| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2018-07-23 | |||
| 14:38:42 | dtroyer | mordred: nice! | |
| 14:42:38 | openstackgerrit | Paolo Smiraglia proposed openstack/osc-lib master: Change issubclass() to isinstance() https://review.openstack.org/584920 | |
| 14:45:16 | pasm | dtroyer: more easy than expected... thanks again! | |
| 14:45:35 | dtroyer | pasm: \o/ thank you! | |
| 14:46:35 | pasm | dtroyer: and now? someone will took in charge my review request? | |
| 14:47:57 | dtroyer | pasm: yes. we are in a freeze on osc-lib so it may be a bit before it merges, unless we decide that a py37 issue is critical :) | |
| 14:49:05 | pasm | dtroyer: ok, I asked just for personal knowledge | |
| 14:50:04 | mordred | lgtm | |
| 14:50:46 | pasm | dtroyer: back to work now... ciao! | |
| 15:15:57 | mordred | dtroyer: I'm looking at the volume snapshot ting ... | |
| 15:17:18 | mordred | ugh. it seems like size is a required value now, even though for "create volume from snapshot" the volume size will be the size of the snapshot | |
| 15:20:56 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Pass volume snapshot size to volume create https://review.openstack.org/584932 | |
| 15:21:06 | mordred | dtroyer: I *think* that will take care of it | |
| 15:33:05 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Pass volume snapshot size to volume create https://review.openstack.org/584932 | |
| 15:35:42 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Pass volume snapshot size to volume create https://review.openstack.org/584932 | |
| 15:36:18 | mordred | dtroyer: also - spoke with smcginnis and jungleboyj in #openstack-cinder - it's a server-side bug they're going to fix | |
| 15:36:32 | mordred | dtroyer: but the osc patch should be harmless, so we might want to land it to unbreak the gate anway | |
| 15:40:13 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Pass volume snapshot size to volume create https://review.openstack.org/584932 | |
| 15:42:50 | mordred | https://review.openstack.org/#/c/584944/ <-- cinder fix | |
| 15:44:01 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: DNM - Testing cinder fix https://review.openstack.org/584948 | |
| 15:49:41 | dtroyer | mordred: ++ on unbreaking. since it doesn't require another API call, its fine to leave in | |
| 15:55:09 | mordred | coolio | |
| 15:56:16 | mordred | dtroyer: for that versions command - I also think being able to see it by single service would be a useful filter - but I need to add a new call to keystoneauth for that (I completely didn't think about that case when adding the original call) - so maybe that'll be an improvement for stein | |
| 15:56:58 | mordred | (obviously someone can filter - but the ksa command sure will still run discovery on every endpoint) | |
| 15:57:37 | dtroyer | Adding a client-side filter would be fine until we get the query argument to the call | |
| 15:58:20 | dtroyer | until ksa is updated it'll take the same amount of time anyway | |
| 15:58:32 | mordred | dtroyer: nod. good call. so you thnk it should just be a filter to that same command? so "openstack versions list --service-type=compute" ? | |
| 15:58:50 | dtroyer | yeah | |
| 15:59:07 | mordred | woot. I'll get that updated during this next phone call while I'm ignoring people | |
| 15:59:21 | mordred | kk | |
| 16:01:23 | dtroyer | mordred: confirm —service, we use it elsewhere for that, —service-type is used in network for a neutron thing | |
| 16:27:15 | openstackgerrit | Monty Taylor proposed openstack/service-types-authority master: Add container-infra as an alias for magnum https://review.openstack.org/584963 | |
| 16:27:34 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Add command to show all service versions https://review.openstack.org/584905 | |
| 16:28:05 | mordred | dtroyer: okie. there's filter by service - as well as by status. it'll let you specify service by any known alias | |
| 16:33:12 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Add command to show all service versions https://review.openstack.org/584905 | |
| 17:06:34 | openstackgerrit | Toure Dunnon proposed openstack-infra/shade master: python-shade expose MTU setting. https://review.openstack.org/578861 | |
| 17:27:33 | lbragstad | curious if i can get some reviews on https://review.openstack.org/#/q/status:open+project:openstack/python-openstackclient+branch:master+topic:bp/unified-limits before client freeze on Thursday? | |
| 17:33:05 | dtroyer | lbragstad: yup, those are on my short list | |
| 17:35:07 | mordred | dtroyer, adriant, rods, slaweq, dtantsur|afk, Qiming: just as a heads-up - Shrews just restarted infra's nodepool ... so as of now nodepool is running on openstacksdk instead of shade in production | |
| 17:35:31 | dtroyer | no pressure there! :) | |
| 17:35:34 | mordred | it's _mostly_ identical code - so that's mostly a "it looks like we didn't break anything in the merge" | |
| 17:35:37 | Shrews | well, 1 node of nodepool | |
| 17:35:37 | mordred | dtroyer: :) | |
| 17:35:53 | mordred | Shrews: ssh. the other thing sounds more impressive ;) | |
| 17:37:35 | mordred | dtroyer: bwahahaha. my DNM patch to test the cinder fix didn't test anything at all because it didn't touch any code files :) | |
| 17:38:59 | dtroyer | lbragstad: did you see my comments on https://review.openstack.org/#/c/574475/ last week? | |
| 17:55:36 | openstackgerrit | Merged openstack/service-types-authority master: Add container-infra as an alias for magnum https://review.openstack.org/584963 | |
| 18:08:53 | PagliaccisCloud | would someone mind giving me a hand with this bugfix? zuul keeps failing at creating a volume snap but I didn't changed anything having to do with cinder :/ https://review.openstack.org/#/c/584484/ | |
