| *** pmannidi is now known as pmannidi|Lunch | 03:17 | |
| *** pmannidi|Lunch is now known as pmannidi | 04:31 | |
| arne_wiebalck | Good morning janders and Ironic! | 06:31 |
|---|---|---|
| *** pmannidi is now known as pmannidi|brb | 06:32 | |
| *** pmannidi|brb is now known as pmannidi | 07:01 | |
| iurygregory | good morning janders arne_wiebalck and Ironic | 07:09 |
| arne_wiebalck | hey iurygregory o/ | 07:10 |
| iurygregory | fyi today is a public holiday in CZ so I won't be around too much, if there is something that needs attention feel free to ping =) | 07:10 |
| dtantsur | morning folks | 07:48 |
| dtantsur | iurygregory: what are you doing here? go visit some Pegas! :) | 07:48 |
| iurygregory | dtantsur, :D Pegas++ (maybe I should grab some lunch there) | 07:54 |
| rpittau | good morning ironic! o/ | 07:57 |
| opendevreview | Aija Jauntēva proposed openstack/ironic master: Fix iDRAC configuration mold docs https://review.opendev.org/c/openstack/ironic/+/811316 | 08:15 |
| opendevreview | Merged openstack/ironic master: Fix iDRAC configuration mold docs https://review.opendev.org/c/openstack/ironic/+/811316 | 09:34 |
| rpittau | mmmm is CI completely foobar? | 10:01 |
| dtantsur | dunno, I'm hacking on ironic-image today | 10:02 |
| iurygregory | there is a know problem in devstack related to placement | 10:03 |
| rpittau | I see the same error in multiple jobs and multiple patches | 10:03 |
| rpittau | https://paste.openstack.org/show/809642/ | 10:05 |
| dtantsur | Oo | 10:20 |
| rpittau | I don't understand, the conductor is actually running :/ | 10:24 |
| dtantsur | and ironic-api is okay? | 10:25 |
| rpittau | I was looking at it now, I see a client-side error | 10:25 |
| rpittau | https://6c6ba23f5d3153a2b236-5726fddc2982c27736234618fe9da818.ssl.cf5.rackcdn.com/810828/1/check/ironic-standalone/dff4d5e/controller/logs/screen-ir-api.txt | 10:25 |
| rpittau | in another patch there's a different error though | 10:27 |
| rpittau | https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_c18/776507/8/check/ipa-tempest-bios-ipmi-direct-tinyipa-src/c18b6b8/controller/logs/screen-ir-api.txt | 10:27 |
| rpittau | mmm wait | 10:27 |
| rpittau | the client error is the same | 10:27 |
| rpittau | but here looks like we have something else | 10:27 |
| dtantsur | I wonder if it's an issue in devstack (around uwsgi configuration) or openstacksdk | 10:51 |
| dtantsur | hmm, wait, ironicclient doesn't use openstacksdk | 10:51 |
| opendevreview | Verification of a change to openstack/ironic master failed: Refactor: move base_iso handling from create_boot_iso https://review.opendev.org/c/openstack/ironic/+/810076 | 11:14 |
| opendevreview | Verification of a change to openstack/ironic master failed: Use an ImageCache for provided boot/deploy ISO images https://review.opendev.org/c/openstack/ironic/+/810241 | 11:14 |
| dtantsur | arne_wiebalck: is https://github.com/metal3-io/ironic-hardware-inventory-recorder-image something you ended up using? | 11:26 |
| dtantsur | I would like to try purging metal3 from unused repositories, and this is a candidate | 11:26 |
| opendevreview | Verification of a change to openstack/ironic master failed: Refactor: move base_iso handling from create_boot_iso https://review.opendev.org/c/openstack/ironic/+/810076 | 11:56 |
| *** jamesdenton_alt is now known as jamesdenton | 12:47 | |
| TheJulia | good morning | 13:16 |
| dtantsur | morning TheJulia | 13:24 |
| TheJulia | arne_wiebalck: I've set a time for the US friendly session | 13:44 |
| arne_wiebalck | TheJulia: thanks! | 13:45 |
