| *** mushrushu_ is now known as mushrushu | 00:08 | |
| *** osmanlicilegi is now known as Guest7485 | 00:18 | |
| *** seebaer is now known as seba | 00:20 | |
| *** shadower9 is now known as shadower | 01:32 | |
| opendevreview | Steve Baker proposed openstack/ironic master: WIP Implement a DHCP driver backed by dnsmasq https://review.opendev.org/c/openstack/ironic/+/851681 | 04:39 |
|---|---|---|
| *** undefined is now known as Guest7674 | 07:59 | |
| rpittau | good morning ironic! o/ | 08:39 |
| kubajj | good morning rpittau | 09:08 |
| rpittau | hey kubajj :) | 09:08 |
| opendevreview | Merged openstack/sushy stable/yoga: Handle mismatch between registry Identity and the way it's referred to https://review.opendev.org/c/openstack/sushy/+/851883 | 10:10 |
| opendevreview | Merged openstack/sushy stable/xena: Handle mismatch between registry Identity and the way it's referred to https://review.opendev.org/c/openstack/sushy/+/851884 | 10:10 |
| opendevreview | Aija Jauntēva proposed openstack/ironic master: Update raid_type handling for Redfish raid_config https://review.opendev.org/c/openstack/ironic/+/852393 | 10:30 |
| opendevreview | Vanou Ishii proposed openstack/ironic master: [WIP] Deal with iRMC virtual media incompatibility https://review.opendev.org/c/openstack/ironic/+/823790 | 10:44 |
| iurygregory | good morning Ironic | 11:25 |
| rpittau | hey iurygregory :) | 11:27 |
| iurygregory | rpittau, if you have some time today can you check https://review.opendev.org/c/openstack/ironic/+/852234 ? o/ | 11:28 |
| kubajj | Is it possible to connect to the devstack nodes somehow? I tried enabling the console through ironic, but it says that the driver ipmi does not support it. | 11:43 |
| dtantsur | kubajj: sudo virsh console <node> maybe? | 11:52 |
| kubajj | dtantsur: does it need to be in an active state for it to work? (I am getting internal error: character device console0 is not using a PTY error) | 12:00 |
| dtantsur | well, something has to output there. also on devstack you can find serial console logs somewhere, I cannot remember where | 12:01 |
| dtantsur | I suspect /opt/stack/data/ironic/logs | 12:02 |
| dtantsur | kubajj: ^^ | 12:03 |
| kubajj | dtantsur: the thing is that I wanted to connect to a node to setup a RAID on it to be able to test what happens during the cleaning process if RAID is present. I assume that the console would not help me with that, right? | 12:08 |
| dtantsur | the console will if you find a way to connect to it (it's very confusing) | 12:08 |
| kubajj | Ok, thanks | 12:09 |
| dtantsur | TheJulia: eat lunch meow! | 12:11 |
| dtantsur | the words to live by | 12:11 |
| *** Guest7674 is now known as rcastillo | 12:35 | |
| TheJulia | wait, I should eat lunch right meow?!? | 12:56 |
| TheJulia | My coffee is still brewing! | 12:56 |
| * TheJulia is sooooo confused | 12:56 | |
| TheJulia | Also, good morning everyone | 12:56 |
| * TheJulia is not sure what else the response could be right now | 13:08 | |
| TheJulia | sweet, wrong window | 13:09 |
| TheJulia | coffee! wake me up! | 13:09 |
| iurygregory | good morning TheJulia =) | 13:09 |
| opendevreview | Julia Kreger proposed openstack/ironic-tempest-plugin master: WIP: Initial tempest test idea anaconda deploy https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/835917 | 13:31 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Redfish: Consider password part of the session cache https://review.opendev.org/c/openstack/ironic/+/850259 | 13:39 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Concurrent Distructive/Intensive ops limits https://review.opendev.org/c/openstack/ironic/+/840006 | 13:44 |
| *** yoctozepto is now known as yoctozepto[pto] | 13:46 | |
| *** hjensas is now known as hjensas|afk | 14:07 | |
| dtantsur | TheJulia: any plans to update https://review.opendev.org/c/openstack/sushy/+/850547/ ? | 14:25 |
