15:00:25 <TheJulia> #startmeeting ironic
15:00:26 <openstack> Meeting started Mon Jun 22 15:00:25 2020 UTC and is due to finish in 60 minutes.  The chair is TheJulia. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:27 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:28 <TheJulia> Good morning everyone
15:00:29 <stendulker> o/
15:00:29 <erbarr> o/
15:00:30 <dtantsur> o/
15:00:30 <openstack> The meeting name has been set to 'ironic'
15:00:31 <iurygregory> o/
15:00:34 <rpioso> \o
15:00:38 <kaifeng> o/
15:00:41 <rpittau> o/
15:00:58 <arne_wiebalck> o/
15:01:06 <rajinir> o/
15:01:33 <TheJulia> Our agenda can be found on the wiki, as always
15:01:35 <TheJulia> #link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting
15:02:21 <TheJulia> I've noticed someone keeps putting in release countdown related stuff for openstack and I'm honestly feeling like it is a bit of noise for us.
15:02:43 <TheJulia> If someone wants to talk to me about that, I would be happy to have that chat
15:02:46 <TheJulia> Anyway!
15:02:48 <TheJulia> #topic Announcements / Reminder
15:03:03 <TheJulia> Someone apparently wishes us to know where we are in the release countdown.
15:03:10 <dtantsur> was not me this time
15:03:15 <rpittau> TheJulia: that was me
15:03:28 <dtantsur> what I do want to point out is that next week we need to do our first coordinated releases
15:03:33 <rpittau> I understand it's not important for us, ok
15:03:35 <TheJulia> #info This week is R-16, Vitoria-2 milestone (which doesn't really apply to us), is July 30th
15:03:44 <rloo> o/
15:03:49 <TheJulia> \o rloo
15:03:50 <TheJulia> rpittau: ok
15:04:29 <cdearborn> o/
15:04:31 <TheJulia> The cycle priorites change is still up in review, I believe we'll merge it today since it has had +2's for over a week now.
15:04:34 <TheJulia> #link https://review.opendev.org/#/c/720100/
15:04:35 <patchbot> patch 720100 - ironic-specs - Victoria Cycle Priorities - 9 patch sets
15:04:51 <iurygregory> ++
15:04:56 <TheJulia> #info The partition discussion from last week generated some notes.
15:04:58 <TheJulia> #link https://etherpad.opendev.org/p/ironic-disk-partitioning-2020
15:05:12 <TheJulia> Does anyone have anything to remind us of this week? Anything to announce?
15:05:32 <dtantsur> the opendev event next week?
15:05:40 <TheJulia> Oh yes!
15:05:55 <dtantsur> #link https://www.openstack.org/events/opendev-2020/opendev-schedule
15:06:01 <TheJulia> #info The first virtual OpenDev event starts next week!
15:06:12 <dtantsur> There is a topic for bare metal scaling on Wednesday
15:06:29 <TheJulia> One _does_ need to register
15:06:58 <TheJulia> #link https://www.openstack.org/events/opendev-2020/
15:07:11 <TheJulia> Anything else?
15:08:14 <TheJulia> Shall we move on?
15:08:44 <iurygregory> yes
15:08:58 <TheJulia> #topic Review action items from previous meeting
15:09:17 <TheJulia> We had one action item from last week, for dtantsur to go talk to the release team. Dmitry can you summarize that for us?
15:09:23 <openstackgerrit> Kaifeng Wang proposed openstack/ironic-inspector master: Fix incorrect pxe-enabled was set during introspection  https://review.opendev.org/737129
15:09:42 <dtantsur> yep
15:10:06 <dtantsur> The biggest outcome, they don't believe it's necessary for us to change the release model to 'independent' to achieve what we agreed on
15:10:25 <dtantsur> They have been doing certain preparations in the release tooling to permit additional branches
15:10:51 <dtantsur> We do need to update it again to accommodate the desire to use 'bugfix/X.Y' naming rather than the currently implemented 'stable/X.Y'
15:11:23 <TheJulia> when you say it, do you mean the release tooling or the plan?
15:12:46 <dtantsur> Well, both?
15:13:02 <dtantsur> The current plan states stable/X.Y, there are pretty reasonable objections to doing it
15:13:25 <TheJulia> objections with-in the openstack release model confines?
15:13:33 <dtantsur> not to conflate the intentions between the traditional stable branches and we're going to add
15:13:46 <dtantsur> it's less of a tooling or model problem, more of people expectations
15:14:07 <TheJulia> from solely an openstack perspective I guess?
15:14:16 <dtantsur> the new branches, per our plan, are not going to be such an LTS things
15:14:23 <TheJulia> where as someone from the outside may be confused
15:14:34 <TheJulia> This is true
15:15:05 <dtantsur> a concern that ttx (?) voiced is that if we treat them similarly, there may be a confusion when some (minor) fixes are merged into stable/ branches, but not into bugfix/ branches
15:15:35 <dtantsur> which is fine, but needs a clear delineation between the goals
15:15:49 <dtantsur> a bugfix/ branch is not something you spend years consuming
15:15:56 <dtantsur> like it is the case with stable/ branches now
15:16:06 <dtantsur> it's just a way for us to produce bugfix (hence the name) releases
15:16:09 <TheJulia> Okay, that makes sense and that will surely happen at some point no matter what we do
15:16:53 * dtantsur needs to amend the spec with this clarification
15:17:03 <TheJulia> okay, sounds good. Thanks dtantsur!
