Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-01-30
14:30:25 mriedem sean-k-mooney: if it makes you feel better huawei will need to take it to mitaka
14:32:40 jangutter mriedem: thanks, checking!
14:34:05 mdbooth lyarwood: Just looking at your patch https://review.openstack.org/#/c/626952/1/nova/api/openstack/wsgi_app.py. IIUC that's basically what we do for all non-wsgi services here: https://github.com/openstack/nova/blob/master/nova/cmd/__init__.py
14:34:49 jaypipes sean-k-mooney: yikes. (on newton). I just backported shit to Ocata and know how painful that is.
14:35:10 mdbooth I'm pretty sure that means that we monkey patch while looking up the entry point, so before anything else.
14:35:57 sean-k-mooney mriedem: well if it was not for the fact that osp 7 went endo f life on august 5 2018 we would have had to backport to kilo
14:36:18 mdbooth 'end of life'
14:38:24 mriedem tssurya: regarding https://review.openstack.org/#/c/614167/ cern might want to investigate that up front before running it on all cells since it iterates all cells but is intended to migrate cell-specific data
14:38:40 mriedem which is kind of different from how we do other online data migrations which only hit the cell db
14:38:41 sean-k-mooney jaypipes: osp 10/netwon is an lts release which will be supported until decmber 15 2021 alther it will move to extend maintant at the end of the year but ya we have to backport to newton faily commonly at present
14:39:03 mriedem tssurya: it uses a marker though so once it runs for all cells on the first run it should be a noop if run on each individual cell,
14:39:20 mriedem which i'm assuming cern does as part of upgrades - run online_data_migrations per cell db config via config management
14:39:43 mdbooth And, on occasion, to Kilo still, despite being 'end of life' :/
14:41:00 sean-k-mooney mdbooth: well there is always the $$$ factor that makes "end of life" more flexible then we might like :)
14:41:04 mdbooth lyarwood: Ah, your patch touches that code too. And.... I'll bet the reason is that nova.utils imports keystoneauth1, which means we've already imported it before running the monkey patching.
14:41:22 mdbooth sean-k-mooney: Indeed.
14:42:54 mriedem jaypipes: i +2ed https://review.openstack.org/#/c/614167/ since i was the one hammering on the limit processing and tests and such
14:42:55 mdbooth lyarwood: I think I'll pull the monkey_patch util method into a separate module which doesn't import anything.
14:44:33 mdbooth This is a fix to change Ie7bf5d012e2ccbcd63c262ddaf739782afcdaf56
14:45:37 jaypipes mriedem, sean-k-mooney: done.
14:46:22 sean-k-mooney maciejjozefczyk: ^
14:56:22 dpawlik jaypipes, cdent, dansmith: Hi, could you check https://review.openstack.org/#/c/628956/ and https://review.openstack.org/#/c/628128/ ? Thanks
14:57:01 cdent ayup, thanks
15:05:41 lyarwood mdbooth: ack yeah feel free to rewrite it and also take over authorship of the actual commit
15:08:20 mdbooth lyarwood: Don't think I can do that, in gerrit at least
15:10:19 lyarwood mdbooth: right, that's why I said commit ;)
15:10:42 lyarwood mdbooth: just git amend with --author
15:11:17 lyarwood ah wait that's a local alias
15:11:45 lyarwood git commit --amend --author=foo
15:11:59 mdbooth lyarwood: Ok, I can take the git commit, but it'll still be yours in gerrit I think. Shouldn't matter, anyway.
15:12:09 mdbooth lyarwood: What's the relevant test for this, btw?
15:13:11 lyarwood mdbooth: ykarel in #tripleo should be able to help with that, iirc there's a fedora28 py3 undercloud job deploying with SSL somewhere.
15:14:09 mdbooth I'm just running functional on my version, and I've run unit tests matching 'wsgi', but that seems little better than a smoke test
15:15:46 tssurya mriedem: thanks for bringing it to our notice, looking into right now
15:23:23 mriedem jangutter: replied in https://review.openstack.org/#/c/572082/
15:23:36 mriedem jangutter: looks like just add a todo and tweak the release note and i'm +2
15:24:01 jangutter mriedem: thanks for risking your sanity!
