| *** gildub has joined #openstack-sdks | 00:12 | |
| *** hoangcx has joined #openstack-sdks | 00:39 | |
| *** chlong_ has joined #openstack-sdks | 01:12 | |
| *** chlong has joined #openstack-sdks | 01:12 | |
| *** zhurong has joined #openstack-sdks | 01:15 | |
| *** kevo has joined #openstack-sdks | 01:19 | |
| *** hoangcx has quit IRC | 01:35 | |
| *** hoangcx has joined #openstack-sdks | 01:36 | |
| *** hoangcx2 has joined #openstack-sdks | 01:38 | |
| *** hoangcx has quit IRC | 01:40 | |
| *** hoangcx has joined #openstack-sdks | 01:42 | |
| *** hoangcx2 has quit IRC | 01:45 | |
| openstackgerrit | ayoung proposed openstack/keystoneauth: Fix parameters for Kerberos Auth Plugin https://review.openstack.org/368288 | 01:48 |
|---|---|---|
| *** hoangcx2 has joined #openstack-sdks | 01:49 | |
| *** hoangcx has quit IRC | 01:51 | |
| *** hieulq has joined #openstack-sdks | 01:52 | |
| openstackgerrit | ayoung proposed openstack/keystoneauth: Fix parameters for Kerberos Auth Plugin https://review.openstack.org/368288 | 01:58 |
| *** jamielennox is now known as jamielennox|away | 02:00 | |
| *** jamielennox|away is now known as jamielennox | 02:04 | |
| openstackgerrit | ayoung proposed openstack/keystoneauth: Fix parameters for Kerberos Auth Plugin https://review.openstack.org/368288 | 02:13 |
| openstackgerrit | QiangTang proposed openstack/python-openstackclient: Add image project set command https://review.openstack.org/367224 | 02:17 |
| *** namnh has joined #openstack-sdks | 02:17 | |
| *** hoangcx has joined #openstack-sdks | 02:17 | |
| *** hoangcx2 has quit IRC | 02:18 | |
| *** lifeless_ has quit IRC | 02:19 | |
| openstackgerrit | QiangTang proposed openstack/python-openstackclient: Add image project list command https://review.openstack.org/366492 | 02:19 |
| openstackgerrit | QiangTang proposed openstack/python-openstackclient: Refactor image project add / remove https://review.openstack.org/365910 | 02:24 |
| *** prg3 has quit IRC | 02:27 | |
| openstackgerrit | QiangTang proposed openstack/python-openstackclient: Support mark volume as bootable in volume set https://review.openstack.org/360191 | 02:28 |
| *** prg3 has joined #openstack-sdks | 02:31 | |
| *** lifeless has joined #openstack-sdks | 02:34 | |
| *** hoangcx has quit IRC | 02:42 | |
| *** hoangcx has joined #openstack-sdks | 02:42 | |
| openstackgerrit | Rui Chen proposed openstack/python-openstackclient: Support to list specified server's ports https://review.openstack.org/357027 | 02:42 |
| openstackgerrit | Tang Chen proposed openstack/python-openstackclient: Set default api version for global options https://review.openstack.org/355417 | 02:43 |
| openstackgerrit | QiangTang proposed openstack/python-openstackclient: Fix openstack quota set/show --class not work https://review.openstack.org/351581 | 02:44 |
| *** yuanying has quit IRC | 02:51 | |
| *** RuiChen has joined #openstack-sdks | 02:52 | |
| openstackgerrit | Tang Chen proposed openstack/python-openstackclient: Trivial: Rename ListTransferRequests to ListTransferRequest https://review.openstack.org/362625 | 02:58 |
| *** hoangcx2 has joined #openstack-sdks | 03:00 | |
| *** hoangcx has quit IRC | 03:00 | |
| openstackgerrit | Rui Chen proposed openstack/python-openstackclient: Set quota "per_volume_gigabytes", "backup_gigabytes" and "backups" https://review.openstack.org/352286 | 03:02 |
| openstackgerrit | Rui Chen proposed openstack/python-openstackclient: Unit tests of quota command refactor https://review.openstack.org/352285 | 03:02 |
| *** hoangcx has joined #openstack-sdks | 03:03 | |
| *** hoangcx2 has quit IRC | 03:04 | |
| openstackgerrit | Rui Chen proposed openstack/python-openstackclient: Support fetching network project default quota https://review.openstack.org/343616 | 03:07 |
| openstackgerrit | Rui Chen proposed openstack/python-openstackclient: Add functional tests for unset multiple volume type props https://review.openstack.org/335338 | 03:09 |
| openstackgerrit | Rui Chen proposed openstack/python-openstackclient: Support "--no" and "--all" option in aggregate set and unset https://review.openstack.org/310193 | 03:15 |
| *** RuiChen has quit IRC | 03:29 | |
| *** RuiChen has joined #openstack-sdks | 03:31 | |
| *** yuanying has joined #openstack-sdks | 03:50 | |
| *** amotoki has joined #openstack-sdks | 04:05 | |
| *** gouthamr has joined #openstack-sdks | 04:19 | |
| openstackgerrit | Merged openstack/python-openstackclient: Add unit test for volume type in volume v1 https://review.openstack.org/367230 | 04:27 |
| *** salv-orlando has joined #openstack-sdks | 04:52 | |
| *** chlong_ has quit IRC | 05:03 | |
| *** chlong has quit IRC | 05:03 | |
| *** chlong_ has joined #openstack-sdks | 05:16 | |
