| openstackgerrit | Monty Taylor proposed openstack-dev/pbr: Write and read more complex git sha info https://review.openstack.org/141666 | 00:02 |
|---|---|---|
| mordred | clarkb: there is the patch ported on to master | 00:03 |
| fungi | mordred: would that also be useful returned as an attribute of the pbr.version.VersionInfo object? | 00:05 |
| fungi | perhaps there's a better option for software which wants to be able to return the git sha along with the human-readable version somewhere which doesn't involve a runtime dependency on pbr maybe | 00:07 |
| fungi | i guess the cli tool provides a good example of how to go about that | 00:09 |
| *** ZZelle_ has quit IRC | 00:11 | |
| *** sabari is now known as zz_sabari | 00:20 | |
| *** ivar-lazzaro has quit IRC | 00:22 | |
| *** ivar-lazzaro has joined #openstack-infra | 00:23 | |
| *** zz_avozza is now known as avozza | 00:23 | |
| *** fandi has joined #openstack-infra | 00:31 | |
| *** zz_sabari is now known as sabari | 00:32 | |
| *** avozza is now known as zz_avozza | 00:33 | |
| *** dpaterson has quit IRC | 00:34 | |
| *** Masahiro has joined #openstack-infra | 00:37 | |
| mordred | fungi: I do not understand the chaos that a runtime dep on pbr causes in some people's brains | 00:37 |
| mordred | fungi: it's a python library | 00:38 |
| mordred | openstack already has >200 python libraries | 00:38 |
| *** otter768 has joined #openstack-infra | 00:38 | |
| *** Masahiro has quit IRC | 00:42 | |
| *** ryanpetrello has joined #openstack-infra | 00:43 | |
| *** dimsum__ has joined #openstack-infra | 00:46 | |
| *** oomichi_ has joined #openstack-infra | 00:46 | |
| *** penguinRaider has quit IRC | 00:47 | |
| *** teran has quit IRC | 00:53 | |
| *** unicell has joined #openstack-infra | 00:54 | |
| *** tylerdurden has quit IRC | 00:54 | |
| *** shakamunyi has joined #openstack-infra | 00:55 | |
| *** Masahiro has joined #openstack-infra | 00:55 | |
| *** ivar-lazzaro has quit IRC | 00:57 | |
| *** sabari is now known as zz_sabari | 00:58 | |
| *** ryanpetrello has quit IRC | 00:59 | |
| *** ryanpetrello has joined #openstack-infra | 01:01 | |
| *** zz_sabari is now known as sabari | 01:08 | |
| *** yaguang has joined #openstack-infra | 01:08 | |
| *** ryanpetrello has quit IRC | 01:12 | |
| *** boris-42 has quit IRC | 01:13 | |
| *** wuhg has joined #openstack-infra | 01:19 | |
| *** yaguang has quit IRC | 01:20 | |
| *** yaguang has joined #openstack-infra | 01:20 | |
| *** stevemar has joined #openstack-infra | 01:20 | |
| *** mriedem has joined #openstack-infra | 01:21 | |
| *** ddieterly has joined #openstack-infra | 01:24 | |
| *** zz_avozza is now known as avozza | 01:27 | |
| *** jaypipes has joined #openstack-infra | 01:28 | |
| *** mriedem has quit IRC | 01:31 | |
| *** zhiwei has joined #openstack-infra | 01:34 | |
| *** otter768 has quit IRC | 01:35 | |
| *** liusheng has quit IRC | 01:37 | |
| *** liusheng has joined #openstack-infra | 01:37 | |
| *** hdd has joined #openstack-infra | 01:41 | |
| *** ivar-lazzaro has joined #openstack-infra | 01:42 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Refactor sources out of triggers https://review.openstack.org/118993 | 01:42 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Fix pep8 issues https://review.openstack.org/115187 | 01:42 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add gerrit reviews into patchset approvals https://review.openstack.org/97390 | 01:42 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add support for negative requirements https://review.openstack.org/102726 | 01:42 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Configure triggers dynamically https://review.openstack.org/119534 | 01:43 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add support for 'connection' concept https://review.openstack.org/121528 | 01:43 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Call driver methods more dynamically https://review.openstack.org/119533 | 01:43 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add base class for triggers https://review.openstack.org/119532 | 01:43 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add base class for sources https://review.openstack.org/119531 | 01:43 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add base class for reporters https://review.openstack.org/119530 | 01:43 |
| *** ivar-lazzaro has quit IRC | 01:44 | |
| *** yaguang has quit IRC | 01:51 | |
| *** hdd has quit IRC | 01:57 | |
| *** talluri has joined #openstack-infra | 02:00 | |
| *** talluri has quit IRC | 02:05 | |
| *** pcrews has joined #openstack-infra | 02:18 | |
| *** fandi has quit IRC | 02:20 | |
| *** fandi has joined #openstack-infra | 02:22 | |
| *** pcrews has quit IRC | 02:27 | |
| *** yaguang has joined #openstack-infra | 02:29 | |
| *** talluri has joined #openstack-infra | 02:47 | |
| *** KanagarajM has joined #openstack-infra | 02:49 | |
| *** ddieterly has quit IRC | 02:52 | |
| *** jerryz has joined #openstack-infra | 02:54 | |
| *** zz_dimtruck is now known as dimtruck | 03:10 | |
| *** dpaterson has joined #openstack-infra | 03:19 | |
