| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2018-07-17 | |||
| 13:57:41 | edleafe | cdent: Still waiting on travel approval for PTG <sigh> | |
| 13:58:38 | cdent | <sigh> | |
| 13:59:02 | mordred | cdent: I'll be at the PTG - if nobody else is, we can just sigh at each other | |
| 13:59:12 | cdent | ✔ | |
| 14:05:42 | dtantsur | folks, is the openstacksdk CI broken? http://logs.openstack.org/95/570895/12/check/nodepool-functional-py35-src/f6058c6/job-output.txt.gz#_2018-07-17_12_58_08_813473 | |
| 14:07:02 | Shrews | dtantsur: dib is | |
| 14:07:13 | Shrews | dtantsur: https://review.openstack.org/583068 | |
| 14:12:23 | dtantsur | is ^^^ the fix? | |
| 14:13:03 | Shrews | dtantsur: it should be since astroid is pulled in by pylint | |
| 14:13:32 | mordred | Shrews: do I want to know how dib's test-requirements somehow break that job? | |
| 14:14:38 | Shrews | mordred: i think pylint didn't cap its requirements (at least for that library) until a later version. dib capped the pylint version so never got the fix | |
| 14:14:54 | Shrews | (from my quick inspection of the situation) | |
| 14:15:24 | Shrews | so the latest astroid release broke things | |
| 14:15:57 | Shrews | mordred: i *still* don't understand why python2 is being used there, but I got fed up investigating | |
| 14:16:19 | dtantsur | magic and ponies | |
| 14:17:15 | mordred | Shrews: nod | |
| 14:17:36 | mordred | Shrews: oh - speaking of giant scary patches ... | |
| 14:18:03 | mordred | Shrews: once the nodepool job is fixed - https://review.openstack.org/#/c/577647/ ? | |
| 14:18:35 | mordred | (or, I guess the first question should be - does my response to your question on the previous patch make sense?) | |
| 14:19:06 | dtantsur | +89, -3051 | |
| 14:19:51 | mordred | also - WOW https://review.openstack.org/#/c/582460/ | |
| 14:20:14 | mordred | "When a floating IP is associated to the network port of a server with a fixed IP that doesn't belong to the server," ... *headdesk* | |
| 14:21:03 | dtantsur | .... because we can! | |
| 14:21:36 | Shrews | lovely | |
| 14:22:21 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Plumb additional_user_agent and collect_timings support throughout https://review.openstack.org/569666 | |
| 15:59:54 | openstackgerrit | wangxiyuan proposed openstack/keystoneauth master: Add netloc and version check for version discovery https://review.openstack.org/583215 | |
| 17:09:45 | openstackgerrit | wangxiyuan proposed openstack/keystoneauth master: Add netloc and version check for version discovery https://review.openstack.org/583215 | |
| 17:14:20 | openstackgerrit | Merged openstack-infra/shade master: Make OpenStackCloud a subclass of Connection https://review.openstack.org/577647 | |
| #openstack-sdks - 2018-07-18 | |||
| 01:56:00 | openstackgerrit | wangxiyuan proposed openstack/keystoneauth master: Add netloc and version check for version discovery https://review.openstack.org/583215 | |
| 02:54:14 | openstackgerrit | ayoung proposed openstack/keystoneauth master: Add netloc and version check for version discovery https://review.openstack.org/583215 | |
| 08:09:03 | openstackgerrit | Josephine Seifert proposed openstack/openstacksdk master: Implement signature generation functionality https://review.openstack.org/580080 | |
| 08:18:19 | openstackgerrit | Dmitry Tantsur proposed openstack/openstacksdk master: [WIP] Support for microversions in base Resource and baremetal resources https://review.openstack.org/583491 | |
| 08:29:02 | openstackgerrit | Josephine Seifert proposed openstack/openstacksdk master: Implement signature generation functionality https://review.openstack.org/580080 | |
| 08:51:27 | openstackgerrit | Duc Nguyen proposed openstack/python-openstackclient master: Replace port 35357 with 5000 for "auth_url" https://review.openstack.org/583514 | |