| *** chlong has joined #openstack-sdks | 05:20 | |
| *** chlong has quit IRC | 05:25 | |
| openstackgerrit | lvdongbing proposed openstack/python-openstacksdk: Fix cluster query mapping https://review.openstack.org/367878 | 05:57 |
| *** chlong_ has quit IRC | 06:05 | |
| reedip | amotoki: ping | 06:13 |
| *** hoangcx2 has joined #openstack-sdks | 06:27 | |
| *** hoangcx has quit IRC | 06:27 | |
| openstackgerrit | Rui Chen proposed openstack/python-openstackclient: Support "--no" and "--all" option in aggregate set and unset https://review.openstack.org/310193 | 06:29 |
| *** gouthamr has quit IRC | 07:01 | |
| *** gouthamr has joined #openstack-sdks | 07:02 | |
| *** gouthamr has quit IRC | 07:03 | |
| *** kevo has quit IRC | 07:15 | |
| *** hoangcx2 has quit IRC | 07:25 | |
| *** hoangcx has joined #openstack-sdks | 07:27 | |
| *** adriant has quit IRC | 07:30 | |
| *** jpenag is now known as jpena | 07:37 | |
| *** hoangcx2 has joined #openstack-sdks | 07:38 | |
| *** hoangcx has quit IRC | 07:39 | |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/python-openstackclient: Add network support for "qos policies" https://review.openstack.org/350655 | 07:59 |
| *** gildub has quit IRC | 08:00 | |
| *** zhurong has quit IRC | 08:03 | |
| *** hoangcx2 has quit IRC | 08:06 | |
| *** hoangcx2 has joined #openstack-sdks | 08:06 | |
| *** hoangcx has joined #openstack-sdks | 08:10 | |
| *** hoangcx2 has quit IRC | 08:10 | |
| *** Nh3 has joined #openstack-sdks | 08:12 | |
| *** Romanenko_K has joined #openstack-sdks | 08:14 | |
| *** lucas-afk is now known as lucasagomes | 08:17 | |
| *** zhurong has joined #openstack-sdks | 08:28 | |
| *** salv-orl_ has joined #openstack-sdks | 08:31 | |
| *** salv-orlando has quit IRC | 08:34 | |
| *** openstackgerrit has quit IRC | 09:04 | |
| *** openstackgerrit has joined #openstack-sdks | 09:05 | |
| *** hoangcx has quit IRC | 09:08 | |
| *** hoangcx has joined #openstack-sdks | 09:09 | |
| *** mfedosin has joined #openstack-sdks | 09:43 | |
| *** hoangcx has quit IRC | 09:58 | |
| *** zhurong has quit IRC | 10:04 | |
| *** markvoelker has joined #openstack-sdks | 10:22 | |
| *** markvoelker has quit IRC | 10:26 | |
| *** e0ne has joined #openstack-sdks | 10:27 | |
| *** Nh3 has quit IRC | 10:33 | |
| *** shu-mutou is now known as shu-mutou-AFK | 10:39 | |
| *** cdent has joined #openstack-sdks | 10:52 | |
| *** fzdarsky has joined #openstack-sdks | 11:02 | |
| *** rtheis has joined #openstack-sdks | 11:06 | |
| *** Nh3 has joined #openstack-sdks | 11:22 | |
| *** lucasagomes is now known as lucas-hungry | 11:25 | |
| *** namnh has quit IRC | 11:27 | |
| *** thrash|g0ne is now known as thrash | 11:32 | |
| *** sdague has joined #openstack-sdks | 11:34 | |
| openstackgerrit | Ukesh Kumar proposed openstack/osc-lib: Incorrect usage message when no auth param passed https://review.openstack.org/368768 | 11:37 |
| *** mliima has joined #openstack-sdks | 11:47 | |
| openstackgerrit | Monty Taylor proposed openstack/os-client-config: Fix a bunch of tests https://review.openstack.org/368776 | 11:51 |
| openstackgerrit | Monty Taylor proposed openstack/os-client-config: Split auth plugin loading into its own method https://review.openstack.org/368777 | 11:51 |
| openstackgerrit | Monty Taylor proposed openstack/os-client-config: Remove validate_auth_ksc https://review.openstack.org/368778 | 11:51 |
| openstackgerrit | Monty Taylor proposed openstack/os-client-config: Defer plugin loading until it's needed https://review.openstack.org/368779 | 11:51 |
| mordred | dtroyer: I poked a little at the occ auth plugin loading thing over the weekend - but I know you were also poking, so I changed the change-id's on https://review.openstack.org/368777 and https://review.openstack.org/368778 to make new changes so I wouldn't step on you | 11:52 |
| mordred | dtroyer: mostly wanted to get to https://review.openstack.org/368779 ... that's an idea I've been mulling recently and wanted to see what you think | 11:52 |
| *** markvoelker has joined #openstack-sdks | 12:02 | |
| *** jpena is now known as jpena|lunch | 12:07 | |
| *** fzdarsky has quit IRC | 12:11 | |
| *** lucas-hungry is now known as lucasagomes | 12:29 | |
| openstackgerrit | Merged openstack/python-openstacksdk: Fix cluster query mapping https://review.openstack.org/367878 | 12:33 |
| *** Nh3 has quit IRC | 12:43 | |
| *** Nh3 has joined #openstack-sdks | 12:43 | |
| openstackgerrit | Merged openstack/python-openstackclient: remove duplicate unit test https://review.openstack.org/368237 | 12:45 |
| reedip | RuiChen : commented on the bug https://review.openstack.org/308201 | 12:48 |
| reedip | tangchen : Can you also see my response on ^^ ?? | 12:48 |
