| JayF | TheJulia: you have any of those drac details? | 00:05 |
|---|---|---|
| opendevreview | Steve Baker proposed openstack/ironic-specs master: 2024.1 vPTG Planning Notes: Ironic Work Items https://review.opendev.org/c/openstack/ironic-specs/+/899297 | 01:26 |
| opendevreview | Steve Baker proposed openstack/ironic-specs master: 2024.1 vPTG Planning Notes: Ironic Work Items https://review.opendev.org/c/openstack/ironic-specs/+/899297 | 01:32 |
| rpittau | good morning ironic! o/ | 08:00 |
| rpittau | JayF: sounds good! | 08:00 |
| iurygregory | good morning Ironic | 11:18 |
| *** Continuity__ is now known as Continuity | 12:19 | |
| opendevreview | Julia Kreger proposed openstack/ironic-inspector master: SQLAlchemy 2.0 - Remove remenents of autocommit https://review.opendev.org/c/openstack/ironic-inspector/+/900413 | 13:50 |
| iurygregory | that was fast | 13:51 |
| iurygregory | I was about to ask here since I just saw the email lol | 13:51 |
| opendevreview | Julia Kreger proposed openstack/ironic-inspector master: [CI] Add sqlalchemy 2.0 unit testing https://review.opendev.org/c/openstack/ironic-inspector/+/900414 | 13:53 |
| dtantsur | JayF: I've created a couple more post-PTG RFEs: https://bugs.launchpad.net/ironic/+bug/2043019 https://bugs.launchpad.net/ironic/+bug/2043020 | 13:58 |
| dtantsur | the latter is really a draft, so not marking as triaged | 13:58 |
| TheJulia | iurygregory: it was much more a "take first sip of coffee" "oh, email from zigo" "rutro!" | 14:01 |
| iurygregory | TheJulia, wow | 14:02 |
| TheJulia | zigo: looks like more work might be required, the force sqlalchemy 2.0 change failed | 14:02 |
| TheJulia | sweet, unit test db mocking blows up | 14:03 |
| iurygregory | yay | 14:03 |
| TheJulia | time to let the caffeine soak in a little bit | 14:03 |
| iurygregory | ++ | 14:03 |
| opendevreview | Mark Goddard proposed openstack/networking-generic-switch stable/2023.2: Fix regression plugging 802.3ad port group https://review.opendev.org/c/openstack/networking-generic-switch/+/900316 | 14:18 |
| opendevreview | Mark Goddard proposed openstack/networking-generic-switch stable/2023.1: Fix regression plugging 802.3ad port group https://review.opendev.org/c/openstack/networking-generic-switch/+/900318 | 14:19 |
| Sandzwerg[m] | Hmmm. Do I see it right that sushy takes care of the abstraction of different redfish versions and vendor dialects in ironic? | 14:58 |
| JayF | TheJulia: there is a known issue in Oslo utils version utils | 15:02 |
| opendevreview | Derek Higgins proposed openstack/ironic master: Ensure enable_netboot_fallback writes out pxe config on adopt. https://review.opendev.org/c/openstack/ironic/+/811989 | 15:03 |
| JayF | That is impacting sqla 2 and sqla master jobs | 15:03 |
| JayF | At least as of the TC meeting yesterday | 15:03 |
| JayF | IDK if that's what you were talking about or not... | 15:03 |
| TheJulia | definitely one of the aspects | 15:04 |
| JayF | Sandzwerg[m]: I don't think it does all the abstractioning for us, based on the fact that we do have a separate redfish driver for idrac would imply to me that there is some separate ironic logic as well | 15:05 |
| TheJulia | the separate idrac driver is more so to have a combinded branded state *and* handle some fun like locking | 15:06 |
| TheJulia | Configuration jobs specifically in idrac is what I mean by locking | 15:06 |
| TheJulia | they effectively lock the BMC while they run | 15:06 |
| TheJulia | speaking of locking! sqlalchemy.exc.OperationalError: (sqlite3.OperationalError) database table is locked | 15:06 |
| Sandzwerg[m] | Hmkay. Because we have some custom tooling and automation around ironic. Now with the new hardware generation, based on Sapphire Rapids, it seems almost all of the three vendors we use use different redfish dialects than before. So we thought using some abstraction might be useful. | 15:17 |
| TheJulia | Sandzwerg[m]: eh, the vendors are remarkably consistent, at least where there is not a standard set by the dmtf | 15:17 |
