15:00:07 #startmeeting ironic 15:00:08 Meeting started Mon Oct 21 15:00:07 2019 UTC and is due to finish in 60 minutes. The chair is TheJulia. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:09 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:10 o/ everyone 15:00:11 o/ 15:00:12 o/ 15:00:12 The meeting name has been set to 'ironic' 15:00:13 o/ 15:00:14 o/ 15:00:16 o/ 15:00:16 o/ 15:00:17 \o 15:00:23 \o 15:00:25 o/ 15:00:28 o/ 15:00:29 \o/ 15:00:32 oh, I'm finally here :D 15:00:42 \o 15:00:45 o/ 15:00:49 o/ 15:01:06 Our agenda can be found on the wiki this week. 15:01:06 o/ 15:01:07 #link https://wiki.openstack.org/wiki/Meetings/Ironic 15:01:12 #undo 15:01:13 Removing item from minutes: #link https://wiki.openstack.org/wiki/Meetings/Ironic 15:01:21 #link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting 15:01:34 o/ 15:01:34 It is our normal agenda, so this should go fairly quick. 15:01:48 #topic Announcements / Reminders 15:02:23 I think the only thing really worth mentioning is that Summit is coming up, so I wonder if we even need to have a meeting that week. 15:02:28 o/ 15:02:33 Does anyone have anything else to mention or remind us of? 15:02:58 Ussuri is open for development, hack-hack-hack! :) 15:03:05 oh yes, we encourage doing so! 15:03:16 Crazy ideas are always welcome! 15:04:04 I guess we can carry on 15:04:17 * kaifeng wonders capturing a live instance sounds like crazy 15:04:57 kaifeng: I believe I have snapshots as a topic for summit :) 15:05:14 #topic Review action items from previous meeting 15:05:39 that's great :D 15:05:49 Nothing really to mention in terms of action item follow-up. One item was to sort out sushy stuffs, which the path we thought... ended up being a path we couldn't take. 15:05:55 Moving on! 15:06:01 #topic Review subteam status reports 15:06:06 I have one item from previous meeting 15:06:19 mkrai_: oh yes! 15:06:22 #action mkrai to check Python testing documentation and let us know when we can begin removing python2 15:06:25 this one 15:06:34 yes, what is the update? 15:06:41 I did some investigation and prepared an etherpad https://etherpad.openstack.org/p/ironic_py2_deprecation 15:07:10 So Jan 1, 2020 is the deadline for dropping py2 support 15:07:28 I also wrote down some py2 porting that might be needed in Ironic 15:07:39 #undo 15:07:40 Removing item from minutes: #action mkrai to check Python testing documentation and let us know when we can begin removing python2 15:07:44 hmm 15:07:46 #undo 15:07:47 Removing item from minutes: #topic Review subteam status reports 15:08:16 FWIW the json module replacement is unlikely to be really needed 15:08:26 The list is not complete yet, I will continue my investigation and keep updating etherpad with items that needs to be done in support of dropping py2 15:08:42 Okay, so I suspect we're going to need to review this. The fact our integration CI jobs work seems to make me think that we should largely be good, but we just found a possible issue this past week with slight behavior differences. 15:09:08 truth 15:09:13 We should migrate the service projects ASAP, since they have a lot of dependencies 15:09:20 and I think Nova is about to drop Python 2 support 15:09:28 (ditto openstacksdk, but I stopped it) 15:09:55 Okay, well, if we can focus on this sort of stuff this week, I think that would be best for the community 15:10:21 dtantsur, which service projects do you mean? 15:10:23 Which means I'll add a category for py2 deprecation patches and anyone can add to that list on the etherpad 15:10:27 sound good? 15:10:34 mkrai_: any services: ironic, ironic-inspector 15:10:41 TheJulia, yes 15:10:50 while python-ironicclient and company should wait until the consumers (e.g. tripleo) catch up 15:11:08 ++ 15:11:14 Are we good to proceed? 