15:24:17 maciejjozefczyk mriedem: jaypipes thanks
15:26:38 mriedem gibi: are you still around today? just wondering if you want me to address my own nits in your series to keep it moving, or just start my own FUP patch (assume you'd prefer the latter)
15:26:50 gibi mriedem: today I'm still here
15:27:10 gibi mriedem: and working on an update but I might not finish it today
15:27:47 mriedem ok because i'm mostly good up through https://review.openstack.org/#/c/619527/ (just looking for replies) and i'm happy to address nits myself
15:28:55 gibi mriedem: then I will switch to reply mode to see if we agree and I let you propose fups while I'm away
15:29:46 mriedem need another core on mel's configurable max disk devices change https://review.openstack.org/#/c/616777/ which will close that bp
15:34:17 gibi mriedem: replies in https://review.openstack.org/#/c/630722
15:38:57 openstackgerrit Matthew Booth proposed openstack/nova master: Fix SSL infinite recursion running WSGI app under python 3.7 https://review.openstack.org/626952
15:39:02 mdbooth lyarwood: ^^^
15:39:32 mdbooth lyarwood: It's your change but with its own module and an expanded commit message
15:39:59 mdbooth And some comments
15:41:29 gibi melwitt, mriedem: I'm +A on https://review.openstack.org/#/c/616777/
15:42:20 mriedem thanks
15:42:23 mriedem i'll fix those typos
15:46:09 mriedem mdbooth: lyarwood: issues in that change
15:47:13 openstackgerrit Matt Riedemann proposed openstack/nova master: Add configuration of maximum disk devices to attach https://review.openstack.org/616777
15:56:16 mdbooth mriedem: Thanks. Principal issue there was changes to monkey_patch() since the version I copied. Updating shortly.
15:59:31 mriedem yeah i figured
16:01:02 mriedem http://status.openstack.org/elastic-recheck/#1813198 has magically dropped off and no one knows why
16:09:39 gibi mriedem: most of those graphs on that page are dropped off around the same time
16:09:54 gibi mriedem: I think we don't have data in elastic re-check
16:10:34 cdent gibi: that reminds me: does that burndown stuff I'm hosting still matter? If so, it's not causing any issues, but if not I supposed I could clean it up.
16:11:26 mriedem gibi: gah good point
16:13:15 gibi cdent: good point. thank you for hosting that you can delete the app
16:13:30 mriedem all of the legacy notifications have been converted yeah?
16:13:39 gibi mriedem: yes
16:13:45 mriedem only open question is what to do about the default in the config option for format
16:13:46 gibi mriedem: a final reno was merged
16:13:54 cdent ✔
16:13:56 gibi mriedem: yes, deprecation is still open
16:13:58 mriedem we could probably close the bp
16:13:58 gibi cdent: thanks
16:14:08 gibi mriedem: I think melwitt closed the bp
16:15:32 cdent gibi: all dead. content is still there if it's ever needed again
16:15:46 gibi cdent: awesome, thanks
16:24:18 tinwood dansmith, hi, I have a review up to add LXD to the HypervisorDriver() enum: https://review.openstack.org/#/c/633279/ -- I bumped the version number (not sure it really needed it!) and was wondering if you might have a moment to take a look please?
16:33:30 mdbooth lyarwood: mriedem asks why Related rather than Closes on those 2 bugs. I wondered the same. Do you recall?
16:35:25 gibi mriedem: I'm finished replying to your comments in the bandwidth series. Thank you very much for those reviews. I agreed with you in every relevant cases
16:35:43 gibi mriedem: and thanks again for proposing fups while I'm away
16:37:28 mriedem np, take care
16:37:47 mriedem i also appreciate your wording here, "I agreed with you in every relevant cases" :)
16:39:01 gibi :)
16:39:51 openstackgerrit Matthew Booth proposed openstack/nova master: Fix SSL infinite recursion running WSGI app under python 3.7 https://review.openstack.org/626952
16:40:32 lyarwood mdbooth: tbh it was a random punt at the underlying problem so I just used related, it should be close now tbh
16:41:05 mdbooth lyarwood: Yeah, that's what I thought too. I updated.
17:58:29 mriedem oh fun
17:58:29 mriedem stack@stein:~/nova$ curl http://162.253.55.42/compute/v2.1
17:58:30 mriedem Out of date osapi_compute service This service is older (v36) than the minimum (v37) version of the rest of the deployment. Unable to continue.
17:58:42 mriedem with all services in nova_cell0 and nova_cell1 being version 37
17:59:01 mriedem which sharts out 500s from the api when listing services
18:00:08 mriedem hell i'm not even listing services there, just getting the version doc
18:00:09 mriedem wtf
18:00:37 mgagne what's the purpose of reservations in Nova? can they all be deleted without impact? I have a situation where *some* reservations where not expired as expected and I'm now stuck with them.
18:00:47 mriedem mgagne: depends on the release
18:00:58 mriedem reservations aren't used starting in pike
18:01:02 mgagne mriedem: mitaka
18:01:04 mriedem they are related to quota
18:01:35 mgagne mriedem: with cellsv1 of course. they are currently located in the API cell database.
18:02:17 mriedem they are meant to be short-term
18:02:29 mgagne right.
18:02:34 mriedem they are a placeholder for quota until a quota 'claim' is either committed or rolled back
18:02:57 mriedem so if the server a given reservation is already gone or active, you don't need the reservation anymore

Earlier   Later