*** ameybhide has quit IRC | 00:05 | |
*** britthouser has joined #openstack-sdks | 00:28 | |
jamielennox | dtroyer_zz, terrylhowe : any reason the volume type create command doesn't exist in cinder v2? | 00:28 |
---|---|---|
dtroyer_zz | jamielennox: no idea, it was a recent add | 00:35 |
dtroyer_zz | v2 volume type commands | 00:35 |
jamielennox | dtroyer_zz: ok, well added a bug https://bugs.launchpad.net/python-openstackclient/+bug/1475060 and related https://bugs.launchpad.net/devstack/+bug/1475062 | 00:38 |
openstack | Launchpad bug 1475060 in python-openstackclient "Volume type create and list commands missing in v2" [Undecided,New] | 00:38 |
openstack | Launchpad bug 1475062 in devstack "Volume type fails with V2 API" [Undecided,New] | 00:38 |
*** yonglihe has joined #openstack-sdks | 00:43 | |
*** Qiming has joined #openstack-sdks | 00:49 | |
*** Qiming has quit IRC | 01:11 | |
*** openstack has joined #openstack-sdks | 01:25 | |
*** Yanyanhu has joined #openstack-sdks | 01:26 | |
*** Qiming has joined #openstack-sdks | 01:53 | |
openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: Add configuration function using os-client-config https://review.openstack.org/191535 | 02:03 |
*** hitalia has quit IRC | 02:49 | |
*** britthou_ has joined #openstack-sdks | 03:30 | |
*** britthouser has quit IRC | 03:33 | |
*** briancurtin has quit IRC | 03:45 | |
*** LinuxHorn has quit IRC | 03:45 | |
*** boris-42 has quit IRC | 03:46 | |
*** stevemar has joined #openstack-sdks | 03:48 | |
*** Klumben has quit IRC | 03:50 | |
*** LinuxHorn has joined #openstack-sdks | 03:56 | |
*** briancurtin has joined #openstack-sdks | 03:57 | |
*** boris-42 has joined #openstack-sdks | 04:07 | |
*** britthouser has joined #openstack-sdks | 04:15 | |
*** britthou_ has quit IRC | 04:16 | |
*** Klumben has joined #openstack-sdks | 04:47 | |
*** ameybhide has joined #openstack-sdks | 05:22 | |
*** hitalia has joined #openstack-sdks | 05:26 | |
*** Caleb-- has quit IRC | 05:51 | |
*** terrylhowe has quit IRC | 05:55 | |
*** Caleb-- has joined #openstack-sdks | 06:00 | |
openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Fix the way we call find_resource when only using ID https://review.openstack.org/202395 | 06:00 |
openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Fix the way we call find_resource when only using ID https://review.openstack.org/202395 | 06:01 |
*** ParsectiX has joined #openstack-sdks | 06:06 | |
*** ameybhide has quit IRC | 06:08 | |
openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Fix the way we call find_resource when only using ID https://review.openstack.org/202395 | 06:22 |
openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Make trustee/trustor/project searchable by ID https://review.openstack.org/200646 | 06:31 |
openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Fix the way we call find_resource when only using ID https://review.openstack.org/202395 | 06:31 |
openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Make trustee/trustor/project searchable by ID https://review.openstack.org/200646 | 06:31 |
*** aufi has joined #openstack-sdks | 06:38 | |
*** aufi has quit IRC | 06:38 | |
*** aufi has joined #openstack-sdks | 06:38 | |
*** stevemar has quit IRC | 06:42 | |
openstackgerrit | Guojian Shao proposed openstack/python-openstackclient: only return endpoints that have url https://review.openstack.org/201911 | 07:25 |