| TheJulia | dtantsur: if you are up to rev it for the comments, I'm cool with that | 14:27 |
| TheJulia | dtantsur: I didn't get braincells on Friday unfortunately | 14:27 |
| dtantsur | okie, lemme see | 14:28 |
| kubajj | dtantsur, TheJulia I modified get_os_install_device to use the helper function I created so that drives are not chosen as root when on skip list. Should I create a new change from this or put it into the old one (I guess it is easier to review a smaller patch)? Should I rename the helper function if it is not used only for cleaning (it has _for_cleaning in the name)? | 14:28 |
| iurygregory | I can update if necessary o/ | 14:28 |
| dtantsur | iurygregory: the sushy patch? would be great | 14:28 |
| iurygregory | yeah | 14:28 |
| dtantsur | kubajj: hmm, is it intended to not choose them as root? | 14:29 |
| dtantsur | I can see pros and cons.. either way the setting name should reflect it, something like skip_block_devices (no reference to cleaning) | 14:29 |
| TheJulia | timeu_: o/ you around? | 14:29 |
| kubajj | dtantsur: I think it should not choose them as root as it would overwrite the data with the image. We had a chat about this with Arne and it could be avoided by leaving the responsibility at the user side (to specifiy root_device), but we agreed on skipping them during deploy as well | 14:32 |
| TheJulia | dtantsur: I just saw the /Pending comment... whiskey tango. | 14:37 |
| dtantsur | :D | 14:37 |
| TheJulia | please insert whiskey! | 14:37 |
| iurygregory | yeah, the Pending is funny | 14:37 |
| TheJulia | seriously though, I did look at the spec documents and nothing like that is described | 14:37 |
| dtantsur | kubajj: I'm trying to think if I see a case to skip cleaning a root disk.. nothing comes to my mind immediately | 14:38 |
| TheJulia | jungleboyj: o/ Do you have some brain cells and potentially a contact we can leverage w/r/t https://bugzilla.redhat.com/show_bug.cgi?id=2103710 | 14:41 |
| TheJulia | hmm... ussuri, why you no like us https://zuul.opendev.org/t/openstack/build/3b885e4f35854345ac41563da952e476/logs | 14:45 |
| iurygregory | #startmeeting ironic | 14:59 |
| opendevmeet | Meeting started Mon Aug 8 14:59:59 2022 UTC and is due to finish in 60 minutes. The chair is iurygregory. Information about MeetBot at http://wiki.debian.org/MeetBot. | 14:59 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 14:59 |
| opendevmeet | The meeting name has been set to 'ironic' | 14:59 |
| rloo | o/ | 15:00 |
| iurygregory | Hello ironicers, welcome to our weekly meeting o/ | 15:00 |
| matfechner | o/ | 15:00 |
| iurygregory | you can find the agenda for the meeting in the wiki | 15:00 |
| iurygregory | #link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting | 15:00 |
| rpittau | o/ | 15:01 |
| * iurygregory gives a few minutes to see if we will have more people | 15:01 | |
| kamlesh6808c | o/ | 15:01 |
| TheJulia | o/ | 15:01 |
| iurygregory | #topic Announcements / Reminder | 15:02 |
| iurygregory | #info PTG will be virtual! | 15:03 |
| iurygregory | if you haven't seen the email in the openstack-discuss about PTG going virtual you can check this link | 15:03 |
| iurygregory | #link https://lists.openstack.org/pipermail/openstack-discuss/2022-August/029879.html | 15:03 |
| dtantsur | o/ | 15:03 |
| iurygregory | Don't forget to register for the PTG! | 15:03 |
| iurygregory | #link https://openinfra-ptg.eventbrite.com/ | 15:03 |
| iurygregory | I think we can continue using the same etherpad (even if now it's virtual) | 15:04 |
| iurygregory | #link https://etherpad.opendev.org/p/ironic-ptg-planing-Columbus-OH | 15:04 |
| iurygregory | #topic Revise Ironic Vision #2 | 15:06 |
| TheJulia | Just a thought w/r/t the ptg etherpad, we should likely change etherpads | 15:06 |
