| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2017-03-14 | |||
| 03:02:16 | stevemar | reedip: same principals i believe, install OSC in the virtualenv then the other repo | |
| 03:03:28 | reedip | lemme try | |
| 03:06:29 | reedip | stevemar : what if requirements are not met :) | |
| 03:09:37 | reedip | stevemar : getting an error | |
| 03:09:42 | reedip | in requirements | |
| 04:13:23 | ankur-gupta-f4 | reedip: it should work in the same way | |
| 04:13:37 | ankur-gupta-f4 | R u talking about the TaaS commands? | |
| 04:23:47 | rabel__b1 | can someone help me with https://review.openstack.org/#/c/444924/ ? two functional tests in the gate job are failing, but it seems to have nothing to do with the change. one of the test i also saw failing for other changes that seem to be unrelated. | |
| 04:24:31 | ankur-gupta-f4 | rabel__b1: the image tests are temporarily broken. | |
| 04:25:00 | ankur-gupta-f4 | Till its resolved avoid rechecking | |
| 04:36:49 | rabel__b1 | ankur-gupta-f4: ok, thank you. | |
| 04:52:14 | reedip | ankur-gupta-f4 : all working :) dont worry .. I worked around stevemar's solution :D | |
| 04:52:53 | ankur-gupta-f4 | K | |
| 05:04:59 | openstackgerrit | Lv Jiawei proposed openstack/python-openstackclient master: Add extra filtering options to qos policy list https://review.openstack.org/401131 | |
| 08:23:29 | openstackgerrit | Colleen Murphy proposed openstack/python-openstackclient master: Narrow expected responses for CheckUserInGroup https://review.openstack.org/445257 | |
| 08:29:25 | openstackgerrit | Lv Jiawei proposed openstack/python-openstackclient master: Add status filtering options to port list https://review.openstack.org/444810 | |
| 11:11:17 | openstackgerrit | Lv Jiawei proposed openstack/python-openstackclient master: Add status filtering options to port list https://review.openstack.org/444810 | |
| 11:53:17 | rabel | does anyone know about the status of the broken image functional tests for osc? | |
| 13:07:59 | elmiko | cdent: saw your ping way too late yesterday, isn't that email basically sending us back down the openapi rabbit hole? | |
| 13:08:29 | cdent | elmiko: I think so, but wasn't sure, so wanted to check in. | |
| 13:08:49 | cdent | Thanks for pinging me about it though, as I completely forgot to follow up like I promised I would | |
| 13:09:00 | elmiko | hehe =) | |
| 13:49:22 | openstackgerrit | Bence Romsics proposed openstack/python-openstacksdk master: Add port property: trunk_details https://review.openstack.org/445501 | |
| 14:50:31 | openstackgerrit | Pepijn Oomen proposed openstack/python-openstacksdk master: Expose ha_state property from HA enabled L3 Agents. https://review.openstack.org/445520 | |
| 14:53:34 | reedip | o/ | |
| 15:14:19 | ynirk | I faced a problem with the command `openstack user show 12345 --domain Default` returning the error "get() got an unexpected keyword argument 'domain_id'" | |
| 15:14:19 | ynirk | I am working on an openstack infrastructure where some usernames are only numeric (12345 for example) in a specific domain | |
| 15:14:19 | ynirk | Hi, I hope I am in the right place | |
| 15:14:20 | ynirk | I am more of an ops so I prefer asking for your advice before submitting a bug and a patch. | |
| 15:14:20 | ynirk | Wouldn't it be more coherent with the other try/except to have a 'pass' at https://github.com/openstack/osc-lib/blob/master/osc_lib/utils.py#L124 instead of a 'raise' and let the function raise a NotFound at the end if no resource was found ? | |
| 15:14:20 | openstack | Launchpad bug 1671179 in python-keystoneclient "Unable to retrieve a numeric only username within a domain" [Undecided,In progress] - Assigned to Julien Lavesque (ynirk) | |
| 15:14:20 | ynirk | All get requests (in all openstack components) made with a numeric name and kwargs might return a TypeError if the method's prototype doesn't handle kwargs. | |
