Earlier  
Later
New Search
Posted Nick Remark
#openstack-sdks - 2019-04-22
23:45:30 openstackgerrit Ghanshyam Mann proposed openstack/python-openstackclient master: Dropping the py35 testing https://review.opendev.org/654660
#openstack-sdks - 2019-04-23
03:21:03 openstackgerrit jacky06 proposed openstack/os-api-ref master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/654687
05:37:03 openstackgerrit Le Hou proposed openstack/api-sig master: Update to opendev https://review.opendev.org/654746
08:21:44 dtantsur gtema: morning! could you check https://review.opendev.org/#/c/650044/ if you have a minute?
08:30:32 gtema morning dtantsur, sure
08:31:19 gtema I am however not sure it can be merged, there are still some stuff broken after opendev migration
08:31:43 gtema I mean today
08:42:37 dtantsur I see, yeah. any CI patches I should pay attention to?
08:43:07 gtema no idea, most lay in the bifrost and zuul
08:43:08 gtema https://review.opendev.org/#/c/654238/
08:43:12 gtema https://review.opendev.org/#/c/654056/
08:43:21 gtema https://review.opendev.org/#/c/654230/
08:43:26 gtema https://review.opendev.org/#/c/654233/
09:52:17 openstackgerrit zhulingjie proposed openstack/os-api-ref master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/654928
09:56:05 openstackgerrit zhulingjie proposed openstack/os-api-ref master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/654928
11:37:14 openstackgerrit caoyuan proposed openstack/api-sig master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/654982
11:38:30 openstackgerrit caoyuan proposed openstack/keystoneauth master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/655016
11:39:44 openstackgerrit caoyuan proposed openstack/os-api-ref master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/655051
11:51:56 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Alter test_volume.py to ensure volume deletion https://review.opendev.org/652682
12:39:23 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Alter test_volume.py to ensure volume deletion https://review.opendev.org/652682
13:14:21 openstackgerrit caoyuan proposed openstack/keystoneauth master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/655016
13:52:16 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Alter test_volume.py to ensure volume deletion https://review.opendev.org/652682
17:39:23 openstackgerrit zhulingjie proposed openstack/os-api-ref master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/654928
17:40:02 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Alter test_volume.py to ensure volume deletion https://review.opendev.org/652682
17:59:19 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Alter test_volume.py to ensure volume deletion https://review.opendev.org/652682
18:01:02 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Alter test_volume.py to ensure volume deletion https://review.opendev.org/652682
18:52:45 openstackgerrit Daniel Speichert proposed openstack/openstacksdk master: Deduplicate next-page URL's query params https://review.opendev.org/643675
19:54:54 efried mordred: You at some point asked to be notified when the nova schedule was solidified so you could be there for [topics that interested you] - but I forgot what those topics would be.
19:55:35 efried mordred: I thought it was like SDK stuff. If that's the case, I *think* the only thing SDK-ish on the agenda is the thing you and I are already working on. Do you want to be around forr that?
19:57:13 mordred efried: well, if you're going to talk about it, then sure - but if you're not going to talk about it, then it's not urgent to be tehre
19:58:57 efried mordred: It'll be on the agenda. Not sure how much discussion will ensue. Having you there to say, "yes, this is a thing we should do," and to answer questions about the mechanics of nova eventually moving to using clouds.yaml would be useful, but not like drop-dead critical if you're swamped with other stuff.
19:59:44 efried that said, what's your schedule like? I can try to coordinate with where you've got an opening. How's your saturday?
20:00:53 mordred efried: see - now you're asking hard questions :)
20:01:13 mordred saturday woudl be less good - but let me good dig up what things are looking like
20:03:15 efried mordred: Can also do Thursday. Friday is fully of xproj stuff but there's a couple of small openings: half hour right before and right after lunch; and after 1615
20:09:41 mordred efried: let's shoot for sometime thursday if that's ok - and I think you're furthr along with scheduling than infra is - so if you tell me a slot I'll just let people know I need to be out of room for a little bit
22:10:38 efried work for you?