*** ParsectiX has quit IRC | 07:29 | |
*** stevemar has joined #openstack-sdks | 07:42 | |
*** stevemar has quit IRC | 07:47 | |
*** ParsectiX has joined #openstack-sdks | 07:51 | |
*** boris-42 has quit IRC | 08:32 | |
*** ParsectiX has quit IRC | 08:53 | |
*** yonglihe has quit IRC | 09:06 | |
*** ParsectiX has joined #openstack-sdks | 09:19 | |
*** stevemar has joined #openstack-sdks | 09:24 | |
*** stevemar has quit IRC | 09:28 | |
openstackgerrit | chengkunye proposed openstack/python-openstackclient: add image share for client https://review.openstack.org/202464 | 09:33 |
*** jaosorior has joined #openstack-sdks | 09:41 | |
*** Yanyanhu has quit IRC | 10:06 | |
*** Qiming has quit IRC | 10:27 | |
*** hitalia has quit IRC | 10:47 | |
openstackgerrit | chengkunye proposed openstack/python-openstackclient: add image share for client https://review.openstack.org/202464 | 11:01 |
*** thrash|g0ne is now known as thrash | 11:18 | |
*** Qiming has joined #openstack-sdks | 11:30 | |
*** Guest___ has joined #openstack-sdks | 11:34 | |
*** stevemar has joined #openstack-sdks | 12:24 | |
*** stevemar has quit IRC | 12:29 | |
*** stevemar has joined #openstack-sdks | 12:30 | |
*** stevemar has quit IRC | 12:35 | |
*** ParsectiX has quit IRC | 12:38 | |
*** terrylhowe has joined #openstack-sdks | 12:41 | |
*** trown|outttypeww is now known as trown | 12:42 | |
*** ParsectiX has joined #openstack-sdks | 12:46 | |
*** Guest___ is now known as smashwilson | 12:49 | |
*** smashwilson has quit IRC | 12:50 | |
*** stevemar has joined #openstack-sdks | 12:50 | |
*** smashwilson has joined #openstack-sdks | 12:50 | |
*** stevemar_ has joined #openstack-sdks | 12:51 | |
*** stevemar has quit IRC | 12:54 | |
*** stevemar_ has quit IRC | 12:55 | |
*** stevemar has joined #openstack-sdks | 12:56 | |
*** stevemar_ has joined #openstack-sdks | 12:57 | |
*** stevemar has quit IRC | 13:01 | |
openstackgerrit | Monty Taylor proposed openstack/os-client-config: Remove region list from single cloud https://review.openstack.org/202546 | 13:10 |
*** ParsectiX has quit IRC | 13:10 | |
*** ParsectiX has joined #openstack-sdks | 13:11 | |
*** bnemec has joined #openstack-sdks | 13:14 | |
*** guojian has joined #openstack-sdks | 13:17 | |
*** ParsectiX has quit IRC | 13:18 | |
*** petertr7_away is now known as petertr7 | 13:21 | |
*** bknudson has joined #openstack-sdks | 13:25 | |
*** pm90_ has joined #openstack-sdks | 13:26 | |
*** petertr7 is now known as petertr7_away | 13:34 | |
*** petertr7_away is now known as petertr7 | 13:35 | |
*** ParsectiX has joined #openstack-sdks | 13:40 | |
*** pm90_ has quit IRC | 13:55 | |
*** pm90_ has joined #openstack-sdks | 14:01 | |
*** sigmavirus24_awa is now known as sigmavirus24 | 14:02 | |
*** ParsectiX has quit IRC | 14:05 | |
*** zz_steveortiz is now known as steveortiz | 14:13 | |
*** fifieldt_ has quit IRC | 14:19 | |
stevemar_ | dtroyer_zz: ping | 14:39 |
stevemar_ | dtroyer_zz: can you 2x check bknudson's comment here: https://review.openstack.org/#/c/196297/3/keystoneclient/auth/cli.py | 14:40 |
dtroyer_zz | we actually don't use that… | 14:49 |
*** notmars has joined #openstack-sdks | 14:59 | |
*** karimb has joined #openstack-sdks | 15:02 | |
*** notmars has quit IRC | 15:09 | |
*** trown is now known as trown|brb | 15:19 | |
*** trown|brb is now known as trown | 15:24 | |
openstackgerrit | Guojian Shao proposed openstack/python-openstackclient: add --project-domain option for user v3 https://review.openstack.org/202645 | 15:27 |