| arne_wiebalck | TheJulia: I will send out a reminder/invitation nearer the time. | 13:45 |
| TheJulia | arne_wiebalck: awesome | 14:50 |
| TheJulia | brraaains | 14:50 |
| TheJulia | dtantsur: we're meeting in a little over an hour to go over open rfe's right? | 14:51 |
| dtantsur | TheJulia: correct | 15:04 |
| rpittau | bye everyone, good night! o/ | 15:33 |
| arne_wiebalck | bye everyone o/ | 15:46 |
| TheJulia | So the proxypass changes really breaks us | 15:51 |
| dtantsur | mmm? | 15:57 |
| TheJulia | https://meetpad.opendev.org/ironic | 16:00 |
| TheJulia | proxypass is now passing /baremetal/ through all requests to ironic | 16:00 |
| dtantsur | so nice of it... | 16:01 |
| dtantsur | rloo: hi! are you still interested in https://review.opendev.org/c/openstack/ironic-specs/+/471174 ? | 16:28 |
| rloo | dtantsur: heh, don't recall. let me see... (in a meeting now) | 16:33 |
| dtantsur | same question re https://review.opendev.org/c/openstack/ironic-specs/+/471174/ | 16:34 |
| dtantsur | wait, it's the same link | 16:35 |
| dtantsur | hmmm, why is it attached to two RFEs? | 16:35 |
| rloo | dtantsur: i have no need for that (not that I am aware). If no one has asked for it, then maybe close it or whatever. (wrt 471174) | 16:35 |
| dtantsur | rloo: I suspect https://storyboard.openstack.org/#!/story/1671549 is something we should eventually fix.. | 16:36 |
| dtantsur | I'll look at the spec in more details | 16:36 |
| rloo | it is definitely something missing from the api but.. ?? and machine-readable i think was maybe the bottleneck... | 16:37 |
| dtantsur | mgoddard: hi! is https://storyboard.openstack.org/#!/story/1737947 still a problem for you? | 16:57 |
| dtantsur | TheJulia: we seem to have closed 14 RFEs, not bad! | 17:05 |
| TheJulia | \o/ | 17:06 |
| TheJulia | dtantsur: I've continued a little :) | 17:21 |
| sam_z | TheJulia: you (rightfully) pointed out that making an Ironic call for every pseudo-Redfish call comes with a lot of overhead, and i just wanted to get your take on a possible solution: since one of the pros of making an Ironic API call for every equivalent Redfish API call is that we won't need to implement the same functionality in two different places (e.g. a database query to get the details of a node perhaps), i was wonderi | 17:23 |
| sam_z | ng if importing the Ironic API controller responsible for said functionality and calling its methods (and maybe transforming the results) is a good idea? | 17:23 |
| sam_z | it seems. very hacky and clunky. but it also feels to me like it would work? | 17:24 |
| sam_z | ^ famous last words | 17:24 |
| TheJulia | sam_z: I would try to avoid transforming it's response results since there is a lot already there you don't need | 17:51 |
| TheJulia | I would honestly just ask for the list directly from the objects | 17:51 |
| TheJulia | or the object itself, and work from there like the API code does today | 17:51 |
| sam_z | ah ok, gotcha! | 17:52 |
| TheJulia | Basically you'd still transform some things, remove some fields or form entirely new response bodies | 17:52 |
| TheJulia | but you wouldn't need entire node objects | 17:52 |
| TheJulia | since a row in the nodes table has ~60 columns | 17:52 |
| sam_z | oh wow | 17:52 |
| sam_z | i uh did not know that | 17:52 |
| TheJulia | and any time you can do ~15 columns instead of ~60, your better off | 17:52 |
| TheJulia | yeah.... its a bit excessive | 17:52 |
| sam_z | i appreciate the feedback; i'll probably still ask for help when it comes time to do the implementation but that's a much better solution | 17:53 |
