Earlier  
Later
New Search
Posted Nick Remark
#openstack-sdks - 2018-06-28
20:56:22 rm_work you have a CR?
20:56:24 rm_work or, SHA
20:56:29 rm_work i can apply
20:56:36 rm_work or i guess i could just checkout master
20:56:46 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Only send force parameter to live migration if supported https://review.openstack.org/578960
20:56:57 mordred mriedem: ^^ updates. thanks for the feedback btw - super helpful!
20:56:59 mordred rm_work: looking
20:57:00 rm_work yeah sec i'll just switch to master
20:57:03 rm_work don't worry about it
20:57:04 mordred ok. cool
20:57:31 rm_work hmmmmmmm
20:57:48 mordred it's also possible that we looked at this, I said "oh, I see the issue and will write a patch" and then didn't
20:57:56 rm_work yeah i think it's the latter
20:58:03 rm_work master seems to be the same
20:58:26 mordred yes!
20:58:30 mordred I remember the thing now
20:58:54 rm_work yeah everyone using this cloud is stuck on 0.9.x which is just sad T_T
20:58:59 mordred kmalloc: ^^ on older clouds, neutron's version discovery document is auth protected (because of course it is)
20:59:06 rm_work lolol yes
20:59:26 mordred kmalloc: I kind of think we should just update ksa to send a token if it has one when doing discovery
20:59:42 mordred kmalloc: (but not to get a token if it doesn't already have one perhaps)
21:00:05 mordred kmalloc: or else I can probably work around it in SDK - but it's definitely a weird gotcha for folks
21:00:08 rm_work wouldn't that just make the error inconsistent then? :(
21:00:39 kmalloc Hmm
21:00:44 mordred rm_work: hrm. good point
21:01:05 kmalloc Yeah it is weird
21:01:53 rm_work can you assume that if it is auth protected, that it's one of the old versions pre-discovery? and that tells you what the endpoint is? :P or is it still variable?
21:01:56 mordred kmalloc: the sdk answer will basically be "if network_endpoint_override is None and not catalog_url.endswith('/v2.0'): network_endpoint_override = catalog_url + '/v2.0'
21:02:20 mordred rm_work: well - *generaly* no - but for neutron it turns out there is one and only one version if it's old
21:02:26 rm_work yes
21:02:30 rm_work that is what i was thinking
21:08:45 mordred AHHHHH
21:12:31 rm_work :P
21:14:07 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Hardcode v2.0 onto end of neutron endpoints https://review.openstack.org/579015
21:14:17 mordred rm_work: ^^ untested yet - but maybe that'll fix it for you
21:14:31 rm_work will try it
21:15:24 rm_work mordred: A++ :)
21:15:31 rm_work thanks
21:15:43 mordred rm_work: it worked?
21:15:45 rm_work yep
21:15:48 mordred yay!
21:15:53 mordred I will update that patch with some tests
21:16:09 rm_work review: A++, would bother mordred about patches again
21:16:14 mordred :)
21:16:28 rm_work thanks muchly
21:16:35 mordred kmalloc: ^^ in case you wanna look at the evil hacky thing
21:16:37 rm_work starting to have stuff we'd like to have past 0.9.x :P
21:27:16 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Hardcode v2.0 onto end of neutron endpoints https://review.openstack.org/579015
21:33:45 kmalloc K
21:57:32 openstackgerrit Sorin Sbarnea proposed openstack/cliff master: Assure executable name is kept when app is called as module https://review.openstack.org/579025
22:11:51 openstackgerrit Lance Bragstad proposed openstack/python-openstackclient master: Implement support for registered limits https://review.openstack.org/574475
22:11:51 openstackgerrit Lance Bragstad proposed openstack/python-openstackclient master: Implement support for project limits https://review.openstack.org/575211
23:08:59 kmalloc mordred: looking at evil hacky-thing now that i am not on mobile
23:09:41 kmalloc mordred: gross, but hey, if it fixes it
#openstack-sdks - 2018-06-29
01:48:13 openstackgerrit Chen proposed openstack/python-openstackclient master: Fix a typo https://review.openstack.org/572771
02:47:42 openstackgerrit Merged openstack/python-openstackclient master: Add ability to filter image list by tag https://review.openstack.org/575256
03:25:26 openstackgerrit tianhui proposed openstack/python-openstackclient master: Compute: Add tags support for server https://review.openstack.org/569386
06:22:47 openstackgerrit Jens Harbott (frickler) proposed openstack/python-openstackclient master: Add DNS support to floating IP commands https://review.openstack.org/558824
08:52:27 openstackgerrit wu.chunyang proposed openstack/openstacksdk master: Add release note link in README https://review.openstack.org/578672
08:59:08 openstackgerrit wu.chunyang proposed openstack/openstacksdk master: Add release note link in README https://review.openstack.org/578672