| *** cleong has joined #openstack-sdks | 12:54 | |
| *** mfedosin has quit IRC | 13:11 | |
| *** openstackgerrit has quit IRC | 13:19 | |
| *** openstackgerrit has joined #openstack-sdks | 13:20 | |
| openstackgerrit | Merged openstack/python-openstacksdk: Rebase keystone resources onto resource2 https://review.openstack.org/365314 | 13:20 |
| openstackgerrit | Qiming Teng proposed openstack/python-openstacksdk: Fix 'config_drive' and 'networks' for compute server https://review.openstack.org/368837 | 13:21 |
| *** jpena|lunch is now known as jpena | 13:22 | |
| openstackgerrit | Monty Taylor proposed openstack/os-client-config: Remove validate_auth_ksc https://review.openstack.org/368778 | 13:23 |
| openstackgerrit | Monty Taylor proposed openstack/os-client-config: Defer plugin loading until it's needed https://review.openstack.org/368779 | 13:23 |
| openstackgerrit | Monty Taylor proposed openstack/os-client-config: Fix a bunch of tests https://review.openstack.org/368776 | 13:23 |
| openstackgerrit | Monty Taylor proposed openstack/os-client-config: Split auth plugin loading into its own method https://review.openstack.org/368777 | 13:23 |
| *** fzdarsky has joined #openstack-sdks | 13:26 | |
| openstackgerrit | Merged openstack/python-openstacksdk: Add network segment create, delete and update support https://review.openstack.org/302223 | 13:29 |
| *** annegentle has joined #openstack-sdks | 13:29 | |
| *** hoangcx has joined #openstack-sdks | 13:35 | |
| *** annegentle has quit IRC | 13:36 | |
| *** annegentle has joined #openstack-sdks | 13:37 | |
| *** mfedosin has joined #openstack-sdks | 13:37 | |
| *** zhurong has joined #openstack-sdks | 13:38 | |
| *** tamilhce has joined #openstack-sdks | 13:44 | |
| *** hoangcx has quit IRC | 13:51 | |
| *** hoangcx has joined #openstack-sdks | 13:52 | |
| *** hoangcx2 has joined #openstack-sdks | 13:57 | |
| *** rvasilets__ has left #openstack-sdks | 13:58 | |
| *** hoangcx has quit IRC | 13:59 | |
| *** fguillot has joined #openstack-sdks | 14:15 | |
| openstackgerrit | Merged openstack/python-openstackclient: Unit tests of quota command refactor https://review.openstack.org/352285 | 14:17 |
| openstackgerrit | Qiming Teng proposed openstack/python-openstacksdk: Fix 'config_drive' and 'networks' for compute server https://review.openstack.org/368837 | 14:30 |
| *** salv-orlando has joined #openstack-sdks | 14:30 | |
| *** salv-orl_ has quit IRC | 14:33 | |
| openstackgerrit | Qiming Teng proposed openstack/python-openstacksdk: Rebase keystone v3 proxy to proxy2 https://review.openstack.org/365331 | 14:38 |
| *** annegentle has quit IRC | 14:39 | |
| *** beekneemech is now known as bnemec | 14:43 | |
| kzaitsev_ws | is check-osc-plugins red by default in stable/mitaka? | 14:52 |
| *** annegentle has joined #openstack-sdks | 14:53 | |
| openstackgerrit | Merged openstack/python-openstacksdk: Fix 'config_drive' and 'networks' for compute server https://review.openstack.org/368837 | 14:54 |
| *** zhurong has quit IRC | 14:55 | |
| *** lucasagomes is now known as lucas-afk | 15:02 | |
| *** jgriffith has quit IRC | 15:03 | |
| *** Nh3 has quit IRC | 15:04 | |
| *** Nh3 has joined #openstack-sdks | 15:06 | |
| openstackgerrit | Qiming Teng proposed openstack/python-openstacksdk: Fix connection init when session is provided https://review.openstack.org/368907 | 15:08 |
| openstackgerrit | Merged openstack/python-openstacksdk: Rebase keystone v3 proxy to proxy2 https://review.openstack.org/365331 | 15:08 |
| Qiming | rtheis, please help take a look at #368907 above when you have bandwidth | 15:09 |
| *** gouthamr has joined #openstack-sdks | 15:09 | |
| Qiming | it is a fix to problems raised in rally | 15:09 |
| Qiming | thanks | 15:09 |
| rtheis | lgtm | 15:12 |
| *** thrash is now known as thrash|biab | 15:13 | |
| openstackgerrit | Qiming Teng proposed openstack/python-openstacksdk: Fix connection init when session is provided https://review.openstack.org/368907 | 15:13 |
| *** Nh3 has quit IRC | 15:22 | |
| *** Nh3 has joined #openstack-sdks | 15:28 | |
| *** e0ne has quit IRC | 15:41 | |
| *** Nh3 has quit IRC | 15:42 | |
| *** hoangcx2 has quit IRC | 15:52 | |
| jidar | is there no way to do this: http://docs.openstack.org/user-guide/cli-nova-launch-instance-from-volume.html#create-volume-from-image-and-boot-instance with the python-openstackclient ? | 15:55 |
| *** kevo has joined #openstack-sdks | 15:56 | |
| *** kevo has quit IRC | 15:57 | |
| *** fzdarsky has quit IRC | 15:57 | |
| openstackgerrit | Brian Curtin proposed openstack/python-openstacksdk: Add extended Glance Image properties https://review.openstack.org/368943 | 16:04 |
| *** lucas-afk is now known as lucasagomes | 16:04 | |