| TheJulia | sam_z: I also did a TON of work on performance of queries earlier in the development cycle | 17:53 |
| TheJulia | largely revolving around handling specific field responses to make it as efficient as possible | 17:53 |
| TheJulia | dtantsur: thoughts on keeping https://storyboard.openstack.org/#!/story/1526757 open or not? | 18:01 |
| sam_z | i will definitely welcome any and all advice for maximizing query efficiency (considering my general lack of experience lol) | 18:01 |
| TheJulia | Oh, and fwiw, your totally free to use flask :) | 18:06 |
| TheJulia | We put up an outreachy item to work towards overall pecan->flask migration | 18:06 |
| TheJulia | ayoung: o/ | 18:06 |
| ayoung | Hey ho. I am now at a Hard ware company and looks like I will be getting intimate with Ironic. | 18:09 |
| TheJulia | ayoung: \o/ | 18:10 |
| * TheJulia dances | 18:10 | |
| TheJulia | I had a feeling! I swear it wasn't bunnies! | 18:10 |
| ayoung | Im now learning how to debug the node cleaning process. | 18:10 |
| ayoung | ipmitool -H 10.76.97.176 -U ADMIN -I lanplus -P ADMIN chassis power status | 18:11 |
| sam_z | TheJulia: was that comment about flask directed at me? i have looked into flask a teensy bit for a personal project idea i had so i would be willing to give that a shot as well! | 18:14 |
| TheJulia | iurygregory: dtantsur: I think we should go ahead and ax motlteniron. Maybe we can get infra to check logs to see if people still clone/use it? | 18:18 |
| TheJulia | sam_z: yes, it was directed at you. Awesome! | 18:18 |
| ayoung | We redid Keystone in Flask a few years back | 18:24 |
| ayoung | it wasn't too bad...although kmalloc did all the work | 18:24 |
| TheJulia | Yeah, it shouldn't be hard now that we've basically killed wsme in ironic | 18:25 |
| ayoung | I had a funny idea for a feature request. The Landing pages for each of the services should tell you what service you are on. If you don't have the ports memorized, you have no idea that port 5000 is Keystone, or port...whatever is Ironic. | 18:33 |
| ayoung | v3! Great I'm on V3 of the API. What API? | 18:33 |
| iurygregory | TheJulia, makes sense to me | 18:34 |
| TheJulia | ayoung: I *thought* we did that | 18:39 |
| TheJulia | or at least include it in the body | 18:39 |
| TheJulia | Maybe not everyone did? | 18:39 |
| TheJulia | dtantsur: could https://storyboard.openstack.org/#!/story/1525233 make sense for outreachy onboarding? Although it involves DB work so maybe not | 18:47 |
| JayF | ayoung: TheJulia: I *think* it's in headers, but not in text | 19:09 |
| JayF | not 100% sure and don't have general access to an ironic api atm to check | 19:09 |
| TheJulia | I don't have a running one at the moment unfortunately | 19:11 |
| TheJulia | I've been cleaning up RFEs | 19:11 |
| ayoung | Why so you do | 19:12 |
| ayoung | "name": "OpenStack Ironic API", "description": "Ironic is an OpenStack project which aims to provision baremetal machines.", | 19:12 |
| ayoung | that isw so RESTful. Lovely | 19:12 |
| * TheJulia should fix "aims" | 19:13 | |
| TheJulia | We do it, we do it well. | 19:13 |
| NobodyCam | Grr to missing hypervisor records. Any trick I’m missing to recreate them ? | 19:26 |
| TheJulia | umm.. undelete them? | 19:33 |
| TheJulia | its a field | 19:33 |
| janders | good morning Ironic o/ | 20:40 |
| janders | TheJulia: dtantsur asked me to check in with you w/r/t NVMe cleaning enhancements | 20:41 |
| janders | apologies for no progress with this - been mostly focused on downstream related stuff lately | 20:41 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!