09:49:14 openstackgerrit tianhui proposed openstack/python-openstackclient master: Compute: Add tags support for server https://review.openstack.org/569386
11:43:07 dtantsur mordred: morning! dumped more thoughts in https://etherpad.openstack.org/p/microversions-better-than-beer
11:59:47 openstackgerrit tianhui proposed openstack/python-openstackclient master: Compute: Add tags support for server https://review.openstack.org/569386
12:01:22 mordred dtantsur: morning! awesome
12:09:57 openstackgerrit tianhui proposed openstack/python-openstackclient master: Compute: Add tags support for server https://review.openstack.org/569386
12:10:36 openstackgerrit Chen proposed openstack/python-openstackclient master: Compute: Add description support for server https://review.openstack.org/568549
12:31:12 dtroyer dhellmann: if you have a minute and need a break today to think about something a bit different than py3 and TC-stuff, I'd love to hear your thoughts on https://etherpad.openstack.org/p/osc-included_image_signing. tl;dr: signing code that wants to be used in both OSC and server-side (nova so far), and an appropriate place for it to live.
12:35:11 mordred dtroyer: seems like functionality we'd want in sdk too
12:36:05 dtroyer mordred: I didn't think about that long, I suppose I had oslo on my brain a bit much
12:36:29 dtroyer is nova ready to use sdk? have they already added it as a dep?
12:36:51 mordred they haven't - I wasn't going to push on it until we cut the 1.0
12:37:27 mordred but it shouldn't be an issue for it to. that said - looking at cursive real quick, I don't think sdk would want to pick it up as a depend
12:37:27 dtroyer agreed there, I wonder if this is enough of a reason to start
12:37:45 mordred dtroyer: I can make some initial patches to check it out
12:37:56 dtroyer I don't think you need to, the signing bits don't use the rest of cursive as far as I saw last night
12:38:04 mordred cool
12:38:05 dtroyer it was just a home
12:38:30 dtroyer pile on to that ML thread and see what Josephine thinks
12:38:34 mordred well - image signing code _definitely_ would be welcome in sdk
12:38:36 mordred will do
12:43:35 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Only send force parameter to live migration if supported https://review.openstack.org/578960
12:43:51 mordred dtantsur: speaking of microversions ^^
12:47:58 openstackgerrit Monty Taylor proposed openstack/osc-lib master: Use openstack.config for config and Session objects https://review.openstack.org/579139
12:48:13 mordred dtroyer: also - ^^ that's still WIP - but I figured I should maybe not just keep it on my laptop
12:59:58 mordred Shrews: if you get a moment this morning, would you take a peek at https://review.openstack.org/#/c/579015/ and https://review.openstack.org/#/c/578960/
13:19:33 Luzi dtroyer, mordred: I think using openstacksdk is a good idea, i already wrote an email. Thank you both :)
13:20:36 dtroyer Luzi: good to hear… and glad you found us here :)
13:26:19 mordred Luzi: \o/ yay
13:30:52 mordred Luzi: I'll try to write down some more notes - but fwiw in the image upload code in the shade portion of sdk (in openstack/cloud/openstackcloud.py) we currently have code that calculates md5 and sha256 hashes for images before uploading them - and it does a similar thing to the main loop in your cursive patch (iterate over file, call hasher.update(chunk))
13:31:02 mordred so I think there is a really great place to hook the code in
13:31:19 mordred in terms of using it in the image upload code there
13:32:46 mordred Shrews: thanks!
13:34:35 Luzi mordred: thanks, I will have a look at this
13:50:26 Shrews mordred: np
13:57:46 openstackgerrit Merged openstack/python-openstackclient master: Display private flavors in server list https://review.openstack.org/540633
14:08:20 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Add connection backreference to proxy instances https://review.openstack.org/579167
14:08:20 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: WIP Use shade create_image method in upload_image https://review.openstack.org/579168
14:09:13 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Only send force parameter to live migration if supported https://review.openstack.org/578960
14:24:26 frickler Luzi: when you talk about nova support, is this just the "openstack server image create" command or something inside nova services? the former would only seem to affect OSC and not nova as a project, I guess
14:27:06 Luzi frickler: it should be just that command, but the compute host has to do the signing

Earlier   Later