15:11:20 one note on this 15:11:30 we aren't required to *drop* py2 support by jan 1 15:11:41 that's just when python 2 is going away 15:11:48 we have till Dec 13 to drop I think 15:11:49 https://etherpad.openstack.org/p/drop-python2-support 15:12:17 I think we're looking at that wrong 15:12:26 jroll is absolutely correct 15:12:28 we aren't required to drop python 2 support 15:12:34 yep, unmaintained doesn't mean it's not available 15:12:53 exactly 15:13:02 yep 15:13:05 we should try to roughly follow that schedule, but it isn't a hard deadline 15:13:10 Oh yes!it's python2 retirement date. 15:13:12 just keeping it doesn't have any benefits 15:13:14 that being said, if any of our dependencies drop the support, we're blocked 15:13:20 jroll, Thanks for the correction! 15:13:22 we already have this situation with python-zeroconf 15:13:41 dtantsur: well, not blocked, it won't hurt to leave py2 support in even if the dependencies drop it 15:13:46 it just won't work 15:14:05 meaning we don't have to drop everything to go work on dropping py2 15:14:08 that means that we can't have testing with that anyway 15:14:13 so we shold still have CI to cover py2 in Ussuri? 15:14:20 iurygregory: I don't think we'll able to 15:14:20 anyway, I think we're on the same page, just wanted to be sure :) 15:14:30 dtantsur: ++ 15:14:43 if Nova drops Python 2 support, our CI will go down, unless we switch to Python 3 (similarly with other critical dependencies) 15:14:50 iurygregory: I'd prefer to move all integration jobs away from python2 15:15:21 TheJulia, agree, i was also thinking about the integration test we run in ironic-tempest-plugin or rocky I think 15:15:24 I believe at some point we'll have to drop py2, so IMHO priority should be in migrating CI to py3 only 15:15:44 I thought all our CI jobs had equivalent py3 already :-( 15:16:10 rloo: not all of them, I believe 15:16:36 rpioso: yes, that is what it sounds like. 15:16:42 not all of them because it would have doubled the number of running jobs in our CI which would have been a disaster 15:16:56 Also some patches to migrate additional jobs didn't go anywhere 15:16:59 in review that is 15:17:21 well, we could start removing the py2 when the equiv p3 one is working. 15:17:44 rloo: I agree, no need to keep both py2 and py3 running at this point 15:18:08 rloo: most jobs have been able to be changed without any problems, turns out the standalone job is the ?only? job being problematic 15:18:24 * dtantsur is worried about grenade (as always) 15:18:34 only for BaremetalIpmiAnsibleWholedisk =) 15:19:25 So anyway, we can continue on trying to move the jobs, trying to identify areas that need fixes, and proceed from there 15:19:25 iurygregory, while investigating I found some reference for porting in ansible playbook as well, can that be an issue? 15:19:57 looking at our whiteboard, I guess since we have 'Python 3', I had assumed (wrongly) that we had everything on some py3.x (L377). https://etherpad.openstack.org/p/IronicWhiteBoard 15:19:58 mkrai_, "maybe" can you point to me the reference please? 15:20:14 could we add a new section/topic for making sure all our CI is on py3.x ? 15:20:24 iurygregory, yes I will do add it in the etherpad 15:20:32 mkrai_, Ty! 15:21:38 rloo: no real need, I think what is needed is for the patches iurygregory has been working on to be listed in that general topic 15:22:10 I think this should have been a discussion topic 15:22:14 * TheJulia sighs 15:22:15 TheJulia: well, whatever works to make it clear which tests still are not on py3. 15:22:41 rloo: I think part of the issue was there was an anticipation of the devstack default changing one day, and that got blocked 15:22:55 shall we move on then? 15:23:00 ++ 15:23:15 #topic Review subteam status reports 15:23:58 * iurygregory added the link in the Python3 section 15:24:02 we did not move on :) 15:24:17 arne_wiebalck: Do you know when you'll be able to revisit the rootfs uuid work? 15:24:20 dtantsur: yeah :\ 15:24:29 TheJulia: this week 15:24:34 arne_wiebalck: awesome 15:24:51 dtantsur: any remaining work on the ironic-python-agent-builder bootstrapping item you have on the whiteboard? 