| iurygregory | TheJulia, ack! | 15:06 |
| iurygregory | I will update and also send an email to openstack-discuss | 15:06 |
| TheJulia | k | 15:07 |
| iurygregory | going back to Ironic Vision #2, we only had 5 people and the result was a day that only 2 could join... | 15:07 |
| TheJulia | Given the PTG is virtual now... perhaps we just allocate time to it? | 15:07 |
| iurygregory | I'm wondering if I should add dates for next week, and see if more people want to vote? | 15:07 |
| iurygregory | hummm | 15:08 |
| iurygregory | sounds like a plan also | 15:08 |
| TheJulia | We are in a tricky time of year too, people still on holidays. Some of us still need to take time off :) | 15:08 |
| iurygregory | yeah =) | 15:08 |
| TheJulia | PTG, as much as I don't like the idea of waiting on visioning stuffs that long, at *least* is a block of time far out that we can schedule and align around | 15:08 |
| iurygregory | and we would have more participants... | 15:09 |
| iurygregory | I do think it would make sense | 15:09 |
| iurygregory | 3yrs ago we did at the PTG if I recall... | 15:09 |
| rloo | ++ for doing it at PTG | 15:10 |
| TheJulia | 2018 I think... | 15:10 |
| iurygregory | yeah =) | 15:10 |
| TheJulia | time flies | 15:10 |
| iurygregory | right after I joined | 15:10 |
| iurygregory | or it was begin of 2019... | 15:11 |
| iurygregory | moving on o/ | 15:11 |
| TheJulia | time has literally blurred | 15:11 |
| iurygregory | #info New bugfix branches next week | 15:11 |
| iurygregory | yeah is that time of the year =) | 15:11 |
| iurygregory | ops of the cycle XD | 15:12 |
| iurygregory | just a heads-up that next week we will be creating the second bugfix branch in the Zed cycle | 15:12 |
| iurygregory | 3months after the first one | 15:12 |
| rpittau | great | 15:13 |
| rpittau | :) | 15:13 |
| iurygregory | Does anyone have something we should add to Announcements / Reminder ? | 15:13 |
| iurygregory | ok, moving o/ | 15:14 |
| iurygregory | #topic Review subteam status reports | 15:15 |
| iurygregory | #link https://etherpad.opendev.org/p/IronicWhiteBoard | 15:15 |
| iurygregory | starting around L90 | 15:15 |
| TheJulia | iurygregory: with us being so close, will it effectively be our final release for the cycle? | 15:16 |
| TheJulia | The OpenStack release cycle that is | 15:16 |
| dtantsur | we can have a shorter 3rd release | 15:17 |
| iurygregory | TheJulia, let me double check the final date for the release... | 15:17 |
| iurygregory | yeah, this is what normally happened in the past cycles I think | 15:17 |
| TheJulia | k | 15:18 |
| iurygregory | TheJulia, in the Anaconda Deploy CI I think we can add the WIP patch you posted in the tempest plugin? | 15:23 |
| TheJulia | we do and it is super close | 15:23 |
| iurygregory | NICE \o/ | 15:23 |
| TheJulia | I should have it working this week | 15:23 |
| TheJulia | if at least, the stars align and all | 15:23 |
| iurygregory | ++ | 15:25 |
| iurygregory | for the OpenConfig I'm going to review more patches this week | 15:26 |
| TheJulia | Yeah, there is a *lot* there, but hjensas|afk has been super responsive | 15:26 |
| iurygregory | agree | 15:27 |
| iurygregory | ok, moving on o/ | 15:27 |
| iurygregory | #topic Deciding on priorities for the coming week | 15:27 |
| iurygregory | #link https://review.opendev.org/q/status:open+hashtag:ironic-week-prio | 15:27 |
| rpittau | I added a bunch of backports! :) | 15:28 |
| iurygregory | backport all the things! | 15:28 |
| iurygregory | rpittau, tks! | 15:28 |
| TheJulia | \o/ | 15:28 |
| TheJulia | I also added a couple fixes for our unit testing which one of the FJ folks noticed and corrected | 15:29 |
| iurygregory | TheJulia, ack | 15:30 |
| TheJulia | if there are no objections, I'd also like to add https://review.opendev.org/c/openstack/ironic/+/851802 | 15:30 |