| openstackgerrit | Merged openstack/python-openstacksdk: Fix connection init when session is provided https://review.openstack.org/368907 | 16:05 |
| *** mfedosin has quit IRC | 16:21 | |
| *** jgriffith_ has joined #openstack-sdks | 16:24 | |
| *** thrash|biab is now known as thrash | 16:26 | |
| *** jgriffith_ has quit IRC | 16:29 | |
| *** jgriffith has joined #openstack-sdks | 16:35 | |
| *** annegentle has quit IRC | 16:40 | |
| *** annegentle has joined #openstack-sdks | 16:42 | |
| *** tamilhce has quit IRC | 16:49 | |
| openstackgerrit | Merged openstack/python-openstackclient: Trivial: Rename ListTransferRequests to ListTransferRequest https://review.openstack.org/362625 | 17:06 |
| *** jpena is now known as jpena|off | 17:06 | |
| *** kevo has joined #openstack-sdks | 17:10 | |
| Anticimex | briancurtin: did you have a chance to look at the compute list issue? | 17:11 |
| briancurtin | Anticimex: i looked briefly and nothing stuck out. i will try to take a look again later | 17:12 |
| openstackgerrit | Merged openstack/python-openstackclient: Add functional tests for unset multiple volume type props https://review.openstack.org/335338 | 17:23 |
| *** fzdarsky has joined #openstack-sdks | 17:23 | |
| *** e0ne has joined #openstack-sdks | 17:38 | |
| openstackgerrit | Merged openstack/python-openstackclient: Set quota "per_volume_gigabytes", "backup_gigabytes" and "backups" https://review.openstack.org/352286 | 17:59 |
| *** ukesh has joined #openstack-sdks | 18:04 | |
| ukesh | join | 18:04 |
| ukesh | https://bugs.launchpad.net/python-ironicclient/+bug/1619091 | 18:06 |
| openstack | Launchpad bug 1619091 in python-ironicclient "[RFE] Extend OSC "node list" command to fetch nodes without instance UUID" [Wishlist,Confirmed] - Assigned to Ukesh (ukeshkumar) | 18:06 |
| ukesh | I couldn't find the code for the command 'openstack baremetal node list' | 18:07 |
| ukesh | in openstackclient project | 18:07 |
| *** ukesh has quit IRC | 18:10 | |
| stevemar | why did ukesh run away :( | 18:15 |
| *** lucasagomes is now known as lucas-dinner | 18:16 | |
| *** fzdarsky has quit IRC | 18:19 | |
| *** fzdarsky has joined #openstack-sdks | 18:21 | |
| dtroyer | stevemar: looks like we're not going to get a release until g-r is branched, end of this week at the earliest | 18:24 |
| dtroyer | so let's start (carefully!) merging stuff. carefully == not all at once, and things that make sense to go together. We can see where we are Wednesday afternoon and re-cut the release review then. no sense in wasting a whole week (again) | 18:25 |
| *** Romanenko_K has quit IRC | 18:34 | |
| stevemar | dtroyer: the next release will be for the "ocata" series, right? | 18:42 |
| dtroyer | yes | 18:42 |
| stevemar | dtroyer: okay, we merged a bunch of patches over the weekend (bug fixes and such) -- i thought the flood gates were open? | 18:44 |
| stevemar | 10 or so | 18:44 |
| dtroyer | I thought for cleanups and tests and whatnot. | 18:45 |
| dtroyer | no reason to hold back now, just not all at once | 18:45 |
| stevemar | dtroyer: okay, i also abandoned as many patches :) | 18:45 |
| dtroyer | I _really_ don't know where we stand these days | 18:46 |
| *** fzdarsky has quit IRC | 18:47 | |
| *** fzdarsky has joined #openstack-sdks | 18:48 | |
| openstackgerrit | Brian Curtin proposed openstack/python-openstacksdk: Add extended Glance Image properties https://review.openstack.org/368943 | 18:55 |
| openstackgerrit | Tom Jose Kalapura proposed openstack/python-openstackclient: Add sort support to project list https://review.openstack.org/335204 | 18:55 |
| openstackgerrit | Merged openstack/python-openstacksdk: Obtain Image checksum via additional GET https://review.openstack.org/367459 | 19:05 |
| *** e0ne has quit IRC | 19:08 | |
| openstackgerrit | Merged openstack/python-openstacksdk: Add extended Glance Image properties https://review.openstack.org/368943 | 19:18 |
| stevemar | dtroyer: sorry, was in a meeting, i figure we are restricted by the release team | 19:31 |
| dtroyer | yes | 19:31 |
| briancurtin | rtheis, Qiming: i just pushed 0.9.6 of the SDK to get a lot of the recent changes out there: https://pypi.python.org/pypi/openstacksdk/0.9.6 | 19:33 |
| stevemar | dtroyer: the osc version in "stable/newton" is 3.2.0, the only critical fix not included is the prompting one; 3.3.0 will be released in ocata; i figure it can contain non-critical fixes too, like we did with the 2.x.0 releases | 19:33 |
| rtheis | briancurtin: thank you | 19:33 |
| dtroyer | stevemar: right, we can go forward sensibly now | 19:33 |
| Anticimex | briancurtin: i guess i can 'triage' with older pip-versions of openstacksdk to find one that works | 19:35 |