| *** koolhead17 has joined #openstack-infra | 03:23 | |
| *** ddieterly has joined #openstack-infra | 03:25 | |
| *** ddieterly has quit IRC | 03:27 | |
| *** ddieterl_ has joined #openstack-infra | 03:27 | |
| *** Masahiro has quit IRC | 03:27 | |
| *** baoli has quit IRC | 03:29 | |
| *** baoli has joined #openstack-infra | 03:29 | |
| *** dkranz has joined #openstack-infra | 03:30 | |
| *** ddieterl_ has quit IRC | 03:31 | |
| *** otter768 has joined #openstack-infra | 03:36 | |
| *** teran has joined #openstack-infra | 03:38 | |
| *** otter768 has quit IRC | 03:40 | |
| *** teran has quit IRC | 03:42 | |
| *** teran has joined #openstack-infra | 03:45 | |
| *** dpaterson has quit IRC | 03:48 | |
| *** oomichi_ has quit IRC | 03:50 | |
| *** mestery has quit IRC | 03:52 | |
| *** ayoung has quit IRC | 03:52 | |
| *** amitgandhinz has joined #openstack-infra | 03:54 | |
| *** shashankhegde has joined #openstack-infra | 03:55 | |
| *** cody-somerville has joined #openstack-infra | 03:56 | |
| *** mestery has joined #openstack-infra | 04:00 | |
| *** yamamoto_ has joined #openstack-infra | 04:04 | |
| *** Masahiro has joined #openstack-infra | 04:04 | |
| *** yfried|afk is now known as yfried | 04:22 | |
| *** amitgandhinz has quit IRC | 04:23 | |
| *** nuritv has quit IRC | 04:24 | |
| *** dimsum__ has quit IRC | 04:25 | |
| *** baoli has quit IRC | 04:26 | |
| *** ddieterly has joined #openstack-infra | 04:26 | |
| *** talluri has quit IRC | 04:28 | |
| *** shashankhegde has quit IRC | 04:30 | |
| *** kingia has joined #openstack-infra | 04:30 | |
| *** kingia is now known as dhp | 04:30 | |
| *** ddieterly has quit IRC | 04:31 | |
| *** nuritv has joined #openstack-infra | 04:32 | |
| *** talluri has joined #openstack-infra | 04:36 | |
| *** talluri has quit IRC | 04:37 | |
| *** talluri has joined #openstack-infra | 04:37 | |
| *** talluri has quit IRC | 04:38 | |
| *** talluri has joined #openstack-infra | 04:39 | |
| *** talluri has quit IRC | 04:40 | |
| *** talluri has joined #openstack-infra | 04:40 | |
| *** otter768 has joined #openstack-infra | 04:41 | |
| *** talluri has quit IRC | 04:44 | |
| *** achanda has joined #openstack-infra | 04:47 | |
| *** otter768 has quit IRC | 04:51 | |
| *** talluri has joined #openstack-infra | 04:52 | |
| *** mrmartin has joined #openstack-infra | 04:53 | |
| *** liusheng has quit IRC | 04:54 | |
| *** liusheng has joined #openstack-infra | 04:54 | |
| *** miqui has quit IRC | 04:56 | |
| *** koolhead17 has quit IRC | 04:58 | |
| *** koolhead17 has joined #openstack-infra | 04:58 | |
| *** koolhead17 has joined #openstack-infra | 04:58 | |
| *** boris-42 has joined #openstack-infra | 05:01 | |
| *** dhp_ has joined #openstack-infra | 05:01 | |
| *** dhp_ is now known as iain-king | 05:01 | |
| *** dhp has quit IRC | 05:01 | |
| *** iain-king is now known as dhp | 05:02 | |
| *** avozza is now known as zz_avozza | 05:03 | |
| *** ddieterly has joined #openstack-infra | 05:08 | |
| *** achanda has quit IRC | 05:08 | |
| *** koolhead_ has joined #openstack-infra | 05:08 | |
| *** achanda has joined #openstack-infra | 05:09 | |
| *** hdd has joined #openstack-infra | 05:10 | |
| *** achanda has quit IRC | 05:11 | |
| *** koolhead17 has quit IRC | 05:11 | |
| *** achanda has joined #openstack-infra | 05:12 | |
| *** ddieterly has quit IRC | 05:13 | |
| *** vigneshvar has joined #openstack-infra | 05:15 | |
| *** stevemar has quit IRC | 05:16 | |
| *** achanda has quit IRC | 05:16 | |
| *** achanda has joined #openstack-infra | 05:16 | |
| *** nuritv has quit IRC | 05:17 | |
| *** dimtruck is now known as zz_dimtruck | 05:18 | |
| *** achanda has quit IRC | 05:18 | |
| *** nuritv has joined #openstack-infra | 05:19 | |
| *** ddieterly has joined #openstack-infra | 05:26 | |
| *** nuritv has quit IRC | 05:28 | |
| *** nikil has joined #openstack-infra | 05:30 | |
| *** ddieterly has quit IRC | 05:31 | |
| *** jyuso has quit IRC | 05:33 | |
| *** jyuso has joined #openstack-infra | 05:33 | |
| *** yaguang has quit IRC | 05:34 | |
| nikil | hi i see that the project/branch/ref are taking from zuul and it written into env $workspace/test-evn.sh. Is there any way in devstack-gate to use specify branch/URL for only one particular project like neutron. That is it should fetch all projects from standard repo and only the neutron should fetch from our inernal repo for the CI | 05:36 |
| *** nuritv has joined #openstack-infra | 05:37 | |
| *** nuritv has quit IRC | 05:38 | |
| *** yaguang has joined #openstack-infra | 05:43 | |
| *** yfried has quit IRC | 05:43 | |
| *** nuritv has joined #openstack-infra | 05:47 | |
| *** nuritv has quit IRC | 05:48 | |
| *** koolhead_ has quit IRC | 05:49 | |
| *** dannywilson has joined #openstack-infra | 05:49 | |
| *** nuritv has joined #openstack-infra | 05:51 | |
| *** nuritv has quit IRC | 05:52 | |
| *** penguinRaider has joined #openstack-infra | 05:53 | |
| *** nuritv has joined #openstack-infra | 05:55 | |
| *** hdd has quit IRC | 05:56 | |
| *** nuritv has quit IRC | 05:57 | |
| *** KanagarajM has quit IRC | 06:02 | |