| Sandzwerg[m] | I mean, maybe we can even replace some of that with ironic functionality but that will take longer than. | 15:18 |
| TheJulia | At least, without knowing what vendors, it is also hard to discuss | 15:18 |
| TheJulia | and examples of differences | 15:18 |
| masghar | JayF: Yes, I will be working on the inspector merge next cycle | 15:19 |
| Sandzwerg[m] | We have HPE, Dell and Lenovo and the colleagues told me only Dell didn't changed between our last hardware and the new version | 15:20 |
| Sandzwerg[m] | It's probably not the biggest issues but stuff like, where to get the MAC of a interface is now found at some/new/path instead of the/old/path | 15:21 |
| JayF | Sandzwerg[m]: are these moves from OEM paths to standard paths? Or something else? | 15:23 |
| JayF | Sandzwerg[m]: I wonder if some of this pain is actually moving *toward* the standard | 15:24 |
| TheJulia | That is my thought, there is also the possibility that something was changed by the DMTF, and some engineer interpretted it wrong and removed the old path | 15:25 |
| TheJulia | The only way to really work through that is on a case by case basis | 15:25 |
| Sandzwerg[m] | Might be. Prior redfish support by vendors was sometimes, ...strange. I'll try to get some concrete examples | 15:25 |
| TheJulia | ++ Good plan | 15:26 |
| JayF | masghar: ack, thanks, will mark you down on the spec :) | 15:37 |
| Sandzwerg[m] | OK, Example for HPE is /redfish/v1/Systems/1/BaseNetworkAdapters/ as old Path and /redfish/v1/Chassis/1/NetworkAdapters/ as new and for Lenovo /redfish/v1/Managers/1/VirtualMedia/EXT1/ as old path vs /redfish/v1/Systems/1/VirtualMedia/EXT1/ as new. I haven't checked the redfish standard to check if one of these is the current standard. | 15:39 |
| TheJulia | Yeah, we've recently become aware of the move of the vmedia path | 15:40 |
| TheJulia | I really wish some of them kept some compatability on that | 15:40 |
| Sandzwerg[m] | Also apparently some values used to be only INT or String but could be an array now | 15:40 |
| TheJulia | iurygregory: your working on that right? | 15:40 |
| iurygregory | planning to work on it, still haven't find time =( | 15:40 |
| TheJulia | the network adapter move is surprising, but that could be dmtf triggered | 15:41 |
| JayF | rpittau: when you say "official redfish support starts at drac 7" | 15:41 |
| JayF | rpittau: I'm trying to make a rec to opers | 15:41 |
| Sandzwerg[m] | The new version is from sapphire rapids, so regarding intel hardware the newest available | 15:41 |
| TheJulia | idrac 7?!? That seems far flung future to me given they have supported redfish on their hardware since the idrac 3 as long as you had 2.67.67.67 | 15:41 |
| JayF | rpittau: so basically, if you have an oper with say, drac {5,6,7} hardware, would they use wsman or redfish for each, based on your post (and sorta back-dating my knowledg,e assuming idrac and drac are the same major versions), I'd think =<6 is wsman, =>7 is redfish | 15:41 |
| JayF | TheJulia: note; marketing version != firmware version for drac | 15:42 |
| rpittau | JayF: redfish is officially supported by Dell starting from idrac 7, everything before that is "experimental" | 15:42 |
| JayF | TheJulia: I had idrac 5 on machines I managed a decade ago, assuming idrac 5 == drac 5 | 15:42 |
| TheJulia | rpittau: that seems... like marketing :) | 15:42 |
| JayF | I mainly want to direct operators to do the thing that's most likely to work. I only care what Dell says insofar as they know the answer to that question. | 15:43 |
| TheJulia | JayF: Maybe your thinking gen 5 idrac version 3? | 15:43 |
| JayF | *blink* | 15:43 |
| JayF | I know they were awful, ipmi+java web console stuff | 15:43 |
| TheJulia | heh | 15:44 |
| JayF | I had a special hacked up web browser to make the web console work | 15:44 |
| TheJulia | I think we need a table of idrac versions at this point | 15:44 |
| JayF | I'm going to leave it vague | 15:44 |
| TheJulia | +++++ | 15:44 |
| TheJulia | likely for the best | 15:44 |
| JayF | and plead with someone who knows the matrix to write a doc | 15:44 |
