| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2019-07-23 | |||
| 14:22:46 | BlackDex | now it seems to just pick the first one in the line (seems to be sorted by network id) | |
| 15:16:56 | mordred | BlackDex: you can configure the network that such an IP should come from - one sec, lemme get a link | |
| 15:17:47 | mordred | BlackDex: https://docs.openstack.org/openstacksdk/latest/user/config/network-config.html see default_interface | |
| 17:36:03 | Shrews | mordred: got a sec for a rax related question? | |
| 17:39:54 | mordred | Shrews: sure! | |
| 17:39:59 | mordred | Shrews: am I going to like it? | |
| 17:40:43 | Shrews | mordred: well, gimme a moment... gotta organize a log paste for you | |
| 17:48:46 | Shrews | nm. maybe later when i unconfuse myself | |
| 18:04:33 | Shrews | mordred: ok, different question | |
| 18:05:33 | Shrews | mordred: keep seeing this error in nodepool logs: http://paste.openstack.org/show/754765/ | |
| 18:07:06 | Shrews | mordred: i totally don't get the organization of sdk these days, but _IMAGE_OBJECT_KEY is waaaaay down in the base proxy class (because not sure why) and referenced way up in the image mixin. I don't see Connection deriving from the proxy class??? | |
| 18:09:03 | Shrews | oh, wow. ok, there are all sorts of Proxy() classes | |
| 18:18:30 | mordred | Shrews: yah - the sdk layer has a Proxy() for each service | |
| 18:19:28 | mordred | Shrews: then on the Connection there shoudl be an instance of each one - so conn.image should be an instance of openstack.image.v2._proxy.Proxy | |
| 18:24:11 | Shrews | mordred: the Connection derives from ImageCloudMixin which, it looks like based on the (self reference to that attribute), assumes to be an instance of that Proxy, yet it is not. It's confusing to me | |
| 18:24:29 | Shrews | oops, extra ()'s there for some reason | |
| 18:25:11 | mordred | Shrews: it shouldn't - it should have a proxy, not is-a proxy - I think that's likely just a bug from pushing logic down into the proxy layer | |
| 18:26:11 | mordred | Shrews: the idea was to move as much logic from the shade layer into the proxy layer as made sense, so that all the basics would work no matter which layer you use - and so that the shade layer could then be business logic on top of a robust and working proxy layer | |
| 18:26:50 | mordred | Shrews: image was one of the first things to transition - so it looks like something got borked - and also that that something doesn't have good test coverage | |
| 18:32:13 | Shrews | well, i have a suspicion there might be other issues, but that seems to be the simplest at this point | |
| 18:32:52 | Shrews | we have a lot of Conflict exceptions on uploading that i don't ever recall seeing | |
| 18:33:32 | Shrews | across multiple providers | |
| 18:33:55 | Shrews | and some leaked images, so maybe some sort of relation there that I'm attempting to figure out | |
| 20:19:16 | openstackgerrit | Merged openstack/openstacksdk master: Update api-ref location https://review.opendev.org/672131 | |
| 20:30:57 | mnaser | http://paste.openstack.org/show/754774/ -- this seems to throw a traceback, am i doing something terribly wrong? the docs seem to say im not | |
| 20:31:01 | mnaser | cc mordred ^ | |
| 20:31:39 | mordred | ugh | |
| 20:31:40 | mordred | I mean | |
| 20:31:42 | mordred | uh | |
| 20:31:44 | mordred | one sec- lemme see | |
| 20:32:01 | mnaser | this si pretty much copy pasted from the docs | |
| 20:32:15 | mordred | I feel like there was a patch from you ages ago that fixed something in the docs related to this | |
| 20:32:26 | mnaser | combination of https://docs.openstack.org/openstacksdk/latest/user/config/using.html and https://docs.openstack.org/openstacksdk/latest/user/guides/connect_from_config.html | |