| briancurtin | Anticimex: i would try 0.9.1 — that’s the release before we changed endpoint determination from straight out of the service catalog to using what the service tells us | 19:36 |
| briancurtin | Anticimex: also, can i ask more about what kind of cloud you’re connecting to? is this one you’ve setup on your own (ie. devstack) or is this something inside your company, or something you’re paying a vendor for? so far our changes have worked everywhere except one case we had to adjust, so I’m wondering what’s going on out in the world that this | 19:38 |
| briancurtin | doesn’t work once again. there are some moving parts that don’t seem to be working together | 19:38 |
| Anticimex | briancurtin: i'm connecting to our own cloud, we're a cloud provider | 19:40 |
| Anticimex | using what the service tells the user would probably be fine, except, the URL fed to connect to the service is wrong in my bug case | 19:41 |
| Anticimex | so problem is before the query to the base API of the compute service | 19:41 |
| Anticimex | where "base API" == https://$nova-url/ | 19:42 |
| stevemar | briancurtin: do you not use the releases repo for sdk releases? | 19:43 |
| briancurtin | we get the URL to connect to the service out of the service catalog, but we just use the root of the service (which then spits out versioned URLs we take). which provider is this? maybe i can just make an account and try it myself? | 19:44 |
| briancurtin | stevemar: no idea what that is | 19:44 |
| stevemar | briancurtin: https://github.com/openstack/releases | 19:44 |
| stevemar | :) | 19:44 |
| briancurtin | i push a tag and magic happens | 19:44 |
| stevemar | briancurtin: its all good | 19:44 |
| Anticimex | 0.9.1: RESP BODY: {"servers": []} | 19:45 |
| stevemar | briancurtin: the upper-constraints won't change: https://github.com/openstack/requirements/blob/master/upper-constraints.txt | 19:45 |
| Anticimex | which is correct in my case, so problem there somewhere | 19:45 |
| Anticimex | diff'ing time | 19:45 |
| stevemar | briancurtin: we'll merge patches in a bit once the requirements repo opens up again | 19:45 |
| briancurtin | Anticimex: yeah, there’s some disconnect then between what your service catalog has for the nova endpoint and what we’re calling and/or what we’re receiving from nova itself. hmm | 19:46 |
| Anticimex | i basically stopped debugging it on the finish line on friday. i think i'll now soon see the changes in the function that i had reached then | 19:48 |
| Anticimex | yeah. one of these: https://github.com/openstack/python-openstacksdk/blob/0.9.5/openstack/session.py#L89 , https://github.com/openstack/python-openstacksdk/blob/0.9.5/openstack/session.py#L110 | 19:51 |
| Anticimex | i think the latter | 19:51 |
| Anticimex | that's where i ended up the debugging on friday | 19:51 |
| Anticimex | i.e. "just take the root" fails in our case for some reason | 19:51 |
| briancurtin | Anticimex: i think you mentioned that some other service worked despite that issue with nova, perhaps it was like neutron list networks or something? i wonder what is different between your nova and neutron roots or url structures? | 19:53 |
| briancurtin | (or whichever one it was that worked) | 19:53 |
| Anticimex | yeah, i can test that too | 19:53 |
| Anticimex | first just figure out why _get_endpoint_versions given the input in this case doesn't give the expected output | 19:54 |
| Anticimex | briancurtin: https://gist.github.com/Millnert/5294ae0e0efc9d574509f04a652afe64 | 20:05 |
| Anticimex | representing https://github.com/openstack/python-openstacksdk/blob/0.9.5/openstack/session.py#L117-L121 | 20:05 |
| *** cleong has quit IRC | 20:05 | |
| Anticimex | the if seems wrong | 20:06 |
| Anticimex | added a comment | 20:07 |
| briancurtin | hmm, yeah…let me see what that’s doing there | 20:07 |
| Anticimex | devstack has : for port always? :] | 20:08 |
| * Anticimex has no idea | 20:08 | |
| Anticimex | :-) | 20:08 |
| Anticimex | ok the Markdown ate my comment.. | 20:08 |
| briancurtin | Anticimex: ah, so that was changed for https://github.com/openstack/python-openstacksdk/commit/56cd1a924725d7b52edeeb984d6e66fd17f42b5c — devstack has port numbers, but apache services don’t end up using ports i guess | 20:09 |
| Anticimex | but anyway | 20:09 |
| Anticimex | right, once it's live on the internets the nicest thing is to put it on default https right | 20:09 |
| briancurtin | so in your case, we really need “https://nova.api.cloud.ipnett.se/“ to be what’s returned there | 20:09 |
| Anticimex | right | 20:09 |
| Anticimex | what's the actual objective of that part of the function? | 20:10 |
| Anticimex | strip away the path and keep url up to first / ? | 20:10 |
| Anticimex | at any rate parts.netloc doesn't appear to be the thing one wants there | 20:11 |