| 15:14:20 | ynirk | I first thought of a keystoneclient problem (with the bug https://bugs.launchpad.net/python-keystoneclient/+bug/1671179) but after deeper analysis I think the problem is in find_resource from osc-lib/utils.py | |
| 15:14:21 | ynirk | What do you think ? | |
| 15:16:45 | breton | ynirk: > All get requests (in all openstack components) made with a numeric name and kwargs might return a TypeError if the method's prototype doesn't handle kwargs. | |
| 15:16:51 | breton | ynirk: why is that? | |
| 15:21:50 | dtroyer | ynirk: I think you may be right there | |
| 15:22:22 | dtroyer | I am trying to recall what exceptions I thought should be raised at that point | |
| 15:23:49 | dtroyer | this function is a combination of 3 or 4 of these from individual libs, so working back to that may be fun. Changing it now may also change things in unexpected ways | |
| 15:24:33 | dtroyer | the choices are to add a type(ex) check for TypeError or to remove them altogether | |
| 15:25:31 | ynirk | dtroyer yes exactly but I was not sure of the impact | |
| 15:25:33 | dtroyer | I'm going to have to think on that a bit. if you have time, you may want to try proposing temoving the if in lines 120-124 and just doing a pass there and see what tests/jobs fail. | |
| 15:26:20 | dtroyer | it would be safer to add the TypeError check though | |
| 15:28:15 | ynirk | Ok I also think it is safer and shouldn't break anything | |
| 15:28:28 | ynirk | I was wondering of the reason for the raise (compared to the other try/except) | |
| 15:31:57 | ynirk | I will try to add the TypeError check and see if some tests fail | |
| 15:32:09 | ynirk | Thanks for the feedback | |
| 15:51:11 | josecastroleon | hi | |
| 15:51:35 | josecastroleon | the gate in openstackclient is failing since devstack removed glance v1 | |
| 15:51:58 | ankur-gupta-f4 | yes | |
| 15:53:05 | josecastroleon | shall we remove the glance v1 test? | |
| 15:53:47 | ankur-gupta-f4 | https://www.irccloud.com/pastebin/p3o3Cgo2/ | |
| 15:54:05 | dtroyer | no, we need to continue to support v1 for some time, for now I'd prefer to re-enable Glance v1 in DevStack | |
| 15:54:20 | dtroyer | at least for a time | |
| 15:54:36 | ankur-gupta-f4 | stevemar suggested doing a check first. If exception raised then skip tests | |
| 15:54:45 | josecastroleon | fine, but then we may need to enable glance v1 in the gate | |
| 15:54:48 | josecastroleon | right? | |
| 15:55:14 | dtroyer | yes, that's on my list, I'm fixing cliff's gate ATM | |
| 15:55:23 | josecastroleon | ok thx | |
| 16:04:59 | openstackgerrit | Dean Troyer proposed openstack/cliff master: Update cmd2 fix to still work with 0.6.7 https://review.openstack.org/445552 | |
| 16:06:10 | dtroyer | josecastroleon, ankur-gupta-f1: https://review.openstack.org/#/c/445553/ re-enables Image v1 for OSC functional jobs | |
| 16:18:55 | openstackgerrit | Merged openstack/golang-client master: Update for CTI ps4 https://review.openstack.org/444499 | |
| 16:27:55 | openstackgerrit | Bence Romsics proposed openstack/python-openstacksdk master: Add port property: trunk_details https://review.openstack.org/445501 | |
| 16:30:59 | ynirk | dtroyer tests pass with the two modifications. To be sure I will add the TypeError check | |
| 16:43:36 | reedip | sindhu : hi | |
| 16:44:45 | sindhu | reedip: hey wass up | |
| 16:45:12 | reedip | nothin much sindhu : just wanted to know about the floating ip patch | |
| 16:45:30 | sindhu | reedip: it's all ready right? | |
| 16:45:53 | reedip | no comments from anyone else, I guess | |
| 16:45:56 | reedip | :| | |
| 16:46:53 | sindhu | reedip: yeah! waiting for reviews/getting merged :) | |
| 16:47:47 | reedip | stevemar , can you come to our rescue ???? :D | |
| 18:52:40 | openstackgerrit | Merged openstack/python-openstacksdk master: Add port property: trunk_details https://review.openstack.org/445501 | |