| 20:32:52 | mnaser | https://review.opendev.org/#/c/627781/ | |
| 20:32:54 | mnaser | oh you're right | |
| 20:33:03 | mnaser | apparently i've fixed it myself | |
| 20:33:17 | mnaser | 7 months ago | |
| 20:34:00 | mnaser | mordred: ive rechecked that, i dunno if its straight forward enough to single approve and ill recheck it around | |
| 20:34:26 | mordred | mnaser: well - I just verified that what you have there in the docs actually works | |
| 20:34:38 | mnaser | yeah same :p | |
| 20:34:46 | mordred | mnaser: I suppose we made things easier at some point and then didnt' update the docs. that's so unlike me :) | |
| 20:34:47 | mnaser | good thing 7 month ago me is a stand up guy | |
| 20:34:56 | mnaser | and fixed it | |
| 20:35:25 | mordred | Shrews: if you have a sec, https://review.opendev.org/#/c/627781 does seem to fix an actual thing I just verified | |
| 20:35:36 | mordred | mnaser: let's rebase that real quick - just so it's nice and clean ... | |
| 20:35:41 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: docs: Add simplified CLI parser docs https://review.opendev.org/627781 | |
| 20:35:49 | mordred | k. there it is | |
| 20:50:45 | mnaser | mordred: ok only because you asked -- http://paste.openstack.org/show/754780/ fails but "openstack stack create --wait -t cluster.yaml k8s" works | |
| 20:50:51 | mnaser | > | |
| 20:50:52 | mnaser | openstack.exceptions.BadRequestException: BadRequestException: 400 | |
| 20:51:05 | mnaser | before i dive server side am i doing something terribly stupid or it feels pretty obvious.. | |
| 20:51:44 | mordred | mnaser: poop | |
| 20:52:03 | mnaser | https://github.com/openstack/openstacksdk/blob/f03651d61e621c6d6d8238ba7f9e676eceb7c2a5/openstack/tests/functional/cloud/test_stack.py#L102-L105 | |
| 20:52:07 | mnaser | this seems to tell me im not doing anything back | |
| 20:52:32 | mordred | mnaser: I agree. I would expect that to work | |
| 20:54:01 | mordred | mnaser: it's dinner time for me - but I'll check post-dinner and see if you've uncovered anything, otherwise I'll give it a shot and see what I can find | |
| 20:54:18 | mnaser | mordred: ok cool no worries, ill let you know or push up a patch or whatever :) | |
| 23:38:56 | openstackgerrit | Merged openstack/openstacksdk master: docs: Add simplified CLI parser docs https://review.opendev.org/627781 | |
| #openstack-sdks - 2019-07-24 | |||
| 00:01:24 | openstackgerrit | Riccardo Pittau proposed openstack/openstacksdk master: Add node traits support to baremetal https://review.opendev.org/672233 | |
| 00:03:48 | openstackgerrit | Riccardo Pittau proposed openstack/openstacksdk master: Add node traits support to baremetal https://review.opendev.org/672233 | |
| 00:05:21 | openstackgerrit | Riccardo Pittau proposed openstack/openstacksdk master: Add node traits support to baremetal https://review.opendev.org/672233 | |
| 06:59:45 | BlackDex | mordred: Ah, i have to set that on the network. And can't configure this per instance it seems | |
| 07:29:28 | BlackDex | mordred: That doesn't seem to be a working solution | |
| 08:42:43 | openstackgerrit | Riccardo Pittau proposed openstack/openstacksdk master: Add node traits support to baremetal https://review.opendev.org/672233 | |
| 10:45:28 | openstackgerrit | Riccardo Pittau proposed openstack/openstacksdk master: Add node traits support to baremetal https://review.opendev.org/672233 | |
| 12:36:41 | openstackgerrit | Merged openstack/shade master: URL encode swift objects endpoints https://review.opendev.org/663922 | |
| 13:18:57 | openstackgerrit | Merged openstack/openstacksdk master: force specify project_id during create Sg_rule may cause issue https://review.opendev.org/667261 | |