| *** nuritv has joined #openstack-infra | 06:02 | |
| *** nuritv has quit IRC | 06:04 | |
| *** nuritv has joined #openstack-infra | 06:05 | |
| *** nuritv has quit IRC | 06:07 | |
| *** nuritv has joined #openstack-infra | 06:12 | |
| *** nuritv has quit IRC | 06:13 | |
| *** nuritv has joined #openstack-infra | 06:16 | |
| *** nuritv has quit IRC | 06:18 | |
| *** nuritv has joined #openstack-infra | 06:20 | |
| *** nuritv has quit IRC | 06:21 | |
| *** ddieterly has joined #openstack-infra | 06:26 | |
| *** Masahiro_ has joined #openstack-infra | 06:26 | |
| *** simonmcc has quit IRC | 06:26 | |
| *** jraim has quit IRC | 06:26 | |
| *** simonmcc_ has joined #openstack-infra | 06:26 | |
| *** jraim_ has joined #openstack-infra | 06:26 | |
| *** nuritv has joined #openstack-infra | 06:27 | |
| *** Masahiro has quit IRC | 06:28 | |
| *** jamespage has quit IRC | 06:28 | |
| *** nuritv has quit IRC | 06:29 | |
| *** ddieterly has quit IRC | 06:31 | |
| *** nuritv has joined #openstack-infra | 06:35 | |
| *** nuritv has quit IRC | 06:37 | |
| openstackgerrit | Colleen Murphy proposed openstack-infra/project-config: Run acceptance tests without rake https://review.openstack.org/141728 | 06:38 |
| *** nuritv has joined #openstack-infra | 06:40 | |
| crinkle | nibalizer: ^ | 06:41 |
| *** nuritv has quit IRC | 06:42 | |
| *** imcsk8 has quit IRC | 06:42 | |
| *** imcsk8 has joined #openstack-infra | 06:43 | |
| *** nuritv has joined #openstack-infra | 06:44 | |
| *** nuritv has quit IRC | 06:45 | |
| * nibalizer looks | 06:46 | |
| *** nuritv has joined #openstack-infra | 06:48 | |
| *** nuritv has quit IRC | 06:49 | |
| *** koolhead17 has joined #openstack-infra | 06:51 | |
| *** otter768 has joined #openstack-infra | 06:52 | |
| *** ddieterly has joined #openstack-infra | 06:53 | |
| *** nuritv has joined #openstack-infra | 06:54 | |
| *** penguinRaider has quit IRC | 06:55 | |
| *** Longgeek has joined #openstack-infra | 06:56 | |
| *** mrunge has joined #openstack-infra | 06:56 | |
| *** nuritv has quit IRC | 06:56 | |
| nikil | hi can some one explain me when pre_test_hook, post_test_hook and gate_hook function executes ? | 06:57 |
| *** otter768 has quit IRC | 06:57 | |
| *** ddieterly has quit IRC | 06:57 | |
| *** nuritv has joined #openstack-infra | 06:58 | |
| *** nuritv has quit IRC | 07:00 | |
| *** vigneshvar has quit IRC | 07:01 | |
| *** yfried has joined #openstack-infra | 07:04 | |
| zhiyan | hi folks, i'm interested in what's the cause of this kind of wired issue http://logs.openstack.org/22/141722/1/check//gate-glance-python27/feece60/console.html#_2014-12-15_06_12_51_737 ... the version rang is not wrong but .. | 07:04 |
| *** nuritv has joined #openstack-infra | 07:05 | |
| *** nuritv has quit IRC | 07:06 | |
| *** nuritv has joined #openstack-infra | 07:13 | |
| *** nuritv has quit IRC | 07:15 | |
| *** teran has quit IRC | 07:15 | |
| *** nuritv has joined #openstack-infra | 07:18 | |
| *** vigneshvar has joined #openstack-infra | 07:20 | |
| *** nuritv has quit IRC | 07:21 | |
| wuhg | hi ,all, how can i get build history of https://jenkins.openstack.org/view/All/job/gate-horizon-dsvm-integration/ | 07:21 |
| *** penguinRaider has joined #openstack-infra | 07:23 | |
| *** dannywilson has quit IRC | 07:23 | |
| *** nuritv has joined #openstack-infra | 07:24 | |
| *** nuritv has quit IRC | 07:26 | |
| *** Longgeek has quit IRC | 07:26 | |
| *** skraynev has joined #openstack-infra | 07:26 | |
| *** nuritv has joined #openstack-infra | 07:27 | |
| *** _nadya_ has joined #openstack-infra | 07:27 | |
| *** nuritv has quit IRC | 07:28 | |
| *** nuritv has joined #openstack-infra | 07:34 | |
| *** Longgeek has joined #openstack-infra | 07:34 | |
| *** nuritv has quit IRC | 07:35 | |
| *** skraynev has quit IRC | 07:36 | |
| *** zhiwei has left #openstack-infra | 07:37 | |
| *** skraynev has joined #openstack-infra | 07:38 | |
| *** _nadya_ has quit IRC | 07:38 | |
| *** skraynev_ has joined #openstack-infra | 07:38 | |
| *** skraynev_ has quit IRC | 07:38 | |
| *** unicell has quit IRC | 07:39 | |
| *** nuritv has joined #openstack-infra | 07:39 | |
| *** nuritv has quit IRC | 07:41 | |
| *** Longgeek has quit IRC | 07:41 | |
| *** nuritv has joined #openstack-infra | 07:42 | |
| *** Longgeek has joined #openstack-infra | 07:42 | |
| *** jgallard_ has joined #openstack-infra | 07:43 | |
| *** nuritv has quit IRC | 07:44 | |
| *** unicell has joined #openstack-infra | 07:45 | |
| *** nuritv has joined #openstack-infra | 07:48 | |
| *** ala_ has joined #openstack-infra | 07:49 | |
| *** nuritv has quit IRC | 07:50 | |
| *** Longgeek has quit IRC | 07:50 | |
| *** Longgeek has joined #openstack-infra | 07:52 | |
| *** nuritv has joined #openstack-infra | 07:52 | |
| *** teran has joined #openstack-infra | 07:53 | |
| *** nuritv has quit IRC | 07:54 | |
| *** jamespage has joined #openstack-infra | 07:55 | |
| *** jamespage has joined #openstack-infra | 07:55 | |
| *** nuritv has joined #openstack-infra | 07:55 | |
| *** nuritv has quit IRC | 07:57 | |
| *** dhp has quit IRC | 08:00 | |