| * JayF pleads | 15:44 | |
| * JayF checks off action item | 15:44 | |
| * TheJulia notes contributors will likely drink whiskey | 15:44 | |
| rpittau | TheJulia: the first version of the firmware to have redfish support is 2.30.30.30 available starting from 12th generation, so starting from idrac 7 | 15:45 |
| JayF | how does someone give a piece of software 5 version numbers and go "this is good" [7,2,30,30,30] | 15:46 |
| rpittau | :D | 15:46 |
| * JayF needs a decoder ring | 15:46 | |
| rpittau | April 2016 btw :) | 15:46 |
| JayF | in that case, I think what I'm thinking is "DRAC5/DRAC6" is drac5/drac6 | 15:47 |
| JayF | and your definition fits with my idea of what the timeline woudl be | 15:47 |
| JayF | but I have so little context I don't know what to believe | 15:47 |
| TheJulia | shoot, i must be mixing ilo versioning and idrac versioning | 15:47 |
| TheJulia | JayF: Did you look inside of the cereal box? | 15:48 |
| * JayF blinks | 15:48 | |
| JayF | it's too early for metaphor | 15:48 |
| TheJulia | for a decoder ring | 15:48 |
| rpittau | lol | 15:48 |
| JayF | no, those come in cracker jack boxes | 15:48 |
| rpittau | JayF, TheJulia -> https://www.dell.com/support/home/en-us/drivers/driversdetails?driverid=5gchc | 15:49 |
| rpittau | "Added support for Redfish 1.0" | 15:49 |
| TheJulia | wow 2.30.30.30 makes me feel old | 15:49 |
| JayF | I think I'm mostly sold since TheJulia is waffling now and this fits with my (limited) knowledge | 15:49 |
| JayF | I only used Dells for like, half my tenure at Rackspace and only got to IPMI them :D | 15:50 |
| TheJulia | JayF: yeah, I'm definitely mixing versions up in my head | 15:50 |
| JayF | > suggestions based on specified system support or vendor recommendations | 15:50 |
| JayF | is literally what we say | 15:50 |
| JayF | so I think we are doin it right | 15:50 |
| rpittau | 👍 | 15:50 |
| JayF | aight, revision incoming once lint passes | 15:51 |
| * TheJulia fires up Rammsttein and downloads a giant payload of logs to dig into a customer case which just became the #1 priority | 15:52 | |
| JayF | TheJulia: only for 8 minutes :D | 15:52 |
| JayF | then we get to asplodes some stars | 15:53 |
| TheJulia | huh, what?!? | 15:53 |
| TheJulia | what is in 8 minutes?!? | 15:53 |
| JayF | that chat with johnthetubaguy ? | 15:53 |
| TheJulia | oh! right | 15:53 |
| * TheJulia puts on human clothing | 15:53 | |
| opendevreview | Jay Faulkner proposed openstack/ironic-specs master: 2024.1 vPTG Planning Notes: Ironic Work Items https://review.opendev.org/c/openstack/ironic-specs/+/899297 | 15:53 |
| JayF | if we wanna get a bunch of +2s on that I can land it today and send the email | 15:54 |
| JayF | tomorrow is nuts for me so I'd actually appreciate that | 15:54 |
| JayF | it's been mildly refreshing to shorten my todo list instead of grow it this week so far :D | 15:54 |
| rpittau | +2ed | 15:54 |
| JayF | dtantsur: tagged both as RFE; I'm not adding them to workstreams for now unless you think it's worthwhile | 15:59 |
| zigo | TheJulia: Thanks for working on this! :) | 16:19 |
| zigo | I wasn't expecting less from the Ironic team, that was the least of my concern. | 16:19 |
| zigo | It's a little bit more scary for other projects ... | 16:19 |
| rpittau | good night! o/ | 16:52 |
| TheJulia | zigo: stupidly close, just a bunch of other stuff popping up today | 17:00 |
| dtantsur | JayF: yeah, I'm not going to work on either, at least for now | 17:48 |
| JayF | Can someone please update https://launchpad.net/~ironic-drivers/+members | 18:40 |
| JayF | to include all existing release liasons (and me!), and remove emeritus core reviews | 18:40 |
| dtantsur | JayF: I've reactived you, now you have powerrzzz | 18:41 |
| JayF | if someone is available to do this now() it would be incredibly helpful | 18:41 |
| JayF | dtantsur: can you make me admin/ | 18:41 |
| dtantsur | yeah, just noticed and fixed | 18:41 |
| JayF | https://launchpad.net/ironic-python-agent-builder | 18:41 |
| iurygregory | powerrzzz is related to IBM? =P | 18:41 |
