| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2020-02-20 | |||
| 20:39:33 | gtema_ | otherwise - no | |
| #openstack-sdks - 2020-02-21 | |||
| 00:01:28 | openstackgerrit | Sam Morrison proposed openstack/python-openstackclient master: Always display direction for security group rules https://review.opendev.org/709017 | |
| 07:50:34 | openstackgerrit | Andreas Jaeger proposed openstack/service-types-authority master: Minor cleanup to test publishing https://review.opendev.org/709062 | |
| 08:35:13 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: Lay a foundation for the project cleanup https://review.opendev.org/700219 | |
| 11:37:06 | openstackgerrit | Bram Verschueren proposed openstack/ansible-collections-openstack master: Cleanup unit test requirements https://review.opendev.org/709113 | |
| 12:18:50 | openstackgerrit | Igor Malinovskiy proposed openstack/python-openstackclient master: Add 'address_scope' type support to network rbac commands https://review.opendev.org/709124 | |
| 12:21:00 | openstackgerrit | Igor Malinovskiy proposed openstack/python-openstackclient master: Add 'address_scope' type support to network rbac commands https://review.opendev.org/709124 | |
| 14:02:30 | openstackgerrit | Roman Dobosz proposed openstack/openstacksdk master: Add bulk port create. https://review.opendev.org/708113 | |
| 14:02:30 | openstackgerrit | Roman Dobosz proposed openstack/openstacksdk master: Add method for bulk creating objects. https://review.opendev.org/708112 | |
| 15:12:05 | gtema | mordred: have you requested any PTG slot for SDK/Ansible @Vancouver? The deadline is near AFAIK | |
| 16:07:25 | openstackgerrit | Andreas Jaeger proposed openstack/service-types-authority master: Minor cleanup to test publishing https://review.opendev.org/709062 | |
| 19:56:00 | openstackgerrit | Merged openstack/service-types-authority master: Minor cleanup to test publishing https://review.opendev.org/709062 | |
| 20:11:27 | openstackgerrit | Andreas Jaeger proposed openstack/service-types-authority master: Remove html_last_updated_fmt https://review.opendev.org/709232 | |
| 20:56:24 | openstackgerrit | Merged openstack/service-types-authority master: Remove html_last_updated_fmt https://review.opendev.org/709232 | |
| #openstack-sdks - 2020-02-24 | |||
| 16:38:47 | openstackgerrit | Merged openstack/openstacksdk master: Add method for bulk creating objects. https://review.opendev.org/708112 | |
| #openstack-sdks - 2020-02-25 | |||
| 02:07:03 | openstackgerrit | Sam Morrison proposed openstack/python-openstackclient master: Always display direction for security group rules https://review.opendev.org/709017 | |
| 10:30:06 | openstackgerrit | Merged openstack/openstacksdk master: Add bulk port create. https://review.opendev.org/708113 | |
| 14:25:45 | openstackgerrit | Bharat Kunwar proposed openstack/ansible-collections-openstack master: Make os_coe_cluster module retrive id/uuid correctly https://review.opendev.org/709734 | |
| 14:26:51 | openstackgerrit | Bharat Kunwar proposed openstack/ansible-collections-openstack master: os_coe_cluster: Retrive id/uuid correctly https://review.opendev.org/709734 | |
| 14:29:40 | openstackgerrit | Bharat Kunwar proposed openstack/ansible-collections-openstack master: os_coe_cluster: Retrive id/uuid correctly https://review.opendev.org/709734 | |
| 14:48:33 | weust | Hello | |
| 14:49:21 | weust | I am trying to install do a pip install python-openstackclient 5.0.0 using Pythong 3.8.1 and that fails with the following: https://pastebin.com/0bVbUfxi | |
| 14:49:32 | weust | Is this already known? | |
| 14:50:30 | weust | And I checked the bug tracker. can't find it there. | |
| 14:53:46 | weust | Not really awake here. This is on Windows 10. | |
| 14:54:55 | gryf | weust, it fails on building dependency | |