22:10:38 efried mordred: Thursday 1415-1430: Integrating openstacksdk and replacing use of python-*client
#openstack-sdks - 2019-04-24
00:21:36 mordred efried: ++
00:22:37 mordred efried: it is on my calendar - I will be there
13:42:15 openstackgerrit Merged openstack/python-openstackclient master: Fix docs bug link to go to storyboard rather than launchpad https://review.opendev.org/653162
14:48:12 openstackgerrit Daniel Speichert proposed openstack/openstacksdk master: Deduplicate next-page URL's query params https://review.opendev.org/643675
15:24:01 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Alter test_volume.py to ensure volume deletion https://review.opendev.org/652682
18:47:40 dtroyer Ponderment of the day: I need help in creating the argument of why one might want to install a CLI (yes, OSC in this case) in a container. Not for use by other processes in the container but to have a wrapper around docker exec to call it inside the container. I am sure there is a use case for that out there somewhere but I apparently lack the imagination to conjure it up…
20:03:17 mordred dtroyer: so that you can install the CLI and its dependencies easily without needing to install dependencies globally. kind of like how we frequently install such tools into virtualenvs
20:04:36 mordred dtroyer: especially for CLIs that aren't really doing much with the filesystem - obviously dealing with config files sucks a little bit more and would require a -v to the docker exec call - which is where a nice wrapper script around the docker exec would ALSO be nice, but the stories for installing such wrapper scripts are ... not awesome
20:05:36 mordred dtroyer: but "docker exec -it --rm -v$HOME/.config/openstack:/home/containeruser/.config/openstack openstack/osc server list"
20:06:13 mordred dtroyer: would save needing to install stuff globally - where installing the latest OSC and its depends might mess somebody up who is otherwise running older openstack and worried about libraries clashing
20:09:59 dtroyer mordred: ok, the image size bothers me a bit, I guess I am just not totally onboard with Containerize All The Things. The odd thing is that it is all installed from RPM anyway, there is no expectation of installing from pypi or anywhere else…
20:10:05 mordred dtroyer: also- I know for a ceph deployment running in a k8s, we've used the ceph-tools container for cli tools. basically, in one of the pods, we run a ceph-tools container, and then can exec in to it when we need to do ceph cli things
20:11:44 mordred dtroyer: oh - well, I mean- using the python:slim base image and the opendev/python-builder image we put together for making images from software managed by pbr - the zuul docker images are about 266M
20:11:54 mordred dtroyer: there's definitely ways to make efficient images
20:11:59 dtroyer mordred: so in that Ceph scenario, is there a chicken-n-egg problem potential there? I recall one of the earliest Nova installations outside NASA had their DNS and LDAP running in the cloud. Then they needed to cold start. Uh oh.
20:12:01 mordred and then ways to make terribly inefficient images
20:12:39 mordred dtroyer: well, in this case the ceph in question is running in a k8s that's running in an openstack that itself has a ceph/cinder install that we're using to get the block devices that the in-k8s is using as raw devices
20:12:48 dtroyer I wish I had kept up the work in the NEbula days of making OpenWRT's Python work well enough to run OSC. I still love the size of thise things…
20:12:53 mordred but that's beause the app layer needs a cephfs instance and it's the only way to get there
20:13:28 mordred it's more just an example of a use case even in a k8s world where having an image whose only pupose in life is to contain the CLI tools has been actually helpful
20:13:52 dtroyer ok, that is actually exactly what I needed to open up my imagination
20:13:54 dtroyer thanks
20:14:07 mordred dtroyer: we started off using the python:alpine base images for the zuul images, but the libc differences wound up being impactful enough that we gave up and switched to python:slim
20:14:12 mordred yay!
20:14:34 dtroyer mordred: I haven't done enough in that kind of multi-layered environment to have all the sharp edges in my head
20:15:39 dtroyer re alpine, libc that was one of the things I ran up against in OpenWRT.