| *** nuritv has joined #openstack-infra | 08:01 | |
| *** nuritv has quit IRC | 08:03 | |
| *** jpich has joined #openstack-infra | 08:04 | |
| *** HeOS has quit IRC | 08:05 | |
| *** nuritv has joined #openstack-infra | 08:07 | |
| *** dhp has joined #openstack-infra | 08:07 | |
| *** willroberts_ has quit IRC | 08:07 | |
| *** nuritv has quit IRC | 08:08 | |
| *** willroberts_ has joined #openstack-infra | 08:09 | |
| *** nuritv has joined #openstack-infra | 08:11 | |
| *** jpich has quit IRC | 08:11 | |
| *** che-arne has joined #openstack-infra | 08:11 | |
| *** che-arn4 has joined #openstack-infra | 08:11 | |
| *** nuritv has quit IRC | 08:12 | |
| *** che-arn4 has quit IRC | 08:12 | |
| *** jpich has joined #openstack-infra | 08:13 | |
| *** nuritv has joined #openstack-infra | 08:13 | |
| *** mrunge has quit IRC | 08:14 | |
| *** dhp has quit IRC | 08:15 | |
| *** nuritv has quit IRC | 08:15 | |
| *** markus_z has joined #openstack-infra | 08:15 | |
| *** salv-orlando has joined #openstack-infra | 08:16 | |
| *** markus_z has quit IRC | 08:16 | |
| *** sabari is now known as zz_sabari | 08:17 | |
| *** nuritv has joined #openstack-infra | 08:19 | |
| *** nuritv has quit IRC | 08:20 | |
| *** miqui_ has quit IRC | 08:22 | |
| *** ivar-lazzaro has joined #openstack-infra | 08:22 | |
| *** nuritv has joined #openstack-infra | 08:22 | |
| *** zz_sabari is now known as sabari | 08:23 | |
| *** kashyap` is now known as kashyap | 08:24 | |
| *** nuritv has quit IRC | 08:24 | |
| *** k4n0 has joined #openstack-infra | 08:25 | |
| *** doude has joined #openstack-infra | 08:26 | |
| *** rlandy has joined #openstack-infra | 08:26 | |
| *** mpaolino has joined #openstack-infra | 08:26 | |
| *** ivar-lazzaro has quit IRC | 08:27 | |
| *** nuritv has joined #openstack-infra | 08:27 | |
| *** vigneshvar_ has joined #openstack-infra | 08:29 | |
| *** achuprin_ has quit IRC | 08:29 | |
| *** nuritv has quit IRC | 08:29 | |
| *** jcoufal has joined #openstack-infra | 08:31 | |
| *** nuritv has joined #openstack-infra | 08:32 | |
| *** SumitNaiksatam has quit IRC | 08:32 | |
| *** vigneshvar has quit IRC | 08:32 | |
| *** SumitNaiksatam has joined #openstack-infra | 08:33 | |
| *** nuritv has quit IRC | 08:34 | |
| *** mpaolino has quit IRC | 08:35 | |
| *** nuritv has joined #openstack-infra | 08:37 | |
| *** mpaolino has joined #openstack-infra | 08:38 | |
| *** nuritv has quit IRC | 08:39 | |
| *** jlibosva has joined #openstack-infra | 08:40 | |
| *** nuritv has joined #openstack-infra | 08:41 | |
| *** achuprin_ has joined #openstack-infra | 08:42 | |
| *** nuritv has quit IRC | 08:43 | |
| *** zz_avozza is now known as avozza | 08:44 | |
| *** skraynev has left #openstack-infra | 08:45 | |
| *** nuritv has joined #openstack-infra | 08:47 | |
| *** salv-orlando has quit IRC | 08:48 | |
| *** nuritv has quit IRC | 08:49 | |
| *** luqas has quit IRC | 08:50 | |
| *** nuritv has joined #openstack-infra | 08:53 | |
| *** otter768 has joined #openstack-infra | 08:53 | |
| *** pblaho has joined #openstack-infra | 08:53 | |
| *** nuritv has quit IRC | 08:55 | |
| *** luqas has joined #openstack-infra | 08:56 | |
| *** sabari is now known as zz_sabari | 08:57 | |
| *** otter768 has quit IRC | 08:58 | |
| *** nuritv has joined #openstack-infra | 08:58 | |
| *** jlibosva has quit IRC | 08:58 | |
| *** nuritv has quit IRC | 09:00 | |
| *** jlibosva has joined #openstack-infra | 09:00 | |
| *** subscope has joined #openstack-infra | 09:00 | |
| *** e0ne has joined #openstack-infra | 09:01 | |
| *** nuritv has joined #openstack-infra | 09:02 | |
| *** nuritv has quit IRC | 09:03 | |
| *** e0ne has quit IRC | 09:07 | |
| *** skraynev has joined #openstack-infra | 09:07 | |
| *** skraynev has left #openstack-infra | 09:08 | |
| *** nuritv has joined #openstack-infra | 09:09 | |
| *** viktors|afk is now known as viktors | 09:10 | |
| *** BobBall_AWOL is now known as BobBall | 09:11 | |
| *** nuritv has quit IRC | 09:12 | |
| *** _nadya_ has joined #openstack-infra | 09:12 | |
| *** jlibosva has quit IRC | 09:12 | |
| *** andreykurilin has joined #openstack-infra | 09:13 | |
| *** jlibosva has joined #openstack-infra | 09:13 | |
| *** nuritv has joined #openstack-infra | 09:13 | |
| *** nuritv has quit IRC | 09:15 | |
| *** hashar has joined #openstack-infra | 09:15 | |
| *** derekh has joined #openstack-infra | 09:17 | |
| *** Guest60167 has joined #openstack-infra | 09:19 | |
| *** jedimike has joined #openstack-infra | 09:19 | |
| *** nuritv has joined #openstack-infra | 09:19 | |
| *** AJaeger has joined #openstack-infra | 09:20 | |
| *** AJaeger has quit IRC | 09:20 | |
| *** AJaeger has joined #openstack-infra | 09:20 | |
| *** MaxV has joined #openstack-infra | 09:20 | |
| *** nuritv has quit IRC | 09:21 | |
| *** nuritv has joined #openstack-infra | 09:23 | |
| *** nuritv has quit IRC | 09:25 | |
| *** HeOS has joined #openstack-infra | 09:26 | |
| *** nuritv has joined #openstack-infra | 09:27 | |
| *** doude has quit IRC | 09:28 | |
| *** doude has joined #openstack-infra | 09:29 | |
| *** dimsum__ has joined #openstack-infra | 09:29 | |
