| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2019-03-18 | |||
| 15:42:05 | mordred | brtknr: which one? | |
| 15:42:46 | openstackgerrit | Eric Fried proposed openstack/keystoneauth master: Factor Adapter conf-processing logic into a helper https://review.openstack.org/644251 | |
| 15:43:06 | brtknr | mordred: the workshop! | |
| 15:44:27 | mordred | sweet! well - hopefully that goes well and you don't all end up throwing rotten vegetables at me | |
| 15:45:28 | brtknr | mordred: you'll do fine I'm sure :) | |
| 15:45:53 | openstackgerrit | Eric Fried proposed openstack/keystoneauth master: Factor Adapter conf-processing logic into a helper https://review.openstack.org/644251 | |
| 15:46:35 | efried | oh look, those get posted here too ^ | |
| 15:46:52 | mordred | efried: it's almost like these pieces all depend on each other | |
| 15:46:58 | efried | heh | |
| 17:49:31 | openstackgerrit | Eric Fried proposed openstack/openstacksdk master: WIP Make factory for a CloudRegion from CONF objects https://review.openstack.org/643601 | |
| 17:49:49 | efried | mordred: Still need to test locally but ^ | |
| 17:50:22 | efried | ahcrap, forgot depends-on... | |
| 17:51:03 | openstackgerrit | Eric Fried proposed openstack/openstacksdk master: WIP Make factory for a CloudRegion from CONF objects https://review.openstack.org/643601 | |
| 17:59:59 | mordred | efried: update looks good | |
| 18:39:43 | efried | mordred: It once again suffers the NoSuchOptError problem, and now I don't have a good way to fix it. Going to reinstate your conditional import thing, unless you think I shouldn't bother and should just trap Exception. | |
| 18:42:18 | mordred | efried: can you not do project_name in config? (I guess that didn't work out?) | |
| 18:43:27 | efried | oslo_config.cfg.NoSuchOptError: no such option valid_interfaces in group [cinder] | |
| 18:43:27 | efried | mordred: that part would work; what breaks is if that group didn't import ksa opts, then you get e.g. | |
| 18:43:28 | mordred | efried: it looks like a ConfigOpts is a collections.Mapping and defines __contains__ | |
| 18:43:39 | mordred | oh! different issue | |
| 18:44:00 | efried | mordred: so I would have to, like, load up the conf opts and query the group whether it has 'em before invoking the helper method, which seemed like horrible man. | |
| 18:45:35 | mordred | efried: we could have the ksa method do a hasattr before every get? is that a terrible idea? | |
| 18:45:36 | efried | I could put something in the helper to convert the exception to something standard, or return a bool, or... but all of that is really ugly too. | |
| 18:45:56 | mordred | efried: or we could just trap Exception in sdk until it becomes an issue | |
| 18:46:04 | efried | yup, that's what I'm setting up now. | |
| 18:50:18 | openstackgerrit | Eric Fried proposed openstack/openstacksdk master: WIP Make factory for a CloudRegion from CONF objects https://review.openstack.org/643601 | |
| 18:50:22 | efried | mordred: thar she blows | |
| 18:50:56 | mordred | efried: I believe in her | |
| 18:51:29 | efried | now dep updates | |
| #openstack-sdks - 2019-03-19 | |||
| 09:51:29 | dtantsur | API SIG people, we seem to have some PTG space on Thu morning (and it does not conflict with ironic) | |
| 10:56:14 | openstackgerrit | Merged openstack/openstacksdk master: Fix wait_for_server docstring https://review.openstack.org/643955 | |
| 11:21:55 | openstackgerrit | Merged openstack/openstacksdk master: Add unit tests for connection.add_service https://review.openstack.org/643604 | |
| 13:37:33 | openstackgerrit | Merged openstack/openstacksdk master: Update master for stable/stein https://review.openstack.org/644016 | |
| 14:13:24 | openstackgerrit | Merged openstack/osc-lib master: Update master for stable/stein https://review.openstack.org/644042 | |
| 14:57:24 | openstackgerrit | Merged openstack/openstacksdk master: added support for binding:profile parameter in create_port/update_port https://review.openstack.org/642418 | |
| 14:57:25 | openstackgerrit | Merged openstack/openstacksdk master: baremetal: implement the correct update of the maintenance_reason field https://review.openstack.org/643639 | |
| 15:54:19 | openstackgerrit | Merged openstack/cliff master: add python 3.7 unit test job https://review.openstack.org/637766 | |
| 19:56:09 | dustinc | mordred: is there a way to define retry policy for sdk proxy? | |
| 19:56:27 | dustinc | related to my question just now in #openstack-ironic | |
| 19:56:46 | dustinc | I didn't see anything when I checked source real fast but might have missed it? 🤞 | |
| 21:50:52 | mordred | dustinc: there is a set of retry support in there that dtantsur|afk added for ironic but works for all the proxies | |
| 21:52:19 | mordred | dustinc: the basic idea is that the proxy knows what status codes are safe to retry, and then it's possible to configure for each service how many retries should be attempted for retriable status codes | |