| iurygregory | none from my side | 15:30 |
| TheJulia | I *think* the FJ patch is not going to land until some of the unit test fixes merge, but they already have 1x+2 on them | 15:32 |
| iurygregory | I will give priority to review the unit tests in this case | 15:33 |
| iurygregory | (right after my lunch, that will be after our upstream meeting) =) | 15:33 |
| iurygregory | I added https://review.opendev.org/c/openstack/ironic/+/852234 on friday | 15:34 |
| iurygregory | thanks for the review dtantsur and rpittau o/ | 15:34 |
| rpittau | iurygregory: np :) | 15:34 |
| TheJulia | I have concerns w/r/t that one, but I need to be fully awake to digest it | 15:35 |
| iurygregory | TheJulia, ack =) | 15:35 |
| TheJulia | I'll try to review it this morning | 15:35 |
| iurygregory | I should probably create a story for it... | 15:35 |
| TheJulia | ++++++++++++++ | 15:35 |
| TheJulia | That would likely help me with context | 15:35 |
| iurygregory | cool =) | 15:36 |
| iurygregory | ok, moving on | 15:36 |
| iurygregory | #topic Baremetal SIG | 15:36 |
| iurygregory | I don't think we have any updates .... | 15:37 |
| TheJulia | arne is back next week and my plate is full | 15:37 |
| iurygregory | ok o/ | 15:37 |
| iurygregory | skipping RFE review (I don't have the one for my patch created...) | 15:38 |
| iurygregory | #topic Open discussion | 15:38 |
| iurygregory | I have one topic to start | 15:38 |
| iurygregory | last week during the meeting we talked about having a pre-ptg meeting to check topics | 15:38 |
| iurygregory | do we think we still need? | 15:39 |
| TheJulia | it has typically helped to have one session to collalese topics | 15:39 |
| iurygregory | I will send the email with the doodle =) | 15:40 |
| TheJulia | At least that is my feeling | 15:40 |
| TheJulia | I don't know how others feel | 15:40 |
| iurygregory | I think it's a good idea to test and see how it goes (I don't remember us doing in the past...) | 15:41 |
| JayF | Is it possible for when those doodles come to a conclusion for whoeever is running them to send an actual-calendar-invite | 15:41 |
| iurygregory | JayF, yes! | 15:41 |
| JayF | I either missed the email or didn't get one last time we did a meeting setup this way; and an actual meeting invite makes it 100x less likely I miss it | 15:41 |
| iurygregory | JayF, I will work on it, thanks for the feedback :D | 15:42 |
| TheJulia | ++++++ | 15:42 |
| TheJulia | I've heard the exact same thing a number of times | 15:42 |
| TheJulia | unfortunately many of us have our day handled by our calendars | 15:42 |
| iurygregory | I need to remember to schedule things using my company email (to avoid the problem in the first ironic vision meeting...) | 15:43 |
| iurygregory | since we are talking about doodle and meetings | 15:45 |
| iurygregory | would be worth trying to reach out to ops meetup folks and see what day/time during the PTG is better for them to join the ironic community? | 15:45 |
| TheJulia | iurygregory: I think that is a fantastic idea! | 15:46 |
| iurygregory | I'm also thinking in keep the same format we did in the past to get APAC folks participating in the PTG | 15:47 |
| TheJulia | sounds like a plan to me | 15:47 |
| iurygregory | what others think about it ? | 15:48 |
| iurygregory | =) | 15:48 |
| rpittau | sounds good to me | 15:49 |
| iurygregory | does anyone have something for Open Discussion? | 15:50 |
| TheJulia | I've got nothing | 15:51 |
| iurygregory | ok, I think this is all for our meeting today o/ | 15:52 |
| iurygregory | tks everyone! | 15:52 |
| iurygregory | #endmeeting | 15:52 |
| opendevmeet | Meeting ended Mon Aug 8 15:52:13 2022 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 15:52 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/ironic/2022/ironic.2022-08-08-14.59.html | 15:52 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/ironic/2022/ironic.2022-08-08-14.59.txt | 15:52 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/ironic/2022/ironic.2022-08-08-14.59.log.html | 15:52 |