| *** nuritv has quit IRC | 09:29 | |
| *** ildikov has quit IRC | 09:30 | |
| *** jp_at_hp has joined #openstack-infra | 09:31 | |
| *** arxcruz has joined #openstack-infra | 09:32 | |
| *** yamamoto_ has quit IRC | 09:32 | |
| *** nuritv has joined #openstack-infra | 09:33 | |
| *** koolhead17 has quit IRC | 09:34 | |
| *** markvan has quit IRC | 09:34 | |
| *** dimsum__ has quit IRC | 09:34 | |
| *** markvan has joined #openstack-infra | 09:35 | |
| *** nuritv has quit IRC | 09:35 | |
| *** _nadya_ has quit IRC | 09:35 | |
| *** nuritv has joined #openstack-infra | 09:37 | |
| *** nuritv has quit IRC | 09:39 | |
| *** nuritv has joined #openstack-infra | 09:42 | |
| *** ildikov has joined #openstack-infra | 09:43 | |
| *** nuritv has quit IRC | 09:44 | |
| *** ildikov has quit IRC | 09:44 | |
| *** pblaho has quit IRC | 09:44 | |
| openstackgerrit | yolanda.robla proposed openstack-infra/infra-specs: Add nodepool disk image builder testing spec. https://review.openstack.org/139598 | 09:47 |
| *** nuritv has joined #openstack-infra | 09:49 | |
| *** pelix has joined #openstack-infra | 09:50 | |
| *** e0ne has joined #openstack-infra | 09:51 | |
| *** nuritv has quit IRC | 09:52 | |
| *** yfried has quit IRC | 09:53 | |
| *** yfried has joined #openstack-infra | 09:53 | |
| *** yaguang has quit IRC | 09:54 | |
| *** nuritv has joined #openstack-infra | 09:54 | |
| *** andreykurilin has quit IRC | 09:55 | |
| *** nuritv has quit IRC | 09:56 | |
| *** Guest60167 has quit IRC | 09:57 | |
| *** nuritv has joined #openstack-infra | 09:58 | |
| *** yamamoto has joined #openstack-infra | 09:58 | |
| *** nuritv has quit IRC | 09:59 | |
| openstackgerrit | yolanda.robla proposed openstack-infra/infra-specs: Add nodepool image testing spec. https://review.openstack.org/139598 | 10:00 |
| *** yolanda has joined #openstack-infra | 10:00 | |
| *** nuritv has joined #openstack-infra | 10:03 | |
| *** ildikov has joined #openstack-infra | 10:03 | |
| *** yamamoto has quit IRC | 10:04 | |
| *** nuritv has quit IRC | 10:05 | |
| *** belmoreira has joined #openstack-infra | 10:06 | |
| *** che-arne has quit IRC | 10:06 | |
| *** nuritv has joined #openstack-infra | 10:07 | |
| *** andreaf has joined #openstack-infra | 10:07 | |
| *** asettle has quit IRC | 10:08 | |
| *** Guest60167 has joined #openstack-infra | 10:09 | |
| *** nuritv has quit IRC | 10:09 | |
| *** nuritv has joined #openstack-infra | 10:15 | |
| *** avozza is now known as zz_avozza | 10:15 | |
| *** zz_avozza is now known as avozza | 10:16 | |
| *** nuritv has quit IRC | 10:16 | |
| *** avozza is now known as zz_avozza | 10:17 | |
| *** nuritv has joined #openstack-infra | 10:18 | |
| *** nuritv has quit IRC | 10:19 | |
| *** ihrachyshka has joined #openstack-infra | 10:20 | |
| *** nuritv has joined #openstack-infra | 10:22 | |
| *** nuritv has quit IRC | 10:24 | |
| *** mrunge has joined #openstack-infra | 10:24 | |
| *** asettle has joined #openstack-infra | 10:25 | |
| *** nuritv has joined #openstack-infra | 10:27 | |
| *** nuritv has quit IRC | 10:29 | |
| *** nuritv has joined #openstack-infra | 10:31 | |
| *** skolekonov has joined #openstack-infra | 10:32 | |
| *** skolekonov has quit IRC | 10:33 | |
| *** nuritv has quit IRC | 10:33 | |
| *** salv-orlando has joined #openstack-infra | 10:33 | |
| *** salv-orl_ has joined #openstack-infra | 10:34 | |
| *** salv-orlando has quit IRC | 10:34 | |
| *** yfried is now known as yfried|afk | 10:34 | |
| *** nuritv has joined #openstack-infra | 10:35 | |
| *** nuritv has quit IRC | 10:37 | |
| *** gema has quit IRC | 10:37 | |
| *** belmoreira has quit IRC | 10:39 | |
| *** skolekonov has joined #openstack-infra | 10:40 | |
| *** nuritv has joined #openstack-infra | 10:40 | |
| *** nuritv has quit IRC | 10:41 | |
| *** jp_at_hp has quit IRC | 10:43 | |
| *** nuritv has joined #openstack-infra | 10:45 | |
| *** Adri2000 has quit IRC | 10:45 | |
| *** nuritv has quit IRC | 10:46 | |
| *** jp_at_hp has joined #openstack-infra | 10:48 | |
| *** nuritv has joined #openstack-infra | 10:49 | |
| *** Adri2000 has joined #openstack-infra | 10:49 | |
| *** nuritv has quit IRC | 10:51 | |
| openstackgerrit | Mike Heald proposed openstack-infra/infra-specs: Result set management and paging specification https://review.openstack.org/139638 | 10:53 |
| *** nuritv has joined #openstack-infra | 10:54 | |
| *** unicell has quit IRC | 10:54 | |
| *** yfried|afk is now known as yfried | 10:54 | |
| *** otter768 has joined #openstack-infra | 10:54 | |
| *** nuritv has quit IRC | 10:56 | |
| *** liusheng has quit IRC | 10:56 | |
| *** liusheng has joined #openstack-infra | 10:57 | |
| *** nuritv has joined #openstack-infra | 10:58 | |
| *** ZZelle has quit IRC | 10:58 | |
| *** ZZelle has joined #openstack-infra | 10:58 | |
| *** jgallard_ has quit IRC | 10:58 | |
| *** otter768 has quit IRC | 10:59 | |
| *** nuritv has quit IRC | 11:00 | |
| *** nuritv has joined #openstack-infra | 11:01 | |
| *** aysyd has joined #openstack-infra | 11:02 | |
| *** nuritv has quit IRC | 11:03 | |