| JayF | aight, I'll clean up the group then, thank you | 18:41 |
| iurygregory | sorry >.< | 18:41 |
| dtantsur | :D | 18:41 |
| opendevreview | Jay Faulkner proposed openstack/ironic-python-agent-builder master: Add link to LP bug tracker https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/900446 | 18:48 |
| TheJulia | iurygregory: I think you mean POWER ;) | 18:50 |
| iurygregory | yeah XD | 18:50 |
| * jlvillal shocked JayF deleted him as an Administrator! :D | 18:52 | |
| JayF | jlvillal: I forgot to put a message in for yours, I did for everyone else | 18:52 |
| jlvillal | No worries and good to do it! | 18:52 |
| dtantsur | jlvillal: a revolt! | 18:52 |
| JayF | jlvillal: here's yours: "Your contributions as a core are appreciated; but since you are in emeritus status your launchpad access is being removed." | 18:52 |
| JayF | ;) | 18:52 |
| JayF | jlvillal: wait, hold on, you want access back?! | 18:52 |
| jlvillal | :) | 18:52 |
| JayF | JOHN VOLUNTEERED TO WORK ON IRONIC AGAIN | 18:52 |
| JayF | I HEARD HIM | 18:52 |
| jlvillal | No thank you! :) | 18:53 |
| dtantsur | \o/ | 18:53 |
| TheJulia | He did! | 18:53 |
| TheJulia | sweet! | 18:53 |
| JayF | quick, don't let him leave | 18:53 |
| JayF | :P | 18:53 |
| TheJulia | wait, this is an irc channel, he can always just time out | 18:53 |
| TheJulia | .... | 18:53 |
| jlvillal | But I possibly learned that maybe JayF is actually JasonF ;) | 18:53 |
| TheJulia | Clearly we need to move to something like SecondLife | 18:53 |
| JayF | has that been an option all along?! | 18:53 |
| *** JayF is now known as JasonF | 18:53 | |
| JasonF | I am me | 18:53 |
| *** JasonF is now known as JayF | 18:53 | |
| JayF | I didn't realize I was ever in OpenStack community as JasonF | 18:54 |
| jlvillal | Well launchpad calls you Jason... | 18:54 |
| JayF | (it's mildly humorous that I have ... digital artifacts of my IRC presence on freenode being older than my IRL nickname that *everyone I know* except my parents call me now) | 18:54 |
| JayF | oh, yeah, good point | 18:54 |
| opendevreview | Julia Kreger proposed openstack/ironic-inspector master: SQLAlchemy 2.0 - Fix items lost in migration https://review.opendev.org/c/openstack/ironic-inspector/+/900413 | 18:55 |
| TheJulia | is freenode still online?! | 18:55 |
| JayF | No; but JasonF was my original registered nick there. I changed it to JayF once my IRL name sorta developed into Jay. | 18:56 |
| JayF | I keep it registered on networks I frequent as my backup. | 18:56 |
| TheJulia | zigo: ^^^ that change above should fix things right up, aside from the known issue JayF was mentioning regarding testing on master branch. Locally it works with latest releases of oslo.db and sqlalchemy installed | 18:56 |
| opendevreview | Julia Kreger proposed openstack/ironic-inspector master: [CI] Add sqlalchemy 2.0 unit testing https://review.opendev.org/c/openstack/ironic-inspector/+/900414 | 18:58 |
| TheJulia | so, lets see, what was it that I was going to work on when I sat down this morning | 18:58 |
| JayF | I'm building the world's worst url... | 19:20 |
| JayF | https://zuul.opendev.org/t/openstack/builds?project=openstack%2Fironic&project=openstack%2Fironic-python-agent&project=openstack%2Fironic-inspector&project=openstack%2Fironic-lib&project=openstack%2Fironic-prometheus-exporter&pipeline=periodic&pipeline=periodic-stable&result=FAILURE&result=RETRY_LIMIT&result=POST_FAILURE&result=SKIPPED&result=NODE_FAILURE&result=MERGE_CONFLI | 19:20 |
| JayF | CT&result=MERGE_FAILURE&result=CONFIG_ERROR&result=TIMED_OUT&result=CANCELED&result=ERROR&result=RETRY&result=DISK_FULL&result=NO_JOBS&result=DISCONNECT&result=ABORTED&result=LOST&result=EXCEPTION&result=NO_HANDLE&skip=0 | 19:20 |
| JayF | yikes, didn't know it was two lines | 19:20 |
| JayF | but the tl;dr is that we have jobs failing all over the place for inspector bugfix/11.2 and ipa bugfix/9.2 | 19:21 |
| JayF | I think those are nearing EOL; so we should either fix or EOL them | 19:21 |
| JayF | also interesting: python-ironic-inspector-client has periodic jobs, python-ironicclient doesn't | 19:31 |