| briancurtin | Anticimex: that part should be returning the root of the service. in your case it should be giving https://nova.api.cloud.ipnett.se/ and in the case of something like devstack it’d give http://mydevstack:1234/ | 20:11 |
| Anticimex | no that's wrong of me. parts.netloc is the thing, but the : isn't always there | 20:12 |
| Anticimex | nod, so it's to catch the port number if there is such a thing? | 20:12 |
| Anticimex | roger | 20:13 |
| briancurtin | yep, so yeah that’s a little off given these cases | 20:13 |
| *** mfedosin has joined #openstack-sdks | 20:14 | |
| briancurtin | hm, although if you have a port in there it keeps it in the netloc anyway…so now I’m not sure why we changed this... | 20:14 |
| Anticimex | so that change has the right intention but, shrug, wasn't tested at whatever set of things we have running that makes our catalogue return something like 'https://nova.api.cloud.ipnett.se/v2/<some_tenant_uuid>' | 20:14 |
| Anticimex | i'm slightly confused as well for the actual code of the change :) | 20:14 |
| Anticimex | netloc isn't complete without the port | 20:15 |
| Anticimex | (as per http or similar) | 20:15 |
| Anticimex | aha, it's by intention preserving the path, because "the path part is the identity of the service" | 20:16 |
| *** fangxu has joined #openstack-sdks | 20:16 | |
| Anticimex | that's a pretty wild assumption | 20:16 |
| *** fguillot has quit IRC | 20:17 | |
| *** mliima has quit IRC | 20:18 | |
| briancurtin | yeah, your case is a mixture of both things. no port, but you also don’t want the path parts. i wonder if we should try to make a request without the path, see if it works, if not make one with the path, and then if *that* one doesn’t work, raise an exception or something saying we can’t figure out wtf is going on? i don’t love this, but don’t have | 20:20 |
| briancurtin | a great suggestion to handle three different cases unless we resort to making users pass in something (which i don’t want to have to rely on) | 20:20 |
| Anticimex | no, code trialing is much better | 20:20 |
| Anticimex | so the assumed counter case to the one with non-default port http/https, seems something like the case where you have something like "api.example.com/nova" or some such? | 20:21 |
| Anticimex | there are quite large dragons around these assumptions :] | 20:21 |
| briancurtin | yeah, i think that’s it | 20:21 |
| briancurtin | Anticimex: yeah, this whole project exists to fight the million openstack dragons | 20:22 |
| Anticimex | one could perhaps add some "intelligence" and make guesses if the first part of the path is something like 'vX[.y]' or so, but that's also errorprone i guess | 20:22 |
| Anticimex | briancurtin: awesome! | 20:22 |
| Anticimex | i'm not sure how much one can rely on the fact that the service catalogue, when having per-service dns names such as in our case, always will contain "$hostname/vX[.y/foo" | 20:23 |
| Anticimex | especially if this is to be generic over all endpoints | 20:24 |
| Anticimex | "... rely on the [presumed] fact ..." | 20:24 |
| openstackgerrit | Merged openstack/python-openstackclient: Fix regular expression for uptime in hypervisor show https://review.openstack.org/353555 | 20:25 |
| Anticimex | i guess the heuristics approach such as i suggested is also insufficient. but it could at least be used as an heuristics to suggest which query to try first | 20:27 |
| briancurtin | Anticimex: so should we do the urlsplit, then make requests for x.scheme+"://“+x.netloc as well as x.scheme+”://“+x.netloc+x.path — one of those should give the root of the service, right? although i guess if there’s a tenant in there, the second way wouldn’t work either | 20:27 |
| Anticimex | i still don't like the whole ':' part either. anyone running api.example.com:8443/nova will be screwed :] | 20:28 |
| Anticimex | right | 20:28 |
| Anticimex | that should work i guess | 20:28 |
| briancurtin | though i guess we could also check if your tenant id is included in the path and remove it? | 20:28 |
| Anticimex | right, tenant one wouldn't be base of query | 20:28 |
| *** d0ugal has quit IRC | 20:28 | |
| *** cdent has quit IRC | 20:28 | |
| Anticimex | briancurtin: yeah, but i guess there are a myriad different ways these can show up | 20:29 |
| Anticimex | i mean, whatever comes after the $service | 20:29 |
| briancurtin | yeah, you can literally do anything you want and it’s called openstack | 20:29 |
| Anticimex | :) | 20:29 |
| Anticimex | the source of the tenant confusion is the request to the service catalogue right | 20:30 |
| Anticimex | is it possibly possible to learn something from the state at the time of that request? | 20:30 |