15:17:26 <TheJulia> Moving on!
15:17:30 <TheJulia> #topic Review subteam status reports
15:17:39 <TheJulia> #link https://etherpad.opendev.org/p/IronicWhiteBoard
15:17:49 <TheJulia> Line 256
15:18:17 <TheJulia> I do anticipate changing this up sometime this week for our next meeting based on the priorities document
15:22:16 <TheJulia> dtantsur: w/r/t the deploy steps work, I saw your comment on interface compatability, and I left a comment regarding maybe posting a release note to current and prior branches so there is sufficient heads-up.
15:22:40 <dtantsur> mmmm, I'll definitely add a note for master
15:22:56 <dtantsur> not sure how to make a note on stable branches actionable
15:23:33 <TheJulia> "Coming soon to the next release" would at least provide a heads up that the next release, if your inheriting internal interfaces, things may require additional work.
15:24:00 <TheJulia> iurygregory: only two patches left \o/ At least Ironic side
15:24:26 <iurygregory> TheJulia, yes
15:24:30 <iurygregory> \o/
15:24:56 <iurygregory> I will probably need to rebase if we merge the autospec for the tests (I saw rpittau comment)
15:24:56 <TheJulia> hjensas: Any update w/r/t the notifier?
15:25:46 <TheJulia> derekh: We have a v6 job now, is the only thing we really need to do at this point is make it voting?
15:26:49 <TheJulia> iurygregory: is the multinode grenade job still disliking us?
15:26:55 <derekh> TheJulia: yup, I made it non voting, I havn't looked at its reliability but can check it out to see if it could be voting
15:27:05 <TheJulia> derekh: Sounds good!
15:27:16 <derekh> ack will do
15:28:03 <iurygregory> TheJulia, yes =(
15:28:17 <TheJulia> I think that is it, at least for me. Are we ready to move ahead to deciding priorites?
15:28:47 <rpittau> iurygregory: I think it would be worth to merge the autospec first, or rebase the patch on top of that, we're going to enforce autospec anyway
15:29:22 <iurygregory> rpittau, yup (but since I didn't got any feedback on the patch I will wait before rebasing on top)
15:29:41 <iurygregory> the multinode fails on the tests with "'Unable to complete operation on network 31ae582c-98fd-4b35-a03b-dbea589e8cee. There are one or more ports still in use on the network.'," =(
15:29:46 <iurygregory> still trying to debug that
15:30:30 <TheJulia> are we expecting the ports to be torn down immediately or are we retrying?
15:31:20 <iurygregory> going to double check the way the test is done, because we run the same set of tests in the multinode job
15:31:23 <iurygregory> and its fine
15:32:22 <TheJulia> We already know neutron may take some time to tear down ports/networks
15:32:42 <iurygregory> yeah
15:33:04 <TheJulia> Are we good to proceed?
15:33:28 <dtantsur> ++
15:33:31 <iurygregory> ++
15:33:35 <TheJulia> #topic Deciding on priorities for the coming week
15:33:40 <TheJulia> #link https://etherpad.opendev.org/p/IronicWhiteBoard
15:33:52 <TheJulia> Starting at line 123
15:33:57 * TheJulia removes merged items
15:34:56 <TheJulia> Should we remove the oslo.privsep change for now?
15:36:11 <iurygregory> yes
15:36:13 <openstackgerrit> Verification of a change to openstack/ironic failed: Add function definition handling  https://review.opendev.org/704488
15:36:14 <TheJulia> Line 171-172
15:36:14 <TheJulia> okay
15:36:27 <iurygregory> this week Im working more on downstream
15:36:42 <TheJulia> okay
15:37:00 <TheJulia> Is there anything that we should add this week?
15:37:20 <TheJulia> WSME patches seem good to try and move forward on
15:37:22 <kaifeng> i have one: https://review.opendev.org/737129
15:37:23 <patchbot> patch 737129 - ironic-inspector - Fix incorrect pxe-enabled was set during introspec... - 4 patch sets
15:37:59 <TheJulia> Added
15:38:06 <kaifeng> thanks :)
15:39:26 <TheJulia> Anything else?
15:39:35 <TheJulia> I noticed a few more items were added to improvements
15:39:41 <dtantsur> yep, added 2 fixes
15:40:00 <TheJulia> Do we formally want to call them bug fixes?
15:40:21 <dtantsur> I'm not sure if it matters too much
15:40:25 <TheJulia> ok
15:40:39 <TheJulia> well, things explicitly called bug fixes are easier to track to backport if needed
15:41:44 <dtantsur> done
15:41:52 <TheJulia> \o/
15:42:02 <TheJulia> Anything else? Are we good with the list?