| 14:55:16 | gryf | obviously you don't have needed header files | |
| 14:55:28 | weust | mkay.. | |
| 14:55:56 | weust | Sounds to me that is something pip should pull in | |
| 14:55:58 | gryf | I'm not even sure, if openstackclient will build on Windows | |
| 14:56:49 | weust | well, it complained before about needing Visual C++ build tools. so I guess it's aware of being run on Windows? | |
| 14:56:52 | gtema_ | I was having troubles running OSC on win7, due to some IPv6 stuff, so also not sure it will work. | |
| 14:57:25 | gtema_ | however my colleagues are running OSC in WSL, and apart of the terrible performance it worked for them | |
| 14:58:10 | weust | yeah, I have that in WSL too. but would be nicer to have it in Windows, instead of a wrapper. | |
| 14:58:34 | weust | also, the openstack doc page shows are installable on Windows. Except it's very old. using python 2.7 | |
| 14:59:34 | gtema | netifaces need to be build also under linux, so this is more a general problem how to build/install that | |
| 14:59:34 | weust | /s/are/as | |
| 15:00:10 | weust | I can see if it will build in my WSL Debian thingy | |
| 15:01:32 | weust | Except I have Pythong 3.7.something there | |
| 15:03:40 | weust | *sigh* debian install pythong 2.7 when installing python-dev and python-pip | |
| 15:06:08 | weust | well, apart from pyopenssl complaining about needing version 2.8 of higher, "but you'll have version older" blah, it seems to install | |
| 15:07:09 | weust | and it's still version 4.0.0 even though I installed with ==5.0.0 | |
| 15:09:20 | openstackgerrit | Bharat Kunwar proposed openstack/ansible-collections-openstack master: os_coe_cluster: Retrive id/uuid correctly https://review.opendev.org/709734 | |
| 19:07:37 | umbSublime | weust, yes most distros from 1+ year ago have python2 being the default /usr/bin/python, just make sure to install `python3 python3-dev python3-venv` then; python3 -m venv venv; source venv/bin/activate; pip install --upgrade pip; pip install python-openstackclient==5.0.0. version 5.0.0 is 3.6+ only hence why you got 4.0.0 when using python2 | |
| 20:14:07 | weust | umbSublime: I was able to get it working on a Ubuntu 18.4 LTS WSL today using python 3.8. Will try your actions too in a new setup. | |
| #openstack-sdks - 2020-02-26 | |||
| 13:12:09 | openstackgerrit | Michał Dulko proposed openstack/openstacksdk master: Implement If-Match support for Neutron resources https://review.opendev.org/710030 | |
| 13:12:09 | openstackgerrit | Michał Dulko proposed openstack/openstacksdk master: Implement If-Match support for Neutron resources https://review.opendev.org/710030 | |
| 13:17:33 | dulek | Folks, I could use some help with the above ^. I explained my concerns in the review. | |
| 13:23:00 | openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/python-openstackclient master: Fix network segment range "_get_ranges" function https://review.opendev.org/710031 | |
| 13:23:49 | openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/python-openstackclient master: Fix network segment range "_get_ranges" function https://review.opendev.org/710031 | |
| 14:10:03 | stephenfin | hberaud, smcginnis: Can one of you hit +W on this? https://review.opendev.org/#/c/705630/ | |
| 14:41:37 | openstackgerrit | Monty Taylor proposed openstack/ansible-collections-openstack master: Add a tool to build collections with pbr https://review.opendev.org/710047 | |
| 14:42:23 | gtema | mordred, would appreciate review on https://review.opendev.org/#/c/700219/ | |
| 14:43:03 | mordred | gtema: yeah - sorry - I keep opening it and then getting scared | |
| 14:43:22 | gtema | why, it's so straight forward ;-) | |
| 14:44:07 | gtema | dtantsur is now happy with it (hopefully). Since he is AFK can't verify, but we had some discussions on that | |