| kamlesh6808c | iurygregory,TheJulia : would you help to review these two patches https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/841601 | 15:54 |
| kamlesh6808c | & https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/848709 | 15:54 |
| rpittau | good night! o/ | 16:04 |
| TheJulia | hjensas: Hey, if you could spend a few minutes and take a look at https://review.opendev.org/c/openstack/ironic-inspector/+/842275 it would be appreciated. Specifically the comments. Maybe something might ring in your brain, but I think we need more info :) | 17:03 |
| hjensas | TheJulia: ok, will take a look. | 17:04 |
| iurygregory | TheJulia, here is the story https://storyboard.openstack.org/#!/story/2010221 | 17:05 |
| TheJulia | iurygregory: awesome | 17:05 |
| TheJulia | I'm going to go to the post office, and then I'll review your v6 change | 17:06 |
| iurygregory | sure =) | 17:07 |
| opendevreview | kamlesh chauvhan proposed openstack/ironic-tempest-plugin master: Add iDRAC RAID cleaning steps tests https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/841601 | 17:15 |
| hjensas | TheJulia: not something I've seen before. Agree it seems it may be something openstack/automation, or a race in the filter initialization. More logs would be good. | 17:44 |
| TheJulia | iurygregory: posted review. I'm super worried over use of the bmc address | 18:03 |
| iurygregory | TheJulia, ack | 18:05 |
| JayF | iurygregory: I am very interested and curious about that separate-v6-url merge req and have opinions about it, but a few questino first: | 18:12 |
| JayF | iurygregory: do we know, for a fact, that with an otherwise properly configured network that a dual-stack'd v6 won't work? Like, I'm having trouble imagining a scenario where absent network setup issues or a BMC bug (always possible) that you'd need to setup v4/v6 URLs separately | 18:13 |
| TheJulia | My perception is this is dns-less... | 18:14 |
| JayF | (maybe ACLs disabling v4 access to anything, but still having interfaces/routes to gateway working anyway?) | 18:14 |
| * TheJulia gets out the "DNS is your FRIEND!" t-shirt | 18:14 | |
| JayF | I hadn't even considered DNS-less use | 18:14 |
| JayF | ugh, I hate there's a use case for this because I hate having "v4" and "v6" settings separately | 18:15 |
| * TheJulia also gets out a "I once deployed bind9 9.0.0b2" | 18:15 | |
| JayF | TheJulia: I think I have a one-up for you: I convinced a DJBDNS fanboy to let me migrate us to a bind cluster | 18:15 |
| TheJulia | Ohhh ahhhh! | 18:16 |
| iurygregory | JayF, the case is that when we have the need to the external url, we publish based on the configuration we have in ironic `external_http_url` (so the image is available in this URL normally is ipv4) | 18:16 |
| JayF | (a 'don't you want to be able to hire people who know what the hell is going on' argument is extremely powerful) | 18:16 |
| JayF | iurygregory: is this with or without DNS? | 18:16 |
| * TheJulia is kind of curious *how* the endpoint is getting config from ironic in the first place... | 18:16 | |
| JayF | Yeah, I want a bigger picture of the use case to review that PR properly | 18:17 |
| JayF | with DNS involved, separate v6 and v4 urls is ... not a good pattern to follow | 18:17 |
| TheJulia | yeah | 18:17 |
| iurygregory | JayF, going to double-check, but from what I remember is without DNS | 18:18 |
| iurygregory | so if we serve the URL to the node that is IPv6 only it doesn't work... | 18:18 |
| JayF | I'm saying that at least if DNS is involved, that strongly implies a misconfigured network | 18:19 |
| JayF | rather than a need for us to manually dual stack | 18:19 |