| 18:54:37 | openstackgerrit | Merged openstack/cliff master: Update cmd2 fix to still work with 0.6.7 https://review.openstack.org/445552 | |
| 21:53:59 | ZZelle | dtroyer, hi | |
| 23:26:44 | openstackgerrit | Julien Lavesque proposed openstack/osc-lib master: Fix find_resource exception handling on numeric names with kwargs https://review.openstack.org/445679 | |
| #openstack-sdks - 2017-03-15 | |||
| 07:27:34 | openstackgerrit | Rui Chen proposed openstack/python-openstackclient master: [Compute]Make column content readable for both human and machine https://review.openstack.org/445794 | |
| 08:46:22 | openstackgerrit | Rui Chen proposed openstack/python-openstackclient master: [Compute]Make column content readable for both human and machine https://review.openstack.org/445794 | |
| 10:07:17 | openstackgerrit | lihaijing proposed openstack/python-openstackclient master: [1/3]Replace six.iteritems() with .items() https://review.openstack.org/445856 | |
| 10:32:22 | openstackgerrit | lihaijing proposed openstack/python-openstackclient master: [2/3]Replace six.iteritems() with .items() https://review.openstack.org/445865 | |
| 11:13:22 | openstackgerrit | lihaijing proposed openstack/python-openstackclient master: [3/3]Replace six.iteritems() with .items() https://review.openstack.org/445878 | |
| 11:58:39 | frickler | so it seems like the project-config patch did not fix the gate: http://logs.openstack.org/75/443075/2/check/gate-osc-dsvm-functional-ubuntu-xenial/5eb7ede/console.html#_2017-03-15_11_12_51_996706 | |
| 12:01:04 | openstackgerrit | Chris Dent proposed openstack/api-wg master: Refactor and re-validate api change guidelines https://review.openstack.org/421846 | |
| 12:06:05 | frickler | I guess https://review.openstack.org/445900 would work better | |
| 16:29:29 | benoit | hello | |
| 16:30:42 | benoit | i have a strange behavior using openstacksdk (network class) : i want to retrieve project_id for a network but it always return None | |
| 16:31:49 | benoit | using **query parameters with project_id also seems to not be taken in account | |
| 16:32:28 | benoit | is this parameter set ? or do i miss something ? | |
| 16:36:38 | cdent | sdague: you have an opinion on what should happen here: https://review.openstack.org/#/c/434999/ there's a conflict between the value of the service type and already existing documentation | |
| 17:03:09 | reedip | benoit : whats the request you are sending and whats the response u r getting | |
| 17:05:25 | benoit | reedip, just digging into RESP BODY, it has tenant_id not project_id | |
| 17:05:53 | benoit | reedip, on a more recent openstack, project_id is set and sdk is working as expected | |
| 17:08:07 | benoit | (first openstack is a liberty, second a newton) not sure how to use openstacksdk with liberty then | |
| 17:41:04 | dtroyer | benoit: what you are seeing is a difference between Identity v2 and v3 behaviour | |
| 17:41:14 | dtroyer | re tenant vs project | |
| 17:41:37 | dtroyer | the SDK is supposed to hide much of that from you | |
| 17:42:39 | briancurtin | benoit: can you share some code? what you’re talking about doesn’t really make sense to me to figure out what’s going on | |
| 17:44:41 | briancurtin | the things that come back in a response body aren’t names you’re going to use, that’s raw data out of the server | |
| 17:48:58 | briancurtin | also, project_id actually has to be a query parameter on the REST end of it in order to send it, and we have to have added support for this, so you can’t arbitrarily add query parameters. if something is documented to accept project_id as a query parameter we can add it, of course | |
| 18:23:10 | openstackgerrit | Pepijn Oomen proposed openstack/python-openstacksdk master: Expose OS-EXT-SRV-ATTR:{hypervisor_hostname,instance_name} for Server. https://review.openstack.org/446104 | |