| openstackgerrit | Thierry Carrez proposed openstack-infra/project-config: Introduce Oslo-specific stable-maint team https://review.openstack.org/141769 | 11:04 |
| *** yamamoto has joined #openstack-infra | 11:05 | |
| *** nuritv has joined #openstack-infra | 11:07 | |
| *** nuritv has quit IRC | 11:08 | |
| *** asettle has quit IRC | 11:08 | |
| *** yamamoto has quit IRC | 11:10 | |
| *** pblaho has joined #openstack-infra | 11:10 | |
| *** nuritv has joined #openstack-infra | 11:11 | |
| *** jp_at_hp has quit IRC | 11:11 | |
| *** nuritv has quit IRC | 11:13 | |
| *** amuller has joined #openstack-infra | 11:13 | |
| *** nuritv has joined #openstack-infra | 11:15 | |
| *** nuritv has quit IRC | 11:17 | |
| *** zz_avozza is now known as avozza | 11:17 | |
| *** Guest60167 has quit IRC | 11:17 | |
| *** Masahiro_ has quit IRC | 11:17 | |
| *** _nadya_ has joined #openstack-infra | 11:18 | |
| *** nuritv has joined #openstack-infra | 11:20 | |
| *** mfink__ has quit IRC | 11:20 | |
| *** asettle has joined #openstack-infra | 11:21 | |
| *** nuritv has quit IRC | 11:22 | |
| *** jp_at_hp has joined #openstack-infra | 11:23 | |
| *** nuritv has joined #openstack-infra | 11:25 | |
| *** avozza is now known as zz_avozza | 11:27 | |
| *** nuritv has quit IRC | 11:27 | |
| *** nuritv has joined #openstack-infra | 11:30 | |
| *** fandi has quit IRC | 11:30 | |
| *** nuritv has quit IRC | 11:31 | |
| *** pc_m has joined #openstack-infra | 11:33 | |
| *** nuritv has joined #openstack-infra | 11:34 | |
| *** mpaolino has quit IRC | 11:35 | |
| *** nuritv has quit IRC | 11:36 | |
| *** mpaolino has joined #openstack-infra | 11:38 | |
| *** nuritv has joined #openstack-infra | 11:39 | |
| *** nuritv has quit IRC | 11:40 | |
| openstackgerrit | yolanda.robla proposed openstack-infra/storyboard: Return task count by status for a story dynamically https://review.openstack.org/139100 | 11:43 |
| *** nuritv has joined #openstack-infra | 11:43 | |
| *** nuritv has quit IRC | 11:45 | |
| openstackgerrit | yolanda.robla proposed openstack-infra/storyboard: Return task count by status for a story dynamically https://review.openstack.org/139100 | 11:46 |
| *** nuritv has joined #openstack-infra | 11:47 | |
| *** zz_avozza is now known as avozza | 11:48 | |
| *** nuritv has quit IRC | 11:49 | |
| *** nuritv has joined #openstack-infra | 11:51 | |
| openstackgerrit | yolanda.robla proposed openstack-infra/storyboard: Return task count by status for a story dynamically https://review.openstack.org/139100 | 11:52 |
| *** nuritv has quit IRC | 11:52 | |
| *** nuritv has joined #openstack-infra | 11:53 | |
| rediskin | hiyo! this already have +2, could someone please give +A to this very simple patch https://review.openstack.org/#/c/140296/ | 11:54 |
| *** nuritv has quit IRC | 11:55 | |
| *** adalbas has joined #openstack-infra | 11:57 | |
| *** _nadya_ has quit IRC | 11:58 | |
| *** gema has joined #openstack-infra | 11:58 | |
| *** nuritv has joined #openstack-infra | 11:59 | |
| *** avozza is now known as zz_avozza | 12:00 | |
| *** nuritv has quit IRC | 12:01 | |
| *** nuritv has joined #openstack-infra | 12:04 | |
| *** nuritv has quit IRC | 12:05 | |
| *** arxcruz has quit IRC | 12:06 | |
| *** nuritv has joined #openstack-infra | 12:08 | |
| *** nuritv has quit IRC | 12:10 | |
| *** penguinRaider has quit IRC | 12:11 | |
| *** nuritv has joined #openstack-infra | 12:12 | |
| *** _nadya_ has joined #openstack-infra | 12:14 | |
| *** nuritv has quit IRC | 12:14 | |
| *** arxcruz has joined #openstack-infra | 12:17 | |
| *** nuritv has joined #openstack-infra | 12:17 | |
| *** Masahiro has joined #openstack-infra | 12:18 | |
| *** nuritv has quit IRC | 12:19 | |
| *** nuritv has joined #openstack-infra | 12:21 | |
| *** Masahiro has quit IRC | 12:23 | |
| *** liusheng has quit IRC | 12:23 | |
| *** nuritv has quit IRC | 12:24 | |
| *** liusheng has joined #openstack-infra | 12:24 | |
| *** talluri has quit IRC | 12:26 | |
| *** nuritv has joined #openstack-infra | 12:27 | |
| *** nuritv has quit IRC | 12:28 | |
| *** jpich has quit IRC | 12:30 | |
| *** dimsum__ has joined #openstack-infra | 12:32 | |
| *** nuritv has joined #openstack-infra | 12:32 | |
| *** jpich has joined #openstack-infra | 12:33 | |
| *** hashar has quit IRC | 12:33 | |
| *** acruz_ has joined #openstack-infra | 12:33 | |
| *** acruz__ has joined #openstack-infra | 12:34 | |
| *** nuritv has quit IRC | 12:34 | |
| *** nuritv has joined #openstack-infra | 12:36 | |
| *** dimsum__ has quit IRC | 12:37 | |
| *** EmilienM is now known as EmilienM|afk | 12:37 | |
| *** mpaolino has quit IRC | 12:38 | |
| *** nuritv has quit IRC | 12:38 | |
| *** nuritv has joined #openstack-infra | 12:43 | |
| *** dpaterson has joined #openstack-infra | 12:43 | |
| *** nuritv has quit IRC | 12:44 | |
| *** jgallard_ has joined #openstack-infra | 12:45 | |
| *** nuritv has joined #openstack-infra | 12:46 | |
| *** zz_avozza is now known as avozza | 12:47 | |
| *** baoli has joined #openstack-infra | 12:48 | |
| *** nuritv has quit IRC | 12:48 | |
| *** baoli_ has joined #openstack-infra | 12:48 | |