15:42:08 <TheJulia> Or shall we move on to take over the world?
15:42:33 <dtantsur> take over the world \o/
15:42:38 * kaifeng wants to take over machines by unified images
15:42:44 <iurygregory> take over the world =D
15:43:11 <TheJulia> kaifeng: unified images? how so?
15:43:58 <kaifeng> I am expecting a pointer for uefi/bios compatible image :)
15:44:07 <TheJulia> Ahh, yes
15:44:12 <kaifeng> we can discuss in the open discussion
15:44:17 <TheJulia> Okay
15:44:31 <TheJulia> Moving on, since we don't have any explicit discussion items, I guess we'll stop by the Baremetal SIG
15:44:36 <TheJulia> #topic Baremetal SIG
15:44:42 <arne_wiebalck> yes!
15:44:43 <TheJulia> arne_wiebalck: news to report? :)
15:44:50 <arne_wiebalck> the white paper is submitted!
15:45:06 <arne_wiebalck> the foundation sends their thanks to all contributors!
15:45:14 <dtantsur> \o/
15:45:21 <cdearborn> great job on the paper everyone!
15:45:33 <TheJulia> That is great news! Thanks everyone!
15:45:34 <arne_wiebalck> I think it has turned out quite well in the end
15:45:39 <arne_wiebalck> Thanks!
15:45:57 <arne_wiebalck> That was it for the SIG I think[
15:46:10 <rloo> thx arne_wiebalck and everyone else for doing that!
15:46:14 <rpittau> nice :)
15:46:37 <TheJulia> Awesome, Well then I believe our *checks notes* next section is Open Discussion since we have no items under RFE review.
15:46:46 <jamesdenton> TheJulia Do you have any cycles available to check out https://review.opendev.org/#/c/557051/? Added a few more tests based on your last feedback.
15:46:47 <patchbot> patch 557051 - networking-generic-switch - Support multiple links in link_local_information d... - 11 patch sets
15:46:53 <TheJulia> #topic Open Discussion
15:46:59 <TheJulia> jamesdenton: Awesome, thanks!
15:47:09 <TheJulia> kaifeng: so unified images!
15:47:46 <kaifeng> TheJulia arne_wiebalck if memory serves, it seems Arne is doing so already
15:47:55 <arne_wiebalck> https://techblog.web.cern.ch/techblog/post/bios_uefi_cloud_image/
15:48:16 <arne_wiebalck> this is a post where we share what we do to have a single image
15:48:51 <kaifeng> The thing is when I wrongly deploy a bios image to the uefi machine, it appears like a bmc issue since there is harddisk in the boot option
15:49:23 <kaifeng> arne_wiebalck: many thanks!
15:49:43 <kaifeng> s/there is/there is no/
15:50:47 <arne_wiebalck> even with the single image we still need to make sure the boot node is correct, at least in our deployment
15:51:33 <arne_wiebalck> correct meaning, the node and the pxe infra need to agree (and Ironic is the mediator, kind of)
15:51:34 <TheJulia> kaifeng: no hard disk option at all?
15:51:46 <TheJulia> I guess it is looking for a UEFI partition?
15:51:49 <kaifeng> so it can't boot in both mode?
15:52:14 <kaifeng> TheJulia: yep, 17 boot devices but no hard disk, it seems uefi firmeware is quite smart
15:52:21 <arne_wiebalck> kaifeng: you mean after the installation?
15:52:36 <TheJulia> kaifeng:... "impressive"
15:53:09 <arne_wiebalck> kaifeng: like, dump the BIOS/UEFI image and then it does not matter how the node boots?
15:53:12 <kaifeng> arne_wiebalck: yes, it's technically possible for an image boot from bios or uefi after installation?
15:53:43 <arne_wiebalck> kaifeng: tbh, I never tried, but it is an interesting question
15:54:05 <arne_wiebalck> for sure this does not work with s/w RAID, since it installs only one bootloader
15:54:31 <arne_wiebalck> but for non s/w RAID, this should actually work, I don't see why not
15:54:35 <kaifeng> arne_wiebalck: I thought so when I heard the unified image, but maybe I am thinkg it wrong, will take a look on the post first.
15:55:04 <openstackgerrit> Merged openstack/ironic-specs master: Victoria Cycle Priorities  https://review.opendev.org/720100
15:55:23 <TheJulia> Well, do we have anything else to discuss this week?
15:55:27 <arne_wiebalck> kaifeng: it works for VMs, so it should work for BM as well, I just never tried to switch after installation
15:55:52 <openstackgerrit> Merged openstack/ironic-python-agent-builder master: Run ensure-pip in the build/check jobs  https://review.opendev.org/737264
15:56:08 <kaifeng> arne_wiebalck: got it, i'll try and feedback if there is something new
15:56:17 <arne_wiebalck> kaifeng: +1, thanks!
15:57:18 <TheJulia> Well, seems like we're winding down. Thanks everyone!
15:58:01 <arne_wiebalck> Thanks, TheJulia
15:58:11 <TheJulia> #endmeeting