| 09:25:58 | openstackgerrit | Dmitry Tantsur proposed openstack/openstacksdk master: [WIP] Support for microversions in base Resource https://review.openstack.org/583491 | |
| 09:26:54 | openstackgerrit | Dmitry Tantsur proposed openstack/openstacksdk master: [WIP] Support for microversions in baremetal resources https://review.openstack.org/583532 | |
| 09:44:55 | openstackgerrit | Dmitry Tantsur proposed openstack/openstacksdk master: [WIP] Support for microversions in base Resource https://review.openstack.org/583491 | |
| 09:45:15 | openstackgerrit | Dmitry Tantsur proposed openstack/openstacksdk master: [WIP] Support for microversions in baremetal resources https://review.openstack.org/583532 | |
| 09:49:46 | openstackgerrit | Dmitry Tantsur proposed openstack/openstacksdk master: [WIP] Support for microversions in baremetal resources https://review.openstack.org/583532 | |
| 11:05:31 | openstackgerrit | Josephine Seifert proposed openstack/openstacksdk master: Implement signature generation functionality https://review.openstack.org/580080 | |
| 13:18:40 | openstackgerrit | Josephine Seifert proposed openstack/openstacksdk master: Implement signature generation functionality https://review.openstack.org/580080 | |
| 14:28:12 | daemontool | hi, anyone around? | |
| 14:29:51 | daemontool | I'm looking to fetch os-hypervisor data from compute API re: https://developer.openstack.org/api-ref/compute/?#list-hypervisors with sdk (all of thhem not only the list with details) | |
| 14:31:28 | daemontool | my understanding is that the interface is still limited from https://github.com/openstack/openstacksdk/blob/master/openstack/compute/v2/hypervisor.py | |
| 14:32:55 | daemontool | is my understanding correct? (i.e. I'd like to retrieve the stats endpoint /os-hypervisors/statistics) | |
| 14:37:12 | daemontool | I see the list_hypervisors method here: https://github.com/openstack/openstacksdk/blob/master/openstack/cloud/openstackcloud.py#L11112 | |
| 15:49:33 | lbragstad | wondering if it would be possible to get a couple reviews here https://review.openstack.org/#/q/topic:bp/unified-limits+status:open+project:openstack/python-openstackclient | |
| 16:36:31 | openstackgerrit | Dmitry Tantsur proposed openstack/openstacksdk master: Support for microversions in base Resource https://review.openstack.org/583491 | |
| 16:36:51 | dtantsur | mordred: would love feedback on the direction ^^^ | |
| 16:39:30 | tellesnobrega | hi, is osc_lib being release tomorrow? | |
| 16:41:06 | dtroyer | tellesnobrega: osc-lib was released late last week, I wasn't planning on another one | |
| 16:41:23 | tellesnobrega | last week is good enough for us | |
| 16:41:24 | tellesnobrega | thanks | |
| 16:41:37 | dtroyer | cool | |
| 17:32:09 | tosky | question about OSC: boolean CLI parameters seems to always follow the argparse convention (a bool --do-something flips the default value) | |
| 17:32:34 | tosky | would the usage of --enable-this=True/False be allowed, or should it be avoided for consistency? | |
| 17:32:58 | tosky | pro: it may help distinguish between an explicit choice and the usage of the internal default | |
| 17:33:46 | tosky | (context: discussion in https://review.openstack.org/#/c/580727/ ) | |
| 17:34:58 | Shrews | that seems like an awkward usage, tbh. one that i personally cannot recall ever seeing in any cli | |
| 18:23:15 | openstackgerrit | Hongbin Lu proposed openstack/python-openstackclient master: [DNM] Test neutron change https://review.openstack.org/583690 | |
| 20:54:22 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Add /v3 to the auth_url for vexxhost https://review.openstack.org/583737 | |
| 21:14:32 | mordred | Shrews: you seem to be back ... have a sec for that ^^ ? | |
| 21:14:45 | mordred | dtantsur|afk: I started looking at it but haven't properly read it yet | |
| 21:15:06 | Shrews | mordred: sure | |
| 21:16:40 | Shrews | mordred: done | |