20:15:56 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Refactor test_snapshot.py and test_transfer_request.py https://review.opendev.org/655554
20:33:26 openstackgerrit Eric Fried proposed openstack/openstacksdk master: WIP Make factory for a CloudRegion from CONF objects https://review.opendev.org/643601
20:33:58 efried mordred: FYI, updated requirements and lower constraints for ksa to newly-released-and-requirements-bumped 3.14.0 and removed the depends-on. No other changes ^
21:35:48 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Refactor test_snapshot.py and test_transfer_request.py https://review.opendev.org/655554
21:38:15 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Refactor test_snapshot.py and test_transfer_request.py https://review.opendev.org/655554
22:15:41 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Refactor test_snapshot.py and test_transfer_request.py https://review.opendev.org/655554
22:26:26 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Refactor test_snapshot.py and test_transfer_request.py https://review.opendev.org/655554
#openstack-sdks - 2019-04-25
02:26:37 openstackgerrit liuwei proposed openstack/openstackclient master: update git.openstack.org to opendev https://review.opendev.org/655598
13:23:14 openstackgerrit Artem Goncharov proposed openstack/openstacksdk master: Update compute.server resource https://review.opendev.org/650868
13:41:48 openstackgerrit liuwei proposed openstack/openstackclient master: update git.openstack.org to opendev https://review.opendev.org/655598
13:45:58 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/openstacksdk master: Add "name" filter in "list" call when retrieving a single register https://review.opendev.org/637238
14:07:08 openstackgerrit Merged openstack/openstackclient master: update git.openstack.org to opendev https://review.opendev.org/655598
14:19:10 openstackgerrit Merged openstack/openstacksdk master: Deduplicate next-page URL's query params https://review.opendev.org/643675
14:35:29 openstackgerrit Merged openstack/openstacksdk master: Support for the baremetal introspection service https://review.opendev.org/650044
14:51:59 openstackgerrit Eric Fried proposed openstack/openstacksdk master: Uncap jsonschema https://review.opendev.org/655714
14:51:59 openstackgerrit Eric Fried proposed openstack/openstacksdk master: WIP Make factory for a CloudRegion from CONF objects https://review.opendev.org/643601
14:53:22 efried mordred: fyi, fix jsonschema requirements job failure in the CloudRegion patch via the jsonschema patch, courtesy of tonyb / smcginnis
14:53:25 efried --^
16:00:02 edleafe The API-SIG Office Hour has now begun. Don't be shy!
16:00:41 edleafe Actually, it's probably going to be the API-SIG Office 50 Minutes, as I have to leave a bit early
16:01:32 dtantsur \o
16:06:32 dtantsur edleafe: shall we merge something while nobody is looking? :)
16:07:05 edleafe We could add something to the guidelines that makes fun of elmiko
16:13:25 dtantsur guidelines on making fun of elmiko?
16:14:21 edleafe Or something like "You should do things this way, unless you're a big dummy like elmiko"
16:15:55 dtantsur lol
16:27:16 openstackgerrit Stephen Finucane proposed openstack/os-api-ref master: Add support for Sphinx 2.0 https://review.opendev.org/655734
16:31:26 openstackgerrit Stephen Finucane proposed openstack/os-api-ref master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/654928
16:52:44 edleafe OK, gotta run. See all (most?) of you at the PTG next week!
#openstack-sdks - 2019-04-26
00:09:32 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: Refactor test_snapshot.py and test_transfer_request.py https://review.opendev.org/655554
00:27:34 openstackgerrit Merged openstack/openstacksdk master: Uncap jsonschema https://review.opendev.org/655714
02:00:20 openstackgerrit Merged openstack/openstacksdk master: Finish updating links to point to opendev https://review.opendev.org/654056
06:23:39 openstackgerrit Merged openstack/openstacksdk master: Remove now unused task_manager file https://review.opendev.org/653731
09:21:39 openstackgerrit Merged openstack/os-api-ref master: Add support for Sphinx 2.0 https://review.opendev.org/655734
09:21:40 openstackgerrit Merged openstack/os-api-ref master: Replace git.openstack.org URLs with opendev.org URLs https://review.opendev.org/654928
09:27:00 openstackgerrit Andreas Jaeger proposed openstack/os-api-ref master: Update Testing for Train https://review.opendev.org/655848

Earlier   Later