| Anticimex | i'm assuming here you can have different OS_ENVs or whatever and perhaps not get a tenant back from the service catalogue always | 20:30 |
| *** salv-orl_ has joined #openstack-sdks | 20:31 | |
| * Anticimex doens't know the strictness of the service catalogue either, so a bit hard to speculate | 20:31 | |
| briancurtin | i think what we could do is make that request with no path, make a request with a path, and if those two fail, rely on the service catalog itself? after that, if your requests aren’t going through, you may have to call an exorcist | 20:31 |
| dtroyer | fwiw, getting $tenant_id out of the service catalog is a goal, but one that has to be done carefully. Part of that is making clients be able to handle both ways… | 20:32 |
| briancurtin | dtroyer: so us trying a few different requests up front (once per Connection instance) depending on what we get is fine/acceptable/what is expected right now? | 20:33 |
| *** salv-orlando has quit IRC | 20:34 | |
| briancurtin | dtroyer: this started because we couldn’t find identity v3 in the SC, so we had to go to the root, and then it was mentioned that we should always go to the root, and now finding the root is sometimes a challenge. i think ultimately we’re moving the right way, right? | 20:34 |
| briancurtin | (the root=the root of a given service) | 20:34 |
| Anticimex | seems to me, more discovery less assumption is better, at least | 20:34 |
| openstackgerrit | Tom Jose Kalapura proposed openstack/python-openstackclient: Add sort support to project list https://review.openstack.org/335204 | 20:35 |
| Anticimex | briancurtin: i guess so, ie the ultimate fallback is what's actually in the service catalogue (i didn't get why that can't be used always) | 20:35 |
| *** fzdarsky has quit IRC | 20:36 | |
| Anticimex | ah, a specific service version wasn't in SC | 20:36 |
| Anticimex | isn't that possibly the problem? :] | 20:36 |
| briancurtin | Anticimex: there are some cases where *only* it can be used (object store), and some cases where it can’t be used at all (identity v3). in the middle of that, some services provide both! | 20:36 |
| Anticimex | ok | 20:36 |
| briancurtin | Anticimex: that one is on purpose, as they’re moving toward the version-less SC | 20:37 |
| Anticimex | oh, ack | 20:37 |
| Anticimex | briancurtin: i think you're on the right track anyway | 20:37 |
| Anticimex | generic first, special case later | 20:38 |
| briancurtin | Anticimex: I’m going to rework that method to try a few cases with the SC fallback. do you have an account on gerrit so i can add you to the review? | 20:38 |
| *** sigmavirus is now known as sigmavirus|awa | 20:38 | |
| Anticimex | yes, uh, not sure what my identifier is. email: martin@millnert.se | 20:39 |
| briancurtin | ok cool | 20:39 |
| openstackgerrit | Merged openstack/python-openstackclient: Display hypervisor information without uptime https://review.openstack.org/353846 | 20:43 |
| *** d0ugal has joined #openstack-sdks | 20:44 | |
| *** fangxu has quit IRC | 20:47 | |
| *** fangxu has joined #openstack-sdks | 20:51 | |
| *** sdague has quit IRC | 21:02 | |
| *** rtheis_ has joined #openstack-sdks | 21:03 | |
| *** rtheis_ has quit IRC | 21:03 | |
| *** rtheis has quit IRC | 21:05 | |
| Anticimex | now the openstacksdk user question becomes, how to query for all projects servers, not just a specific tenant's, using the "conn.compute.servers()" iterator | 21:13 |
| briancurtin | Anticimex: can you do that with anything else? (i have no idea, though maybe with admin level creds?) | 21:26 |
| Anticimex | yes, openstack server list --all-projects | 21:28 |
| briancurtin | if you can create a URL at which a request can be made and a response can be returned, we can make it work. we just haven’t done it all, through a combination of manpower and lack of documentation | 21:28 |
| Anticimex | by debug/timing output, it seems the secret(?) parameter ?all_tenants=True is appended | 21:28 |
| Anticimex | i was just this very second checking the compute api spec though | 21:28 |
| Anticimex | and don't see it :) | 21:28 |
| briancurtin | then we probably just need to add all_tenants as a query param and allow you to then call compute.servers(all_tenants=True) | 21:29 |
| Anticimex | http://developer.openstack.org/api-ref-compute-v2.1.html#listServers doesn't list it | 21:29 |
| Anticimex | briancurtin: yep | 21:29 |
| Anticimex | perhaps these operator things are hidden? a bit weird tho | 21:29 |
| Anticimex | (or someone removed that since liberty - we run liberty) | 21:29 |
| briancurtin | Anticimex: if “all_tenants” was added to this https://github.com/openstack/python-openstacksdk/blob/master/openstack/compute/v2/server.py#L32 you would be able to get what you want | 21:30 |