| *** nuritv has joined #openstack-infra | 12:49 | |
| *** nuritv has quit IRC | 12:51 | |
| *** penguinRaider has joined #openstack-infra | 12:52 | |
| *** baoli has quit IRC | 12:52 | |
| *** baoli__ has joined #openstack-infra | 12:52 | |
| *** dimsum__ has joined #openstack-infra | 12:52 | |
| *** baoli_ has quit IRC | 12:53 | |
| *** nuritv has joined #openstack-infra | 12:54 | |
| *** baoli__ has quit IRC | 12:54 | |
| *** baoli has joined #openstack-infra | 12:55 | |
| *** otter768 has joined #openstack-infra | 12:55 | |
| *** nuritv has quit IRC | 12:56 | |
| *** yolanda has quit IRC | 12:56 | |
| *** e0ne has quit IRC | 12:57 | |
| *** e0ne has joined #openstack-infra | 12:57 | |
| *** mfink has joined #openstack-infra | 12:57 | |
| *** nuritv has joined #openstack-infra | 12:58 | |
| *** nuritv has quit IRC | 13:00 | |
| *** otter768 has quit IRC | 13:00 | |
| *** erlon has joined #openstack-infra | 13:01 | |
| *** nuritv has joined #openstack-infra | 13:01 | |
| *** andreaf has quit IRC | 13:02 | |
| *** nuritv has quit IRC | 13:03 | |
| *** salv-orl_ has quit IRC | 13:04 | |
| *** jcoufal_ has joined #openstack-infra | 13:05 | |
| *** tnovacik has joined #openstack-infra | 13:06 | |
| *** kgiusti has joined #openstack-infra | 13:06 | |
| *** dprince has joined #openstack-infra | 13:08 | |
| *** jcoufal has quit IRC | 13:08 | |
| *** mbacchi has joined #openstack-infra | 13:09 | |
| *** nuritv has joined #openstack-infra | 13:10 | |
| *** Hal has joined #openstack-infra | 13:11 | |
| *** Hal is now known as Guest43007 | 13:11 | |
| *** Guest43007 has quit IRC | 13:12 | |
| *** Hal_ has joined #openstack-infra | 13:12 | |
| *** nuritv has quit IRC | 13:12 | |
| *** weshay has joined #openstack-infra | 13:13 | |
| *** nuritv has joined #openstack-infra | 13:14 | |
| *** nuritv has quit IRC | 13:16 | |
| *** EmilienM|afk is now known as EmilienM | 13:17 | |
| *** nuritv has joined #openstack-infra | 13:18 | |
| *** Masahiro has joined #openstack-infra | 13:19 | |
| *** nuritv has quit IRC | 13:20 | |
| *** yamamoto has joined #openstack-infra | 13:21 | |
| *** nuritv has joined #openstack-infra | 13:22 | |
| *** jp_at_hp has quit IRC | 13:22 | |
| *** ddieterly has joined #openstack-infra | 13:22 | |
| *** acruz__ has quit IRC | 13:23 | |
| *** mpaolino has joined #openstack-infra | 13:23 | |
| *** nuritv has quit IRC | 13:24 | |
| *** acruz_ has quit IRC | 13:24 | |
| *** Masahiro has quit IRC | 13:24 | |
| *** acruz_ has joined #openstack-infra | 13:24 | |
| *** koolhead17 has joined #openstack-infra | 13:24 | |
| *** bswartz has quit IRC | 13:25 | |
| *** acruz_ has quit IRC | 13:25 | |
| *** arxcruz has quit IRC | 13:25 | |
| *** nuritv has joined #openstack-infra | 13:26 | |
| *** arxcruz has joined #openstack-infra | 13:27 | |
| *** nuritv has quit IRC | 13:28 | |
| *** jistr has joined #openstack-infra | 13:28 | |
| *** arxcruz has quit IRC | 13:31 | |
| *** arxcruz has joined #openstack-infra | 13:31 | |
| *** mfink has quit IRC | 13:32 | |
| *** mfink has joined #openstack-infra | 13:33 | |
| *** nuritv has joined #openstack-infra | 13:34 | |
| *** jp_at_hp has joined #openstack-infra | 13:34 | |
| *** hashar has joined #openstack-infra | 13:35 | |
| *** nuritv has quit IRC | 13:35 | |
| *** jistr has quit IRC | 13:36 | |
| *** nuritv has joined #openstack-infra | 13:38 | |
| *** nuritv has quit IRC | 13:39 | |
| *** salv-orlando has joined #openstack-infra | 13:40 | |
| *** yamamoto has quit IRC | 13:40 | |
| *** yamamoto has joined #openstack-infra | 13:41 | |
| *** avozza is now known as zz_avozza | 13:41 | |
| *** nuritv has joined #openstack-infra | 13:41 | |
| *** andreaf has joined #openstack-infra | 13:42 | |
| *** nuritv has quit IRC | 13:43 | |
| *** _nadya_ has quit IRC | 13:44 | |
| *** nuritv has joined #openstack-infra | 13:47 | |
| *** nuritv has quit IRC | 13:48 | |
| *** dkranz has quit IRC | 13:48 | |
| *** dkranz has joined #openstack-infra | 13:49 | |
| *** unicell has joined #openstack-infra | 13:50 | |
| *** nuritv has joined #openstack-infra | 13:51 | |
| *** nuritv has quit IRC | 13:53 | |
| *** dizquierdo has joined #openstack-infra | 13:54 | |
| *** mugsie_ is now known as mugsie | 13:55 | |
| *** nuritv has joined #openstack-infra | 13:56 | |
| andreaf | hi - I'm seeing a weird dependency issue when running devstack - http://paste.openstack.org/show/151187/ - it seems that the version from novaclient (installed from git) cannot be matched against the version expected by openstackclient (also installed from git). Has anyone seen this before? | 13:56 |
| *** nuritv has quit IRC | 13:57 | |
| sdague | andreaf: what run is this in? | 13:59 |
| *** nuritv has joined #openstack-infra | 14:00 | |
| *** sabeen1 has joined #openstack-infra | 14:01 | |
| *** mattfarina has joined #openstack-infra | 14:01 | |
| *** pblaho has quit IRC | 14:01 | |
| *** pblaho has joined #openstack-infra | 14:01 | |
| *** nuritv has quit IRC | 14:02 | |
| *** yaguang has joined #openstack-infra | 14:02 | |
| *** dimsum__ is now known as dims | 14:02 | |
| *** bswartz has joined #openstack-infra | 14:03 | |