| opendevreview | Merged openstack/ironic-inspector master: Add DnsmasqFilter error - my fsm encountered an exception https://review.opendev.org/c/openstack/ironic-inspector/+/842275 | 18:19 |
| iurygregory | right | 18:20 |
| JayF | So I'm just pondering how we can support this use case without directing users towards a bad pattern | 18:20 |
| TheJulia | well, *if* it was neutron integrated | 18:21 |
| TheJulia | we could jsut ask neutron | 18:21 |
| iurygregory | their argument when I was trying to understand the use case was: "we can serve the image from many places, but we can only pass one URL to the BMC. currently there is only 1 config option in Ironic. So we have to change Ironic." | 18:21 |
| TheJulia | but in this case, we can't.... | 18:21 |
| JayF | iurygregory: the place we'd normally do such overrides is in the node object, right? | 18:21 |
| JayF | iurygregory: so you'd have a global URL set, and you could override it per-node? | 18:21 |
| JayF | Hmmm. | 18:22 |
| iurygregory | hummm | 18:22 |
| JayF | It *is* weird of us to assume one URL is good everywhere. | 18:22 |
| JayF | Which is possible to make work but I could absolutely see some folks bristling against doing | 18:22 |
| TheJulia | I was htinking the same thing in my review | 18:22 |
| iurygregory | "one url to rule them all" | 18:22 |
| iurygregory | XD | 18:22 |
| TheJulia | One url to rule them all... is like.... saying there is one ring, when in reality we need a ring for each finger | 18:22 |
| TheJulia | And *then* we need to work on our hand coordination to make music with them | 18:23 |
| * TheJulia shows herself the door | 18:23 | |
| iurygregory | wondering if would be overkill to override for the node... | 18:23 |
| JayF | Override per node is the way it "fits" into an existing model, but I realize is really ugly in practice | 18:23 |
| JayF | I almost wish we could make Ironic aware of network domains | 18:24 |
| iurygregory | but we would need to publish the image by checking the node information if there is an override right? | 18:24 |
| iurygregory | instead of just trying to guess if they have IPv6 address... | 18:24 |
| JayF | like BMC network domain = v6-section-of-network then you go look in the url-map = {v6-section-of-network: https://[dead:beef:cafe::]/images} | 18:24 |
| TheJulia | I like the "tell us the address which should work" path | 18:24 |
| iurygregory | yeah | 18:24 |
| JayF | are you saying like, there'd be a node.internal_driver_info['address_to_get_image'] | 18:28 |
| JayF | likely named better | 18:28 |
| JayF | and you set that on nodes that have a nondefault address | 18:28 |
| JayF | that fits with our current pattern, but as a former operator of ironic at scale, this would be a GIGANTIC PAIN if I ever had to change them in bulk | 18:28 |
| iurygregory | correct | 18:28 |
| JayF | That's probably the least controversial route in the short term, but I really wish there was a way for us to scope in some idea of network domain to try and tackle this problem overall | 18:29 |
| JayF | iurygregory: TheJulia: hmm. I wonder if conductor group is a reasonable place to separate settings like this | 18:29 |
| JayF | conductor group *usually* implies some kind of physical separation, right? but I know not all users enable groups | 18:30 |
| iurygregory | hummmm | 18:30 |
| opendevreview | Julia Kreger proposed openstack/ironic-tempest-plugin master: WIP: Initial tempest test idea anaconda deploy https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/835917 | 18:30 |
| iurygregory | but can we override the external url based on conductor? | 18:30 |
| TheJulia | so conductor group are hash ring delineators | 18:31 |
| JayF | That ability doesn't exist today... but neither does any of the things we've talked about, right? | 18:31 |
| iurygregory | JayF, right =) | 18:31 |