15:25:25 TheJulia: there are always small improvements, but the main task is done 15:25:30 i.e. the bootstrapping itself 15:25:38 dtantsur: do you consider that we can remove it from the whiteboard then? 15:25:49 as a separate priority? yep 15:25:54 okay, removing 15:26:27 something I forgot about is bifrost. I'll try to update it this week. 15:26:27 dtantsur: can you update line 361 on the etherpad? 15:26:40 #link https://etherpad.openstack.org/p/IronicWhiteBoard 15:26:42 sure 15:26:45 Thanks 15:27:12 I'm good to proceed to priorities for the week, everyone else good to proceed? 15:27:31 ++ 15:27:51 #topic Priorities for the week 15:27:54 #link https://etherpad.openstack.org/p/IronicWhiteBoard 15:28:05 Line 153 15:28:09 Removing merged items 15:29:19 And proposed items moved up 15:29:26 Is everyone good with this list for this week? 15:30:06 Ilya Etingof proposed openstack/ironic master: Add virtual media boot section to the docs https://review.opendev.org/689822 15:30:07 looks good 15:30:10 yep 15:30:15 Awesome 15:30:21 do we have in-band deploy steps on anyone's radar? 15:30:28 like mgoddard's? :) 15:30:32 mkrai_ I think 15:30:59 needs a spec 15:31:06 dtantsur, Yes the reconfigure spec got a +2 15:31:18 waiting for more reviews 15:31:21 we may also be able to help with this 15:31:23 that's a different feature 15:31:41 somewhat, they intertwined I think 15:31:48 mm, no, I'm talking about having in-band deploy steps at all 15:31:56 mkrai_ has a spec about re-applying deploy steps 15:31:59 they're barely related 15:32:08 (in-band deploy steps cannot be reapplied) 15:32:39 arne_wiebalck: that would be great 15:32:39 dtantsur, got it. Thanks! 15:32:40 I still think they are intertwined as in they will merge conflict each other most likely 15:32:51 well, that's likely :) 15:33:07 my point is that without in-band support deploy steps are not complete 15:33:23 is anything ever complete? :p 15:33:26 Okay, if we're good to proceed, I have no update on the baremetal logo program or whitepaper. We had no discussion items, so I think we're good to proceed to open discussion 15:33:29 TheJulia: oh, let's put https://review.opendev.org/#/c/689751/ (and backports) on priorities 15:33:29 but +1 15:33:29 patch 689751 - ironic - Mask secrets when logging in json_rpc - 1 patch set 15:33:30 Will Szumski proposed openstack/tenks master: Do not install Open vSwitch https://review.opendev.org/689347 15:33:40 dtantsur: oh yes 15:33:46 dtantsur: can you add please ? :) 15:33:59 Thanks 15:34:11 #topic Open Discussion 15:34:21 done 15:34:39 \o/ 15:35:21 do we have https://review.opendev.org/#/c/689603 listed as a priority somewhere? 15:35:21 patch 689603 - ironic (stable/train) - Update sushy library version - 1 patch set 15:36:06 cdearborn: we do not, but I also haven't looked at the mailing list latest on it yet 15:36:40 k - just don't want that to get lost in the shuffle 15:36:40 maybe we should update global-requirements with !=1.9.0 first? 15:37:00 dtantsur: +1 15:37:28 dtantsur, seems reasonable to me 15:37:37 we likely should 15:37:49 that is the first step tony asked us to do 15:38:05 although, there is disagreement with some of that on stable branches 15:38:06 * rpioso thinks tonyb suggested that 15:38:06 but, I think we'll still need to wait for james page to respond about the UCA situation 15:38:20 if ubuntu goes to 2.0.0 we should be clear to go 15:38:28 indeed 15:38:33 So only time will tell 15:39:55 doesn't hurt to have the PR all lined up :) 15:40:07 driveby -1s... 15:40:12 I think we only need an acknowledgement from UCA 15:40:20 it's not that they can/should block it 15:40:26 +1 rloo, +1 dtantsur 15:41:19 Do we have anything else to discuss today? 15:42:51 * TheJulia hears crickets 15:43:47 hi jroll, do you have plan set this cycle for the polling mode deployment? 15:44:02 kaifeng: I do not, sorry 15:45:18 np, I may have some time for the feature this cycle 15:45:40 kaifeng: good to know 15:46:03 :) 15:46:08 Thanks everyone! 15:46:23 thanks! 15:46:32 thanks! 15:46:35 o/ 15:46:40 #endmeeting