*** stevelle_ is now known as stevelle | 15:27 | |
*** petertr7 is now known as petertr7_away | 15:34 | |
*** petertr7_away is now known as petertr7 | 15:35 | |
*** guojian has quit IRC | 15:37 | |
*** terrylhowe has left #openstack-sdks | 15:40 | |
*** ig0r_ has quit IRC | 15:40 | |
*** smashwilson has quit IRC | 15:44 | |
*** pm90_ has quit IRC | 15:44 | |
*** smashwilson has joined #openstack-sdks | 15:44 | |
*** pm90_ has joined #openstack-sdks | 15:44 | |
*** trown is now known as trown|lunch | 15:58 | |
*** pm90__ has joined #openstack-sdks | 16:01 | |
*** terrylhowe has joined #openstack-sdks | 16:01 | |
*** pm90_ has quit IRC | 16:04 | |
*** hitalia has joined #openstack-sdks | 16:05 | |
*** petertr7 is now known as petertr7_away | 16:18 | |
*** smashwilson has quit IRC | 16:22 | |
*** pm90__ has quit IRC | 16:25 | |
*** pm90_ has joined #openstack-sdks | 16:26 | |
*** pm90_ has quit IRC | 16:26 | |
*** pm90_ has joined #openstack-sdks | 16:27 | |
*** smashwilson has joined #openstack-sdks | 16:27 | |
*** notmars has joined #openstack-sdks | 16:30 | |
*** stevemar_ has quit IRC | 16:32 | |
openstackgerrit | Guojian Shao proposed openstack/python-openstackclient: enhance tests for user v3 https://review.openstack.org/202687 | 16:35 |
*** Qiming has quit IRC | 16:44 | |
*** pm90_ has quit IRC | 16:52 | |
*** notmars has quit IRC | 17:06 | |
*** pm90_ has joined #openstack-sdks | 17:07 | |
*** pm90_ has quit IRC | 17:08 | |
*** pm90_ has joined #openstack-sdks | 17:08 | |
*** petertr7_away is now known as petertr7 | 17:09 | |
*** petertr7 is now known as petertr7_away | 17:19 | |
*** pm90_ has quit IRC | 17:29 | |
*** pm90_ has joined #openstack-sdks | 17:29 | |
*** karimb has quit IRC | 17:30 | |
*** hitalia has quit IRC | 17:38 | |
*** aufi has quit IRC | 17:43 | |
*** bknudson has quit IRC | 17:45 | |
*** trown|lunch is now known as trown | 17:46 | |
*** stevemar has joined #openstack-sdks | 18:08 | |
*** petertr7_away is now known as petertr7 | 18:09 | |
openstackgerrit | Doug Hellmann proposed openstack/cliff: Set demo app up with deferred help https://review.openstack.org/202730 | 18:13 |
shaleh | question for changes to OSC. Is it preferred to catch things like OSError in the take_action methods or to let them bubble up to the higher level handler? I do not see much error checking in many of the classes. | 18:22 |
*** pm90_ has quit IRC | 18:24 | |
terrylhowe | mostly errors are fatal, so they should just bubble upt shaleh | 18:27 |
shaleh | terrylhowe: ok. thanks. | 18:29 |
*** hitalia has joined #openstack-sdks | 18:38 | |
*** pm90_ has joined #openstack-sdks | 18:43 | |
shaleh | what about a case where there is a partial failure? | 18:44 |
shaleh | I noticed that OSC does not support image create for v2 glance api, so I added it with a little guidance from dtroyer. | 18:45 |
shaleh | a significant change in the API is that now the image create just sets up a queued image. The upload is needed to complete it. | 18:46 |
*** jaosorior has quit IRC | 18:46 | |
shaleh | But obviously the upload can fail. So I would like to show the queued image ID so it can either be passed in or deleted via a call to image delete. | 18:46 |
shaleh | I usually use existing behavior as a clue but I am not finding many existing classes with rich error handling. | 18:47 |
shaleh | Ideally OSC could say "image created with id <uuid>, pass that in the next time" | 18:48 |
dtroyer_zz | shaleh: in most cases we just pass up the returned errors from the API as they are 'good enough'. in this case I agree it needs to be more graceful, especially as it is doing more than one operation per command | 18:49 |
dtroyer_zz | the actual formatting of the output isn't obvious since I don't think we've needed to provide both returned info and error info | 18:50 |
shaleh | dtroyer_zz: exactly | 18:50 |
*** petertr7 is now known as petertr7_away | 18:50 | |
dtroyer_zz | but the first step could be to return the image ID (and whatever else) as usal with the error on stdout | 18:51 |