| 13:18:58 | openstackgerrit | Merged openstack/openstacksdk master: URL encode swift objects endpoints https://review.opendev.org/663925 | |
| 13:35:33 | Shrews | mordred: gtema did the move of the image code to the proxy correct? I was hoping to point him to that issue we discussed yesterday. Doesn't seem to be around though | |
| 13:57:51 | openstackgerrit | Merged openstack/openstacksdk master: Use Resource layer for compute KeyPairs https://review.opendev.org/662729 | |
| 13:57:52 | openstackgerrit | Merged openstack/openstacksdk master: Specify store when importing an image https://review.opendev.org/670309 | |
| 15:47:18 | mordred | Shrews: yes it was gtema | |
| 15:52:19 | openstackgerrit | Riccardo Pittau proposed openstack/openstacksdk master: Add node traits support to baremetal https://review.opendev.org/672233 | |
| 16:03:40 | openstackgerrit | Matt Riedemann proposed openstack/python-openstackclient master: Mention compute service set --up|--down requires 2.11 or greater https://review.opendev.org/672554 | |
| 17:06:48 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: Rework statistics reporting https://review.opendev.org/659841 | |
| 17:08:01 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: Rework statistics reporting https://review.opendev.org/659841 | |
| 17:20:31 | Shrews | gtema: have you seen the scrollback on the _IMAGE_OBJECT_KEY errors? | |
| 17:21:02 | gtema | cool. When was it and what was it about? | |
| 17:21:23 | Shrews | gtema: http://eavesdrop.openstack.org/irclogs/%23openstack-sdks/%23openstack-sdks.2019-07-23.log.html#t2019-07-23T18:05:33 | |
| 17:22:30 | Shrews | gtema: tl;dr is the paste of that error http://paste.openstack.org/show/754765/ | |
| 17:22:59 | gtema | ok, reading | |
| 17:23:08 | Shrews | seems to be occurring since the image code reorg | |
| 17:24:47 | gtema | yeah, seems to be | |
| 17:25:00 | gtema | lemme fast finish some other patch and I will look | |
| 17:25:08 | Shrews | gtema: thx! :) | |
| 17:29:33 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: Rework statistics reporting https://review.opendev.org/659841 | |
| 18:14:30 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: Fix image deletion with tasks_api enabled https://review.opendev.org/672573 | |
| 18:14:44 | gtema | Shrews: ^^^ | |
| 18:44:01 | openstackgerrit | Matt Riedemann proposed openstack/python-openstackclient master: WIP: Fix compute service set handling for 2.53+ https://review.opendev.org/672577 | |
| 19:12:47 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: Rework statistics reporting https://review.opendev.org/659841 | |
| 19:15:53 | Shrews | mordred: gtema did a thing https://review.opendev.org/672573 | |
| 19:16:20 | Shrews | thx gtema | |
| 19:17:01 | gtema | welcome. Hope it helps. If I can help debugging those image issues for nodepool - let me know, will try as much as time allows | |
| 19:24:21 | mordred | gtema: thanks! +A - looks solid | |
| 19:24:49 | mordred | gtema: I landed a few more of the reorg patches this morning - sorry that's taking so long | |
| 19:25:25 | gtema | no problem. I'm myself terribly busy last month with my "API-Monitoring" project :-) | |
| 19:34:32 | mordred | :) | |
| 19:35:55 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: Volume.Backup restore fixes https://review.opendev.org/665110 | |
| 21:17:06 | openstackgerrit | Merged openstack/openstacksdk master: Fix image deletion with tasks_api enabled https://review.opendev.org/672573 | |
| 22:46:57 | openstackgerrit | Riccardo Pittau proposed openstack/openstacksdk master: Add node traits support to baremetal https://review.opendev.org/672233 | |