| JayF | that seems weird to me | 19:31 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Document new bug deputy role for Ironic https://review.opendev.org/c/openstack/ironic/+/900449 | 19:53 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Document new bug deputy role for Ironic https://review.opendev.org/c/openstack/ironic/+/900449 | 20:01 |
| JayF | and this is when I learn IPE has no bugtracker, too, gonna fix that | 20:05 |
| JayF | same for metalsmith | 20:06 |
| JayF | both of those still link to storyboard, it seems | 20:06 |
| JayF | and tenks | 20:06 |
| JayF | mgoddard: do you all use bugtracking anywhere for tenks? | 20:06 |
| JayF | mgoddard: I'm trying to standardize bug handling and triage across Ironic, don't wanna stop on you all if you've been using storyboard | 20:07 |
| opendevreview | Julia Kreger proposed openstack/ironic-python-agent master: Fix vmedia network config drive handling https://review.opendev.org/c/openstack/ironic-python-agent/+/895519 | 20:11 |
| TheJulia | Okay, I think that fixes that change's issues | 20:12 |
| JayF | TheJulia: just sanity check: there's no IPE/metalsmith bugtracker in LP? | 20:14 |
| * JayF about to go make them | 20:15 | |
| TheJulia | afaik no, and honestly not really sure everything needs a tracker, but whatever | 20:15 |
| JayF | that was mostly my take, then someone asked me to make one for ipa-b | 20:15 |
| * TheJulia shrugs | 20:16 | |
| JayF | yeah, that's where I am, at this point it's easier to have it as an entrypoint | 20:16 |
| JayF | and at least that bug deputy doc I'm building gives us a place to aggregate the links easily | 20:16 |
| opendevreview | Jay Faulkner proposed openstack/metalsmith master: Document LP as official bugtracker https://review.opendev.org/c/openstack/metalsmith/+/900452 | 20:20 |
| JayF | TheJulia: you want me to wait on your review of vptg spec before I land it? | 20:24 |
| JayF | steve just gave me the 2nd +2 | 20:24 |
| TheJulia | I thought I had already +2'ed it | 20:25 |
| JayF | you had, I kicked it off with an update | 20:26 |
| JayF | gonna land it, taking that as a yes | 20:26 |
| JayF | revisions are always a thing :D | 20:26 |
| * JayF is always reminded this time of year, with email summaries and etherpads being thrown around, it warms my heart that we commit a spec | 20:27 | |
| JayF | fungi: for these new LP projects; I gave ironic-drivers access, which gives openstack release team access. Is that sufficient for opendev or do I need to take addl. steps? | 20:29 |
| JayF | fungi: two new examples being bugs.launchpad.net/metalsmith and bugs.launchpad.net/ironic-python-agent-builder | 20:29 |
| opendevreview | Merged openstack/ironic-specs master: 2024.1 vPTG Planning Notes: Ironic Work Items https://review.opendev.org/c/openstack/ironic-specs/+/899297 | 20:40 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Advanced vmedia deployment test ops https://review.opendev.org/c/openstack/ironic/+/898010 | 20:41 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Revert "Unit tests: Isolate mysql test migrations" https://review.opendev.org/c/openstack/ironic/+/900454 | 20:44 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Change snmp job to not use a focal node https://review.opendev.org/c/openstack/ironic/+/893824 | 20:54 |
| opendevreview | Jay Faulkner proposed openstack/ironic-prometheus-exporter master: Document new LP bug tracker https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/900460 | 21:39 |
| opendevreview | Jay Faulkner proposed openstack/ironic-specs master: Quickfix: Correct rendering level of UEFI boot rec https://review.opendev.org/c/openstack/ironic-specs/+/900481 | 21:40 |
| JayF | someone wanna give me a quick +2A on that ^ just a render fix | 21:40 |
| TheJulia | Hey folks! So I took a look at CI jobs running, and then looked at the config, albeit not *super* deep | 21:42 |
| TheJulia | I came up with https://etherpad.opendev.org/p/ironic-ci-jobs | 21:42 |
| TheJulia | and the tl;dr is we do definitely have some cleanup to do | 21:42 |