| 18:14:43 | mordred | PagliaccisCloud: there is a recent cinder bug that is causing it. a patch is up to fix it | |
| 18:15:09 | PagliaccisCloud | ah ok, thanks mordred. i'll wait it out then. | |
| 18:15:12 | mordred | PagliaccisCloud: https://review.openstack.org/#/c/584944/ | |
| 18:15:17 | mordred | is the patch in question | |
| 18:32:55 | lbragstad | dtroyer: oh - i didn't (my dashboard query is sensitive to votes < 0) | |
| 18:33:10 | lbragstad | dtroyer: thanks for the feedback, i'll get those addressed or responded to today | |
| 18:34:34 | dtroyer | lbragstad:, thanks, IIRC the idea applies to both review | |
| 18:36:01 | slaweq | mordred: nice :) | |
| 18:39:09 | openstackgerrit | Dean Troyer proposed openstack/python-openstackclient master: Pass volume snapshot size to volume create https://review.openstack.org/584932 | |
| 18:41:15 | dtroyer | mordred: minor fix ^^^^ | |
| 18:43:21 | mordred | dtroyer: ++ thanks! | |
| 19:16:58 | openstackgerrit | Dean Troyer proposed openstack/python-openstackclient master: Fix error with image show when image name is None https://review.openstack.org/529464 | |
| 19:17:59 | openstackgerrit | Dean Troyer proposed openstack/python-openstackclient master: Fix error with image show when image name is None https://review.openstack.org/529464 | |
| 19:29:17 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Add command to show all service versions https://review.openstack.org/584905 | |
| 20:12:15 | lbragstad | dtroyer: i'm getting an odd error addressing one of your comments | |
| 20:12:45 | lbragstad | it's related to making --default-limit a required argument and it seems to stumble with integer types | |
| 20:13:54 | lbragstad | http://paste.openstack.org/show/726476/ is the implementation and http://paste.openstack.org/show/726475/ is a trace of the failure | |
| 20:14:01 | lbragstad | does the failure jump out to you at all? | |
| 20:15:01 | lbragstad | http://paste.openstack.org/show/726477/ is what i have locally for the test | |
| 20:47:23 | openstackgerrit | Monty Taylor proposed openstack/keystoneauth master: Add ability to filter version data by service-type https://review.openstack.org/585029 | |
| 20:51:47 | dtroyer | lbragstad: in tests when setting arglist we make them all strings and let argparse do the conversion. | |
| 20:52:07 | lbragstad | oh | |
| 20:52:22 | mordred | dtroyer: ^^ that ksa patch above is the one we'll need to make that osc service-type filter be more efficient | |
| 20:52:23 | dtroyer | verifylist has the correct types | |
| 20:52:55 | dtroyer | lbragstad: its odd and I've forgotten exactly why but its that way everywhere I look | |
| 20:53:32 | lbragstad | huh - interesting... i figured the tests would expect the same type passed in as what's declared in the option definition | |
| 20:54:53 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Make use of keystoneauth service-type filtering for versions https://review.openstack.org/585030 | |
| 20:55:07 | lbragstad | dtroyer: that seems to have fixed it, thanks! | |
| 20:55:52 | dtroyer | lbragstad: I think it is because it passes through cliff and argparse and they expect to operate on strings | |
| 20:56:00 | dtroyer | as input | |
| 20:56:08 | lbragstad | that makes sense | |
| 21:05:25 | mordred | dtroyer: https://review.openstack.org/#/c/584932/ is green - you wanna click the +A? (I'm not sure my +2 counts since you just uploaded a fix to the original patch) | |
| 21:08:46 | dtroyer | mordred: got it. Thanks for beating me to that :) | |
| 21:09:27 | mordred | \o/ | |
| 21:09:48 | mordred | dtroyer: at least it wasn't another race-condition bug | |
| 21:14:39 | dtroyer | no kidding | |
| 21:18:58 | openstackgerrit | Lance Bragstad proposed openstack/python-openstackclient master: Implement support for registered limits https://review.openstack.org/574475 | |
| 21:20:16 | lbragstad | dtroyer: for https://review.openstack.org/#/c/575211/5/doc/source/cli/command-objects/limit.rst@17 would you like project to be the non-named argument? | |
| 21:21:09 | dtroyer | the rule is the positional arg should be the name or ID of the resource named in the command… that isn't always possible though | |
| 21:21:13 | lbragstad | so - openstack limit create --service compute --resource-name cores --resource-limit 20 project_foo | |
| 21:22:38 | dtroyer | I'm thinking it might be resource-name | |
| 21:22:57 | lbragstad | ok - so same pattern as registered limits then | |
| 21:23:15 | dtroyer | yes… that's what you use in the other commands to identify the particular limit | |
| 21:23:27 | dtroyer | also, can the name be used in the other commands or must it be ID only? | |
| 21:23:57 | lbragstad | well - limits are a little strange in that they have a rather complex uniqueness constraint | |
| 21:25:26 | lbragstad | each limit must be unique wrt service, default_limit (in the case of registered limits), resource name, and optionally region | |
| 21:25:43 | dtroyer | I wondered… I wouldn't mind doing a bit of work to make UX better, but that can come later, just curious | |
| 21:26:05 | lbragstad | so asking keystone for a list of all limits with resource names of 'cores' is more of a filter | |
| 21:26:24 | lbragstad | yeah - i hear you | |
| 21:26:44 | lbragstad | is that something we should track in a bug/RFE? | |
| 21:26:54 | lbragstad | if it isn't a blocker for these patches? | |
| 21:27:42 | dtroyer | it is certainly not a blocker, unless it was just a find_resource() call away :) sure, having a tracking story would be helpful | |