shaleh | dtroyer_zz: one option is making it part of the exception text | 18:51 |
shaleh | I will start there and let the review suggest improvements | 18:51 |
dtroyer_zz | I thought image create v1 already had an option to pass in an image ID to handle the reservations business | 18:52 |
shaleh | dtroyer_zz: I have it working, now for the pretty | 18:52 |
shaleh | dtroyer_zz: yes, --id exists in both API | 18:52 |
*** petertr7_away is now known as petertr7 | 18:52 | |
dtroyer_zz | does that over the 'handle previouus attempt' case? | 18:52 |
shaleh | point for me is OSC image create --file foo.img called twice, first is a failure leads to a queued image and another complete image | 18:52 |
shaleh | dtroyer_zz: in APIv1, the upload is part of the create so it is folded into the API error return | 18:53 |
shaleh | I suppose passing --id the next time does the right thing | 18:53 |
shaleh | here the client is responsible for informing the user or deleting the queued image | 18:53 |
shaleh | both are valid options | 18:53 |
shaleh | if the id came back for the create I could just invoke a delete on it | 18:54 |
dtroyer_zz | right. maybe deleting as part of the error handler is the right thing? | 18:54 |
shaleh | if the id did not come back then we can assume bad args or busted network | 18:54 |
dtroyer_zz | do we have any idea what glance users actually want to happen here? | 18:54 |
dtroyer_zz | would be a good thing to see what mordred and shade folk are doing… | 18:55 |
terrylhowe | I guess that is the usecase for creating empty images. create and upload later so you delete if it fails | 18:55 |
shaleh | if the id was not provided by the user, and one comes back then I will delete it on error. Otherwise I will include the id in the error output. | 18:55 |
dtroyer_zz | they may still be using v1 though | 18:55 |
shaleh | others can comment from there | 18:55 |
mordred | uhoh | 18:55 |
mordred | what did I do? | 18:55 |
dtroyer_zz | shaleh: I think that's a good place to start | 18:55 |
dtroyer_zz | mordred: image uploads… in v2 it is a two step process, if step 2 fails (the actual image upload) it leaves a queued image ID behind | 18:56 |
mordred | yes. | 18:56 |
* shaleh is adding support to OSC for image create/upload | 18:56 | |
dtroyer_zz | I'm thinking we should delete the queued image and start over rather than return the image ID and tell the user to try again but specify the ID | 18:56 |
dtroyer_zz | but first, is shade using image v2 yet? | 18:56 |
shaleh | and pointing out that error handling in OSC leaves something to be desired | 18:56 |
mordred | dtroyer_zz: yes - for PUT I agree | 18:56 |
mordred | dtroyer_zz: yes - we support ALL of the image upload combinations | 18:57 |
mordred | it's a LOT of logic to do so, fyi | 18:57 |
mordred | because there are two ways to upload images in v2 | 18:57 |
mordred | both of them are 2 steps | 18:57 |
dtroyer_zz | right, one reason I'm opinionated about this | 18:57 |
mordred | and you kind of have to know which one you want | 18:57 |
mordred | if you do the task version | 18:57 |
mordred | you actually don't want to delete, you just want to resubmit the task | 18:57 |
mordred | dtroyer_zz: oh - so - I left out a thign we do | 18:58 |
openstackgerrit | Brant Knudson proposed openstack/python-openstackclient: Query args for get-only https://review.openstack.org/202748 | 18:58 |
mordred | which is we do an md5 and sha1 calculation so that we can skip unnecessary parts of re-uploading | 18:58 |
shaleh | mordred: what is the "we" here? | 18:58 |