| JayF | TheJulia: there is some interesting stuff along those lines in my bug deputy PR | 21:43 |
| JayF | TheJulia: basically links to see periodic job failures for all ironic projects | 21:43 |
| JayF | which is just convienient lol | 21:43 |
| TheJulia | The short take away is I think we should dial back ipmi testing | 21:43 |
| TheJulia | consolidate a couple things | 21:44 |
| TheJulia | *likely* have a few jobs to fix and make voting again | 21:44 |
| TheJulia | and we can likely run things like anaconda+advanced ops which is in review now for vmedia in the same CI job | 21:44 |
| JayF | https://c66b1587a27ab576fc67-06d350e249369bd6ea2abe2080d2269e.ssl.cf5.rackcdn.com/900449/2/check/openstack-tox-docs/fd6d1af/docs/contributor/bug-deputy.html#review-periodic-stable-ci-jobs | 21:44 |
| JayF | (note that last link *does work* just all those projects are green) | 21:44 |
| JayF | I'm basically trying to apply some of the same heuristics I'd apply to ask "is a project active?" from a TC chair perspective for things like, PTL appointments, and finding we've got some dark dirty corners that need some febreeze lol | 21:45 |
| JayF | not even meaning we have to fix 'em right now, but it's nice to just know what's going on | 21:45 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Document new bug deputy role for Ironic https://review.opendev.org/c/openstack/ironic/+/900449 | 21:46 |
| TheJulia | I'd be wary of adding requirements which requires a bunch of new work, fwiw | 21:47 |
| TheJulia | Granted, some might view it as table takes, but it may just depend on how big the table is and disconnects in that table's size | 21:47 |
| JayF | the doc itself basically says: 1) triage new bugs 2) run a bug jam 3) look at periodics once a week and report if new stuff kaboom | 21:47 |
| JayF | if that's too much, please comment it on the review | 21:47 |
| TheJulia | "this is a tiny table, one project1" "OMG, that table is the size of a ranch and we have whole sections hiding ponies we haven't visited in ages" | 21:47 |
| JayF | honestly if not for metal3 picking up on bugfix branches | 21:48 |
| JayF | I'd exclude them | 21:48 |
| JayF | I really, really dislike that move | 21:48 |
| JayF | bceause we do not run CI as well on those branches | 21:48 |
| TheJulia | crazy question, what purpose does the periodic stable job *actually* serve | 21:49 |
| TheJulia | who does it serve | 21:49 |
| TheJulia | why does it serve | 21:49 |
| TheJulia | etc | 21:49 |
| JayF | If the answer to those is "nobody", then we remove it, we don't let it fail and waste CI resources endlessly. | 21:49 |
| * TheJulia doesn't have a candlestick nor is hiding in the lounge | 21:49 | |
| JayF | I think the idea is to identify failures early when they are easier to fix, as those jobs do bitrot over time. | 21:49 |
| TheJulia | JayF: exactly! | 21:50 |
| JayF | If we fixed them and checked them periodically, they could serve that role. | 21:50 |
| JayF | The point is more: status quo is no good, proposing that as part of bug deputy is a shake up to the status quo | 21:50 |
| JayF | if we go the other direction it's still an improvement | 21:50 |
| JayF | I don't know what direction is right; but I know status quo is worse of the three opts I see | 21:50 |
| TheJulia | okay, I'm seriously wondering why we have any periodics, because I thought we didn't have any periodic jobs and I can't answer the who/what/why questions | 21:51 |
| TheJulia | sure, to prevent bitrot might be a why | 21:51 |
| TheJulia | but at some point, that is always to be expected | 21:51 |
| TheJulia | and it still doesn't return value without the who portion | 21:51 |
| JayF | I think about it this way: if I have to maintain stable/OLD | 21:52 |
| JayF | I want to fix the stable/OLD job on the first day it breaks | 21:52 |
| JayF | not fourteen months after it breaks and I'm trying to backport a CVE | 21:52 |
| TheJulia | at what cost of resources then | 21:52 |
| JayF | I see that as a potential efficiency play, with a net resource gain by being proactive | 21:52 |