| 21:18:48 | mordred | Shrews: thanks | |
| 21:18:55 | mordred | Shrews: I think we need to split that into a library | |
| 21:19:09 | mordred | Shrews: like service-types-authority/os-service-types | |
| 21:19:24 | mordred | Shrews: because it kinda sucks to need to release a whole new sdk just to fix a vendor profile | |
| 22:01:43 | corvus | con = sw.create_container(name='corvustest', read_ACL=".r:*") | |
| 22:01:53 | corvus | when i do that, the read acl isn't set | |
| 22:02:02 | corvus | i have to do a set_container_metadata call | |
| 22:02:26 | corvus | (sw = conn.object_store) | |
| 22:26:06 | mordred | corvus: so - I'll have to dig in on that one a bit - another option is conn.create_container(name='corvustest', public=True) | |
| 22:26:30 | corvus | oh neat, i didn't see that | |
| 22:27:21 | corvus | mordred: also, the next thing i was about to dig into (once it's unbroken) is how do i set X-Container-Meta-Web-Index ? | |
| 22:27:37 | mordred | corvus: it's the shade layer rather than the sdk object layer - also, fwiw, conn.create_object will automatically use large object segments for you if any of your files are over the object size limit for the cloud | |
| 22:28:59 | corvus | mordred: oh, i feel like i missed an important fork in the road :) | |
| 22:29:13 | mordred | corvus: conn.update_container(headers={'X-Container-Meta-Web-Index': 'blah'}) should do the trick | |
| 22:29:14 | corvus | mordred: i went to https://docs.openstack.org/openstacksdk/latest/user/index.html#api-documentation which led me to https://docs.openstack.org/openstacksdk/latest/user/proxies/object_store.html and that's why i thought i should be using conn.object_store | |
| 22:29:26 | mordred | corvus: ah - that's a great feedback note | |
| 22:30:24 | mordred | corvus: I'll work on some better entrypoint roadmapping ... | |
| 22:30:33 | mordred | corvus: https://docs.openstack.org/openstacksdk/latest/user/connection.html might be interesting to look at | |
| 22:30:36 | corvus | mordred: i see https://docs.openstack.org/openstacksdk/latest/user/connection.html now | |
| 22:30:37 | corvus | ya | |
| 22:30:57 | mordred | corvus: although - I'm now going to apologize for lacking some docstrings | |
| 22:31:05 | mordred | corvus: you're making several todo list items for me :) | |
| 22:31:14 | corvus | mordred: i think i would have needed that to be the first thing, and probably the text "The Connection class is the primary interface to the Python SDK" in the introduction page. that would have gotten me to the right place. | |
| 22:31:47 | mordred | corvus: yeah ... there's a piece of narrative that needs to be written | |
| 22:32:20 | mordred | corvus: that describes the different interfaces in some short but descriptive terms and leads you to the one that you want | |
| 22:33:34 | corvus | mordred: when you said you were going to apologize for lacking some docstrings, was the docstring for "create_container" one of them? :) | |
| 22:34:34 | mordred | corvus: yes. also for set_container_access and get_container and update_container and get_object_metadata | |
| 22:45:53 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Add missing swift docstrings https://review.openstack.org/583790 | |
| 22:45:57 | mordred | corvus: ^^ | |
| 22:46:00 | mordred | corvus: sorry about that | |
| 22:48:25 | corvus | mordred: oh noes you have the whitespace in that change | |
| 22:48:46 | mordred | NOT WHITESPACE OH NO WE'RE ALL DOOMED | |
| 22:49:25 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Add missing swift docstrings https://review.openstack.org/583790 | |
| 22:50:09 | mordred | corvus: now - your use case brings up an interesting point ... which is that I could see you _not_ wanting the "create client-side-checksums and upload as metadata" behavior | |
| 22:50:28 | mordred | corvus: which is on by default to help prevent double-uploads of large files | |