mordred | oh -sorry - shade | 18:58 |
dtroyer_zz | I don't see the CLI using task version much, we're stateless between commands, so it would be in a singe command invocation | 18:58 |
terrylhowe | osc meeting in 2 minutes? | 18:58 |
mordred | dtroyer_zz: ++ | 18:58 |
mordred | dtroyer_zz: if I were you, I'd implemet the task logic but hide it | 18:58 |
dtroyer_zz | yes, thanks for the reminder terrylhowe | 18:58 |
*** bknudson has joined #openstack-sdks | 18:59 | |
mordred | dtroyer_zz: so that OSC shows "image upload" and does what it needed | 18:59 |
shaleh | terrylhowe: in -meeting? | 18:59 |
mordred | but that might just be me | 18:59 |
dtroyer_zz | shaleh: make sense? and yes, -meeting | 18:59 |
terrylhowe | yeh shaleh | 18:59 |
dtroyer_zz | mordred: you have much more actual user experience here than most of us ;) | 18:59 |
dtroyer_zz | thanks | 18:59 |
mordred | dtroyer_zz: :) | 18:59 |
shaleh | thank you morded | 18:59 |
shaleh | lots and lots to learn still | 19:00 |
mordred | shaleh: I recommend (highly) at least reading through teh image_create codepath in shade | 19:00 |
shaleh | mordred: shall do. Thanks for the pointer. | 19:00 |
shaleh | where do I find shade? | 19:00 |
mordred | shaleh: http://git.openstack.org/cgit/openstack-infra/shade/tree/shade/__init__.py#n1391 | 19:00 |
shaleh | ah, infra | 19:00 |
*** pm90_ has quit IRC | 19:01 | |
mordred | it runs from line 1391 to 1552 | 19:01 |
*** pm90_ has joined #openstack-sdks | 19:01 | |
*** rbrooker has joined #openstack-sdks | 19:01 | |
shaleh | eewww, lots of code in __init__ | 19:03 |
*** trown is now known as trown|outttypeww | 19:03 | |
mordred | shaleh: yeah - ignore that part - we've got some reorg to do later | 19:03 |
*** pm90__ has joined #openstack-sdks | 19:08 | |
openstackgerrit | Merged openstack/cliff: Allow subcommands to accept --help when using 'deferred_help'. https://review.openstack.org/201258 | 19:08 |
*** pm90_ has quit IRC | 19:10 | |
*** bknudson has quit IRC | 19:19 | |
*** petertr7 is now known as petertr7_away | 19:23 | |
*** petertr7_away is now known as petertr7 | 19:23 | |
*** dguerri` is now known as dguerri | 19:26 | |
*** petertr7 is now known as petertr7_away | 19:31 | |
openstackgerrit | Brant Knudson proposed openstack/python-openstackclient: Same exception handling for gets() in find_resource https://review.openstack.org/202760 | 19:50 |
openstackgerrit | Brant Knudson proposed openstack/python-openstackclient: Same exception handling for gets() in find_resource https://review.openstack.org/202760 | 19:52 |
shaleh | dtroyer_zz: re: api-list. If you want to talk about implementing it drop me an email or whatever communication you prefer. | 19:55 |
dhellmann | dtroyer_zz, stevemar, terrylhowe : instead of the prefix based match, what do you think about using levenshtein distance, like git does? | 20:05 |
dhellmann | that's regarding https://review.openstack.org/#/c/202053/13 | 20:05 |
terrylhowe | I guess it should at least share the same algorithm with help.py dhellmann | 20:08 |
*** smashwilson has quit IRC | 20:10 | |
dhellmann | terrylhowe: heh, true. Both could share the implementation | 20:10 |
openstackgerrit | Brant Knudson proposed openstack/python-openstackclient: Same exception handling for gets() in find_resource https://review.openstack.org/202760 | 20:11 |
*** pm90__ has quit IRC | 20:26 | |
openstackgerrit | Terry Howe proposed openstack/os-client-config: Rename 'endpoint_type' to 'interface' https://review.openstack.org/201704 | 20:41 |
openstackgerrit | Terry Howe proposed openstack/os-client-config: Have service name default to None https://review.openstack.org/201311 | 20:42 |