| 14:46:11 | mordred | gtema: yeah - it's looking good to me as I read it - nice work | |
| 14:46:19 | gtema | thks | |
| 14:46:47 | gtema | when this lands wanted to raise discussion where to make an agreed OSC plugin | |
| 14:47:02 | gtema | but anyway wanted to have a session on PTG wrt that | |
| 14:47:35 | mordred | gtema: one quibble - in project_cleanup - you've got code to handle people who are for some reason biased against threads - and that's the default code path ... why not make it use a thread pool by default and make people who want to "avoid" threads do work? (we use threads for chunked uploads to swift whether a user wants them or not, so I'm personally not even convinced we need to support avoiding | |
| 14:47:36 | mordred | them at all) | |
| 14:48:14 | gtema | dtantsur was vomplaining a lot agains that | |
| 14:48:19 | gtema | complaining | |
| 14:48:38 | mordred | one sec - gotta step away for 5 mins... | |
| 14:48:40 | gtema | what you said was a 2nd approach, but he still didn't like it | |
| 14:48:43 | gtema | ok | |
| 14:54:30 | mordred | k. back | |
| 14:54:43 | gtema | good | |
| 14:54:46 | mordred | gtema: so what's the issue - he's opposed to threads for some reason? | |
| 14:55:04 | mordred | oh - I see - kemme read the comments on the patch | |
| 14:55:14 | gtema | yeah, he was agains having it used in the default path | |
| 14:56:49 | mordred | gtema: his original comment was just about allowing the possibilities to pass in a manager: https://review.opendev.org/#/c/700219/7/openstack/cloud/openstackcloud.py@796 | |
| 14:57:07 | gtema | I think he was explaining me more concerns in the private chat, but I lost it and forgot | |
| 14:57:13 | mordred | nod | |
| 14:57:26 | gtema | and then he said - please avoid it completely | |
| 14:57:41 | gtema | we can wait for him then for details | |
| 14:57:56 | mordred | so - I completely agree about the greenlet concerns - because people using greenlet stuff are in a general world of compexity and pain as they have no clue what their code is doing at any time :) | |
| 14:58:05 | mordred | so definitely it's important to give those folks an out | |
| 14:58:11 | gtema | :D | |
| 14:58:35 | mordred | but - I think someone writing some straightforward code should not need to opt-in to what should be normal safe behavior | |
| 14:59:05 | mordred | they shold get the best experience out of the box just by running the cleanup method | |
| 14:59:08 | gtema | I share same opinion, so we are 2 against 1. | |
| 14:59:13 | mordred | and only need to do special things if they are in special circumstances | |
| 14:59:29 | gtema | then post a review to patch and we see what he says | |
| 14:59:40 | mordred | kk | |
| 15:00:00 | mordred | btw - the greenlet concern is good - we should maybe make sure we're providing a similar escape hatch for people for other uses of executors | |
| 15:00:10 | mordred | in fact- I _think_ you can pass an executor to Connection? | |
| 15:00:18 | mordred | one sec- lemme look | |
| 15:01:12 | mordred | nope. it's a TODO | |
| 15:01:29 | mordred | gtema: so - if you look in openstack/cloud/_object_store.py | |
| 15:01:49 | mordred | there's a TODO about making self.__pool_executor configurable - for the folks in the greenlet situation | |
| 15:02:13 | mordred | maybe we should just make that a proper Connection param- then we can use self.__pool_executor in your other code | |
| 15:02:33 | mordred | and it'll also be clear to people using sdk from services how to set up a connection safely | |
| 15:02:52 | gtema | a sec, have a call | |
| 15:02:53 | mordred | (maybe this is why we were considering switching to futurist at one point?) | |
| 15:02:59 | mordred | kk. I'll leave a review | |
| 15:06:06 | gtema | back. | |
| 15:06:09 | gtema | okay, great | |
| 15:06:12 | gtema | thanks | |