| JayF | but maybe I also see unicorns and pots of gold :D | 21:53 |
| TheJulia | so we spend more time context switching then | 21:53 |
| TheJulia | well, in theory, to save us the gap and loss later | 21:53 |
| JayF | it's all tradeoffs. always. | 21:53 |
| TheJulia | dunno, that is a... fun tradeoff I guess | 21:53 |
| TheJulia | yeah | 21:53 |
| JayF | I just am feeling less and less okay over time with letting things lie in the corner | 21:54 |
| TheJulia | ++ | 21:54 |
| JayF | they need to be supported or they need to be gotten rid of | 21:54 |
| * JayF looks at the scary etherpad of scariness | 21:54 | |
| fungi | JayF: https://launchpad.net/~ironic-drivers has openstack administrators as its owner, so should be fine | 21:54 |
| opendevreview | Verification of a change to openstack/bifrost master failed: Restart systemd services on failure https://review.opendev.org/c/openstack/bifrost/+/884198 | 21:54 |
| JayF | fungi: cool, that works! I've been using that as the owner of all the projects, so I think we're good there | 21:55 |
| fungi | JayF: for succession planning, you might want to also make ironic-drivers the driver of those projects too, instead of leaving it as your account | 21:55 |
| TheJulia | ++ | 21:55 |
| TheJulia | That is basically the only way we've kept continuity without chasing folks down and... we all know what happens when someone disappears unexpectedly :( | 21:56 |
| * TheJulia has thought of Ilya a few times in the last couple of days | 21:56 | |
| JayF | fungi: I think I did that, but accidentally made myself the bug driver for a couple | 21:56 |
| fungi | the maintainer field should suffice, but i'm mainly suggesting not leaving personal accounts as owning anything in particular | 21:57 |
| JayF | fungi: literally when it's like "who is the maintainer of the project" I put in "ironic-drivers" group | 21:57 |
| JayF | I didn't put ironic-drivers there for bug driver simply because I didn't know if I was signing up 11 people for a crapton of email | 21:57 |
| fungi | oh, yeah either unset the bug supervisor (if you want open bug triage), or set it to a team | 21:57 |
| JayF | yeah, I'll unset it, that's what I did on the ones where I saw it was defaulting to me | 21:57 |
| * JayF can recheck the others | 21:58 | |
| fungi | JayF: also, in those projects' config pages, you can set them as "part of" openstack, which is what the other openstack deliverables do | 21:58 |
| fungi | i don't know that it does anything useful, other than show the "part of" like you see at https://launchpad.net/nova | 21:59 |
| fungi | and adds the project to https://launchpad.net/openstack | 21:59 |
| JayF | already done :) | 21:59 |
| JayF | before you even were completed explaining the benefits | 21:59 |
| JayF | (both things; I'm not bug driver; it's null, and part of->openstack) | 22:00 |
| TheJulia | masghar: dtantsur: Regarding the etherpad https://etherpad.opendev.org/p/ironic-ci-jobs I created and mentioned earlier, and the ironic-inspector jobs, any chance we can get a slice of time to figure out which of the jobs on master branch should say for inspector since we run 3, one while non-voting is consistently failing "non-standalone" | 22:00 |
| TheJulia | Also, one big question for anyone watching, has there been agreement to default to the new RBAC policies yet across the community? | 22:01 |
| JayF | Explicitly no | 22:01 |
| JayF | meaning; the agreement right now is implicitly to not default to that | 22:01 |
| TheJulia | is there consensus to ever make the default change? | 22:01 |
| JayF | knikolla would be the best person to have this conversation with | 22:02 |
| JayF | but my understanding is that the conversation has moved to either "we have to do the work ourselves to RBAC these non-migrated projects" and/or "we have to have a technical solution to support both at the same time" | 22:02 |
| TheJulia | tonyb: when is the next security meeting? | 22:02 |
| JayF | and move of the chatter I've seen was of the latter variety, and all ideas were shot down | 22:02 |