openstackgerrit | Terry Howe proposed openstack/os-client-config: Rename 'endpoint_type' to 'interface' https://review.openstack.org/201704 | 20:47 |
openstackgerrit | Terry Howe proposed openstack/os-client-config: Remove py26 and py33 from tox.ini https://review.openstack.org/202784 | 20:52 |
*** hitalia has quit IRC | 21:09 | |
*** stevemar has quit IRC | 21:09 | |
*** stevemar has joined #openstack-sdks | 21:12 | |
*** stevemar has quit IRC | 21:17 | |
openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: Add vendor to service filter https://review.openstack.org/197741 | 21:21 |
openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: Add provider to service filter https://review.openstack.org/197741 | 21:21 |
*** notmars has joined #openstack-sdks | 21:28 | |
*** pm90_ has joined #openstack-sdks | 21:38 | |
*** pm90__ has joined #openstack-sdks | 21:40 | |
*** hitalia has joined #openstack-sdks | 21:41 | |
*** boris-42 has joined #openstack-sdks | 21:42 | |
*** pm90_ has quit IRC | 21:43 | |
*** bnemec has quit IRC | 21:46 | |
openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: Dynamically load services https://review.openstack.org/155362 | 21:56 |
*** pm90__ has quit IRC | 21:56 | |
*** notmars has quit IRC | 21:56 | |
openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: Dynamically load services https://review.openstack.org/155362 | 21:57 |
openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: Add provider to service filter https://review.openstack.org/197741 | 21:58 |
*** pm90_ has joined #openstack-sdks | 22:02 | |
*** pm90__ has joined #openstack-sdks | 22:03 | |
*** pm90_ has quit IRC | 22:06 | |
shaleh | back to ask about UI some more | 22:31 |
shaleh | glance image-create has a --progress to show as the upload of the image occurs | 22:31 |
shaleh | ideas on what that should look like in OSC? | 22:31 |
dtroyer_zz | I had not planned on implementing that, any idea how popular it really is? | 22:32 |
shaleh | dtroyer_zz: not really | 22:32 |
shaleh | I was considering punting on it for now to be honest | 22:32 |
dtroyer_zz | that's one of the things we can ask in the UX study | 22:32 |
dtroyer_zz | it's easier to add later than to take it away when it becomes a problem | 22:33 |
shaleh | dtroyer_zz: ++ | 22:33 |
terrylhowe | progress bar on object store might be nice too | 22:34 |
shaleh | terrylhowe: that was my concern too. It needed to be added and shared. | 22:34 |
shaleh | In theory we could re-use the one in python-glanceclient BTW. | 22:41 |
shaleh | it is right there to import | 22:41 |
terrylhowe | well, using glanceclient for object store would be awkward if that is what you mean | 22:45 |
shaleh | terrylhowe: :-) no. | 22:49 |
shaleh | I mean the progress meter that glance uses is in its common/ dir and we already use other code from that module. | 22:49 |
shaleh | So we could borrow it for now and port it to either a common lib or OSC directly | 22:50 |
*** openstackstatus has joined #openstack-sdks | 23:34 | |
*** ChanServ sets mode: +v openstackstatus | 23:34 | |
openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: Add wait for delete method https://review.openstack.org/193793 | 23:38 |
*** steveortiz is now known as zz_steveortiz | 23:40 | |
*** pm90__ has quit IRC | 23:44 | |
*** britthouser has quit IRC | 23:47 | |
*** britthouser has joined #openstack-sdks | 23:50 | |
*** dguerri is now known as dguerri` | 23:54 | |
*** barra204 has joined #openstack-sdks | 23:54 | |
*** pm90_ has joined #openstack-sdks | 23:56 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!