| Anticimex | oh, that's it? | 21:30 |
| Anticimex | such abstraction | 21:30 |
| Anticimex | i'll try :] | 21:30 |
| Anticimex | worked :) | 21:31 |
| Anticimex | ok let me contribute by adding that :] | 21:32 |
| Anticimex | do you have tests for these? | 21:32 |
| briancurtin | Anticimex: yep, for that you should just update https://github.com/openstack/python-openstacksdk/blob/master/openstack/tests/unit/compute/v2/test_server.py#L80 and that should be it | 21:33 |
| Anticimex | ah, i was inside functional tests folder looking | 21:34 |
| Anticimex | ok | 21:34 |
| Anticimex | thx | 21:34 |
| Anticimex | briancurtin: does the project use launchpad or.. something else? | 21:34 |
| briancurtin | we might have functional tests that would apply, but i don’t think so. we’re kind of light on functional tests | 21:34 |
| Anticimex | i mean, the parameter is quite simple data, so unit's probably fine | 21:35 |
| briancurtin | Anticimex: we do use launchpad, sort of. https://bugs.launchpad.net/python-openstacksdk is the bug tracker, but if you have a code fix we just usually go straight into gerrit with it | 21:35 |
| Anticimex | ok | 21:35 |
| briancurtin | before you said “let me contribute” i was about to say feel free to file a bug report there, but code in hand works quite well | 21:36 |
| *** mfedosin has quit IRC | 21:44 | |
| Anticimex | briancurtin: openstack workflow i'm used to is basically: git checkout -B bug/foo ; make_changes ; git review | 21:44 |
| Anticimex | (it's been 6 months) | 21:45 |
| Anticimex | gerrit now told me "FU, i see no change-id" | 21:45 |
| * Anticimex goes to openstack/git review docs :s | 21:45 | |
| briancurtin | Anticimex: that’s what i do. in your “make_changes” step, do you have just one commit to represent both the code and test changes? i think if you make multiple commits it might complain about that | 21:47 |
| Anticimex | nod, just one commit | 21:47 |
| *** fzdarsky has joined #openstack-sdks | 21:48 | |
| Anticimex | do you setup gerrit with commit hook? perhaps i missed a flag in "git review -s" to set it up :s | 21:48 |
| briancurtin | Anticimex: im not really sure then. maybe something wrong with git review? i think you need to do a git review -s? | 21:48 |
| briancurtin | nah we don’t have any hooks going on | 21:48 |
| Anticimex | yeah i'll google it, not your fault :) | 21:48 |
| briancurtin | i just did a git commit —amend on my my last change and it throws the Change-Id automatically at the end of the message | 21:49 |
| Anticimex | that may be what i missed | 21:49 |
| Anticimex | hm nope. found the relevant docs: https://review.openstack.org/Documentation/user-changeid.html#creation | 21:50 |
| Anticimex | out of curiosity, do you have a $repo/.git/hooks/commit-msg there? | 21:50 |
| briancurtin | i do | 21:51 |
| Anticimex | me too actually. perhaps it didn't like my branch name. *reading commit-msg' | 21:51 |
| Anticimex | ok now it appeared on git commit --amend. this was really odd. oh well | 21:53 |
| openstackgerrit | Martin Millnert proposed openstack/python-openstacksdk: compute/v2/server: add ?all_tenants=bool to list https://review.openstack.org/369057 | 21:53 |
| *** annegentle has quit IRC | 21:54 | |
| briancurtin | Anticimex: cool, looks good now. since this is trivial and i can see it’s fine, sending it through now (though it’ll still run the jenkins tests and wait to pass for anything to move forward) | 21:56 |
| Anticimex | ack | 21:56 |
| *** craigs has joined #openstack-sdks | 21:57 | |
| *** amotoki has quit IRC | 21:58 | |
| *** zigo has quit IRC | 22:08 | |
| *** craigs has quit IRC | 22:09 | |
| *** csterret_ has joined #openstack-sdks | 22:09 | |
| *** zigo has joined #openstack-sdks | 22:11 | |
| *** zigo is now known as Guest41388 | 22:11 | |
| openstackgerrit | Merged openstack/python-openstacksdk: compute/v2/server: add ?all_tenants=bool to list https://review.openstack.org/369057 | 22:16 |
| *** Guest41388 is now known as zigo_ | 22:17 | |
| *** adriant has joined #openstack-sdks | 22:17 | |
| *** openstackgerrit has quit IRC | 22:48 | |
| *** openstackgerrit has joined #openstack-sdks | 22:49 | |
| *** salv-orlando has joined #openstack-sdks | 22:53 | |
| *** salv-orl_ has quit IRC | 22:54 | |
| *** sdague has joined #openstack-sdks | 22:56 | |
| *** annegentle has joined #openstack-sdks | 22:59 | |
| *** annegentle has quit IRC | 23:00 | |
| *** sdague has quit IRC | 23:04 | |
| *** fzdarsky has quit IRC | 23:14 | |
| *** sdague has joined #openstack-sdks | 23:15 | |
| *** annegentle has joined #openstack-sdks | 23:43 | |
| *** zhurong has joined #openstack-sdks | 23:48 | |
| *** annegentle has quit IRC | 23:48 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!