| TheJulia | they can mean... anything else to the user | 18:31 |
| JayF | are conductors allowed to be in >1 group? | 18:31 |
| TheJulia | so, fwiw, I need simple pxe code fix https://review.opendev.org/c/openstack/ironic/+/852201 to make anaconda be happy | 18:31 |
| TheJulia | JayF: no | 18:31 |
| JayF | yeh, so probably not a great fit then | 18:32 |
| JayF | looking julia | 18:32 |
| TheJulia | JayF: thanks! | 18:32 |
| JayF | TheJulia: +2A | 18:33 |
| TheJulia | awesome, thanks | 18:35 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: OpenConfig YANG Model, python-bindings releasenote https://review.opendev.org/c/openstack/networking-baremetal/+/852468 | 18:40 |
| iurygregory | soo... do we have a better idea? .-. maybe having a field in the node that would tell if is ipv6? (and in this case if the config is set we publish the image using that value for the external url?) | 18:44 |
| JayF | My primary issue with the PR, as written, is that it's explicitly written for v4 vs v6 | 18:53 |
| JayF | when this exact same issue could occur via network partitioning, completely unrelated to v4/v6/dual stack | 18:53 |
| JayF | and I'd strongly prefer our solution not encode v4/v6 in the names and proscribed solution, because I think that'll be extremely misleading to folks | 18:53 |
| iurygregory | so having on the node level something to override would solve the problem for "v4/v6" but when you look at scale this would be a big problem .-. | 18:59 |
| JayF | which is why I've been proposing larger options | 18:59 |
| JayF | e.g. node.internal_driver_info.bmc_network_domain = my-fancy-network-domain | 18:59 |
| JayF | then having my-fancy-network-domain map appropriate values such as, e.g. hostname to fetch images from | 19:00 |
| JayF | that way you only have to update it when physical moves happen, and you could update all nodes in a given domain at once to a new hostname | 19:00 |
| iurygregory | JayF, can you add some of this in the patch? o/ | 19:03 |
| JayF | yep, I wanted to have the chat first | 19:03 |
| JayF | I try not to like, -1 a patch with a lot of feedback without a chat first | 19:03 |
| JayF | and TBH I've been splitting attention, trying hard to get my armada-upstream thing done so I can focus on giving your patch a full review | 19:04 |
| iurygregory | JayF, feel free to -1 | 19:06 |
| iurygregory | I totally understand your points | 19:07 |
| JayF | I likely will with some of the comments put here :D | 19:07 |
| JayF | I just like having the chat first | 19:07 |
| iurygregory | ++ =) | 19:07 |
| *** tosky_ is now known as tosky | 19:25 | |
| *** undefined is now known as Guest7733 | 19:30 | |
| JayF | Aight, it's posted there with a -1. Most of my -1s I'm usually OK with being overridden, I'd strongly prefer getting a look at this patch and flipping my vote before any version of it merges. | 19:31 |
| *** Guest7733 is now known as rcastillo | 19:31 | |
| iurygregory | JayF, tks! dtantsur rpittau (FYI so you can check in your morning some of the discussion and also the patch) | 19:41 |
| opendevreview | Merged openstack/python-ironicclient master: Remove unused check_python_version https://review.opendev.org/c/openstack/python-ironicclient/+/850102 | 19:54 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Device management driver iface https://review.opendev.org/c/openstack/networking-baremetal/+/835170 | 20:28 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add netconf-openconfig device driver https://review.opendev.org/c/openstack/networking-baremetal/+/835324 | 20:28 |
| opendevreview | Merged openstack/ironic master: Fix pxe image lookups https://review.opendev.org/c/openstack/ironic/+/852201 | 20:32 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/851802 | 23:39 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!