| knikolla | o/ | 22:02 |
| TheJulia | hi! | 22:03 |
| JayF | knikolla: how did you do that | 22:03 |
| JayF | you are more magical than I knew | 22:03 |
| TheJulia | so, when can I nuke or switch primary focus for rbac to the SRBAC model? | 22:03 |
| knikolla | JayF: you won't believe the amount of free time I have after you took on all my responsibilities (jk) | 22:03 |
| * TheJulia thinks of old rbac as the trinity site site with the gadget on the tower | 22:04 | |
| JayF | between actual physical weight loss, and the mental weight loss of having PTG + recaps done and my SeaGL talk done, I feel about 400 lbs lighter and like my days have 500 hours in them | 22:04 |
| * JayF got more upstream stuff done today than he did all last week combined | 22:05 | |
| tonyb | TheJulia: Ummm currently unscedhuled, but I appreciate the ping. | 22:05 |
| TheJulia | tonyb: anytime | 22:05 |
| knikolla | I haven't followed the RBAC work closely in the last cycle, that would be gmann . But with all the pings here we can host a party. | 22:06 |
| TheJulia | It might be more a question for the security sig to push the envelop forward, but the TC is going to have to be mean at some point of not all projects align | 22:06 |
| * JayF keeps trying to find the right incantation to get chatgpt to write CRs to update RBAC models in all straggler projects /s/s | 22:07 | |
| TheJulia | ... and there is always the possibility that some projects (actually, I'm sure this is the case) will have spoons or capability to do so | 22:07 |
| TheJulia | JayF: writing a policy is actually on my todo list for tomorrow | 22:07 |
| TheJulia | so. thanks! | 22:07 |
| TheJulia | wow, that medivac was moving | 22:08 |
| * TheJulia just got buzzed by a medivac helicopter | 22:08 | |
| TheJulia | ... and also crossed the runway at the airport *blink* *blink* | 22:08 |
| opendevreview | Verification of a change to openstack/bifrost master failed: ironic: Perform online data migrations with localhost DB https://review.opendev.org/c/openstack/bifrost/+/895948 | 22:15 |
| JayF | we shouldn't be approving things right now | 22:15 |
| JayF | gate is busted :( | 22:15 |
| TheJulia | when was that annoucement made? | 22:15 |
| JayF | the broken sql2x job I mentioned this morning | 22:15 |
| JayF | > ironic-tox-unit-with-sqlalchemy-2x https://zuul.opendev.org/t/openstack/build/b328ff3b97c2429faf7a1b94f9b52251 : FAILURE in 6m 41s | 22:16 |
| TheJulia | unrelated to biforst | 22:16 |
| JayF | ahh | 22:16 |
| JayF | that's sad then, that might be a real problem | 22:16 |
| JayF | > bifrost-tox-linters | 22:16 |
| JayF | WTF | 22:16 |
| JayF | yikes, I'm going to go look at why that is unhappy | 22:16 |
| JayF | I'm at a stopping point on my current task and real-technical-work instead of writing another email would be nice | 22:17 |
| TheJulia | that was why I went down the list of jobs, tbh | 22:17 |
| JayF | well tox -elinters for bifrost has an external depednency, that's not awesome | 22:18 |
| JayF | > Cannot find ansible-collections-openstack at ../ansible-collections-openstack | 22:18 |
| TheJulia | wheeeeeeeeeeeee | 22:19 |
| JayF | it looks like they added some new lint rules that broke things afaict | 22:21 |
| JayF | going to disable them because they are ... plentiful | 22:22 |
| opendevreview | Jay Faulkner proposed openstack/bifrost master: [CI] Disable new ansible lint rule breaking gate https://review.opendev.org/c/openstack/bifrost/+/900487 | 22:29 |
| JayF | there's the gate fix, we can recheck stuff once that lands | 22:29 |
| * TheJulia finally takes migraine meds | 22:29 | |
| iurygregory | JayF, I just hope the empty space wont cause problems, I've +2 https://review.opendev.org/c/openstack/bifrost/+/900487 | 23:17 |
| JayF | it passes locally | 23:17 |
| JayF | just land it? trivial ci fix, one core is permitted | 23:18 |
| JayF | it's blocking gate | 23:18 |
| JayF | and kicked two things outta gate due to timing (grr) | 23:18 |
| iurygregory | yeah sure =) | 23:18 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!