| 21:52:40 | mordred | baremteal_status_code_retries would be the config setting you'd be looking for for ironic - and it defaults to 5 | |
| 22:05:09 | dustinc | oh thanks, I will go looking for that | |
| 22:06:01 | efried | nice! | |
| 22:06:39 | efried | ...before, that is. | |
| 22:09:50 | efried | dustinc: So you don't have to do a thing - it's already handled and defaulted appropriately. | |
| 22:09:54 | efried | \o/ | |
| 22:10:10 | dustinc | !!! | |
| 22:10:11 | openstack | dustinc: Error: "!!" is not a valid command. | |
| #openstack-sdks - 2019-03-20 | |||
| 13:41:16 | nullscan | hi | |
| 13:42:07 | nullscan | i have been trying to use the python openstack client with osprofiler but when i try to execute any command with --os-profile HMAC_KEY the client responds that the --os-profile switch is invalid | |
| 13:42:19 | nullscan | am i missing something ? | |
| 13:43:18 | nullscan | the client is installed using debian package and the version is 3.16.2 | |
| 13:43:32 | nullscan | the same happens when i try to use the latest version installed via pip | |
| 13:48:15 | dtroyer | nullscan: what version of osc-lib? That is where the —os-profile option is actually declared, although it has been around for a while... | |
| 13:48:52 | nullscan | 1.11.1 | |
| 13:50:53 | dtroyer | That is a global option so it needs to appear on the command line before the actual command: openstack —os-profile qaz server list | |
| 13:51:23 | nullscan | yes, i have actually been testing with openstack --os-profile qwerty image list | |
| 13:51:27 | nullscan | but still no luck | |
| 13:51:52 | dtroyer | hmmm… it's giving me trouble too (I've never actually used it before) | |
| 13:52:13 | dtroyer | oh, duh, liiking at the code... | |
| 13:52:14 | nullscan | openstack: '--os-profile qqqq image list' is not an openstack command. See 'openstack --help'. Did you mean one of these? | |
| 13:52:17 | dtroyer | *looking | |
| 13:52:18 | nullscan | that is what i am getting | |
| 13:52:25 | dtroyer | do you have osprofiler installed? | |
| 13:52:32 | nullscan | yes i do | |
| 13:52:48 | dtroyer | if impporting osprofiler.profiler fails the option isn't defined | |
| 13:52:56 | nullscan | but from what i understand osprofiler is used only to get information from trackIDs | |
| 13:53:26 | nullscan | and the actual documentation does not list any other steps, just enable profiling in the services i want and run any command with --os-profile switch | |
| 13:53:35 | dtroyer | so if OSc is in a venv that venv also needs to have osprofiler in it, etc | |
| 13:54:37 | nullscan | no its not in venv, its installed along with the rest of my system | |
| 13:56:20 | dtroyer | what do you get with: python -m osprofiler | |
| 13:56:21 | nullscan | ah i just found it | |
| 13:56:40 | nullscan | nevemind, i had the python3 version of osc-lib installed instead of the python v2 ones | |
| 13:56:41 | nullscan | :) | |
| 13:56:49 | dtroyer | aha! | |
| 13:56:54 | nullscan | thanks for your help though! | |
| 13:56:59 | dtroyer | np | |
| 23:49:11 | zer0c00l | Is there an example of how to mock openstacksdk calls | |
| 23:49:20 | zer0c00l | For example i mocked openstack.connect | |
| 23:49:43 | zer0c00l | But how do i mock openstack.baremetal.chassis() call | |
| 23:52:12 | zer0c00l | The main app connects to a cluster using sdk = openstack.connect(cloud="a.b.c") | |
| 23:53:00 | zer0c00l | with the sdk object i fetch chassis by running sdk.baremetal.chassis() . Now i would like to mock both openstack.connect and sdk.baremetal.chassis calls | |
| 23:53:08 | zer0c00l | Is there any examples somewhere i can look? | |
| #openstack-sdks - 2019-03-21 | |||
| 00:59:52 | zer0c00l | Looks like i just needed to mock openstack.connection.Connection | |
| 00:59:58 | zer0c00l | Not anything else | |
| 10:39:16 | cdent | dtantsur: that wsme fix needed a requirements change before it was good, so I went ahead and did it | |
| 10:39:26 | cdent | which has also reminded me to remove myself from wsme core ... | |
| 10:39:47 | dtantsur | heh | |
| 15:19:12 | openstackgerrit | Merged openstack-infra/shade master: Update master for stable/stein https://review.openstack.org/644221 | |
| 15:21:19 | openstackgerrit | Merged openstack/os-client-config master: Update master for stable/stein https://review.openstack.org/644024 | |
| 15:59:29 | elmiko | i feel like we need some intro music | |
| 16:00:01 | edleafe | The API-SIG Office Hour has begun. Don't crowd; we'll get to all of you in turn. | |
| 16:00:02 | cdent | "she's the one" by bruce springsteen just came around randomly on my musc. how about that? | |
| 16:00:16 | elmiko | lol cdent | |
| 16:00:23 | cdent | it has good intro piano | |
| 16:00:43 | edleafe | I prefer the Ramones version | |
| 16:01:30 | edleafe | https://www.youtube.com/watch?v=E5Ql6tOPM90 | |
| 16:01:30 | elmiko | wow, there's a ramones version!?! | |