partition image is 959M, a debian one is 270M. maybe something is just missing there. | 19:43 |
| opendevreview | Merged openstack/ironic master: Return non-zero exit code on failures https://review.opendev.org/c/openstack/ironic/+/826467 | 19:45 |
| opendevreview | Merged openstack/ironic master: Wait for conductor start before notifying systemd https://review.opendev.org/c/openstack/ironic/+/826470 | 19:45 |
| dtantsur | will try a normal debian, not debian-minimal, tomorrow | 19:46 |
| dtantsur | see you o/ | 19:46 |
| iurygregory | quite a lot of diff between the sizes, bye o/ | 19:50 |
| opendevreview | Merged openstack/ironic master: deploy_utils: only check glance for image properties kernel/ramdisk https://review.opendev.org/c/openstack/ironic/+/826927 | 19:53 |
| opendevreview | Julia Kreger proposed openstack/sushy master: Fix session authentication issues https://review.opendev.org/c/openstack/sushy/+/821324 | 19:57 |
| opendevreview | Merged openstack/ironic master: Explicit parameter to distinguish partition/whole-disk images https://review.opendev.org/c/openstack/ironic/+/825305 | 20:04 |
| TheJulia | arne_wiebalck: did you, by chance, identify any exceptions? | 22:13 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