| *** lttrl has joined #openstack-infra | 14:04 | |
| AJaeger | andreaf, that's a known issue with setuptools 8.0 that clarkb, dstufft and fungi worked on over the weekend. | 14:04 |
| *** marcusvrn has quit IRC | 14:05 | |
| AJaeger | andreaf, please recheck, I expect it's worked around now. If it still occurs, please tell us again. | 14:05 |
| *** nuritv has joined #openstack-infra | 14:06 | |
| *** marcusvrn has joined #openstack-infra | 14:06 | |
| odyssey4me | might it be this issue? https://bugs.launchpad.net/python-openstackclient/+bug/1402679 | 14:06 |
| uvirtbot | Launchpad bug 1402679 in python-openstackclient "setuptools 8 breaks python-openstackclient" [Undecided,New] | 14:07 |
| openstackgerrit | yolanda.robla proposed openstack-infra/storyboard: Return task count by status for a story dynamically https://review.openstack.org/139100 | 14:07 |
| *** nuritv has quit IRC | 14:08 | |
| *** fandi has joined #openstack-infra | 14:10 | |
| *** _nadya_ has joined #openstack-infra | 14:10 | |
| *** mjturek has joined #openstack-infra | 14:10 | |
| *** nuritv has joined #openstack-infra | 14:11 | |
| *** mpaolino has quit IRC | 14:11 | |
| openstackgerrit | Dmitry Teselkin proposed openstack-infra/zuul: Process event multiple times. https://review.openstack.org/136027 | 14:11 |
| AJaeger | odyssey4me, yeah, same problem. Read the discussion on this channel from Saturday especially. | 14:12 |
| *** nuritv has quit IRC | 14:12 | |
| *** yolanda has joined #openstack-infra | 14:13 | |
| *** erikmwilson has joined #openstack-infra | 14:13 | |
| *** doug-fish has joined #openstack-infra | 14:13 | |
| AJaeger | odyssey4me, There's a new version of pbr out since yesterday that will handle this. | 14:13 |
| *** nuritv has joined #openstack-infra | 14:14 | |
| AJaeger | andreaf, sdague, odyssey4me: pep440 changes, basically the get versions like python-glanceclient 0.14.2.2.gb126351 are not pep440 compliant. | 14:14 |
| *** dkliban has joined #openstack-infra | 14:14 | |
| AJaeger | and thus setuptools 8 will have this version tuple: (-1, 0, 14, 2,2,...) for it - and thus 0.14.2.2.gb12... is less than 0.0 ;( | 14:15 |
| odyssey4me | AJaeger: thanks for the pointer - will check on that. | 14:15 |
| AJaeger | Using "0.14.2.2+gb..." fixes this | 14:15 |
| AJaeger | and pbr 0.10.2 contains a fix. | 14:15 |
| *** nuritv has quit IRC | 14:16 | |
| AJaeger | .. to use "+" instead "." in that place | 14:16 |
| *** erikmwilson has joined #openstack-infra | 14:16 | |
| *** dizquierdo has quit IRC | 14:16 | |
| fungi | AJaeger: odyssey4me: actually it doesn't. we emergency released that as 0.10.1 but then reverted it immediately and released 0.10.2 without it, because it turns out that setuptools is unable to parse requirement file entries containing a + | 14:17 |
| odyssey4me | AJaeger: I just tested using pbr 0.10.2 and setuptools 8.0.2 and it doesn't work | 14:17 |
| fungi | right now we have setuptools pinned <8 in our tests | 14:17 |
| fungi | andreaf: sdague: ^ | 14:18 |
| odyssey4me | thanks fungi - we'll do the same for now until a fix is identified | 14:19 |
| fungi | there's another proposed workaround to get rid of the git sha in pbr version strings entirely and move them into egg info metadata | 14:19 |
| sdague | fungi: so that does mean however that all deployed users that get a setuptools upgrade are broken, right? | 14:19 |
| *** julim has joined #openstack-infra | 14:19 | |
| fungi | sdague: yes, setuptools 8 is currently operating in a non-backward-compatible fashion | 14:20 |
| *** nuritv has joined #openstack-infra | 14:20 | |
| fungi | and also there was some talk of loosening the pep 440 normalization in another setuptools point release to allow arbitrary alphanumeric string subcomponents | 14:20 |
| fungi | which would be another way out of this | 14:20 |
| *** nuritv has quit IRC | 14:21 | |
| fungi | sdague: keep in mind this specific issue should only surface when trying to pip-install from non-released source and then require newer than the released version in your dependencies | 14:22 |
| fungi | which unfortunately we do in our test infrastructure | 14:22 |
| fungi | the setuptools pin was to allow us more time to figure out a way forward between pbr and setuptools/pip developers | 14:23 |
| *** aysyd has quit IRC | 14:23 | |
| clarkb | also oslo.db | 14:24 |
| openstackgerrit | Mike Heald proposed openstack-infra/infra-specs: Proposal for a 'fail-fast' option on jobs https://review.openstack.org/129255 | 14:24 |
| clarkb | we should fix the sqlalchemy requirement there asap | 14:24 |
| fungi | oh, has that not merged yet? | 14:24 |
| fungi | crap | 14:24 |
| *** nuritv has joined #openstack-infra | 14:24 | |
| clarkb | master needs a release | 14:24 |
| sdague | fungi: actually, I thought it was the opposite | 14:24 |
| clarkb | juno needs merging | 14:24 |
| sdague | clarkb: which repo? | 14:25 |
| clarkb | oslo.db | 14:25 |
| *** sdake has quit IRC | 14:25 | |
| sdague | ah | 14:25 |
| fungi | ahh, that's why i was confused. forgot about the backport | 14:25 |
| jd__ | can someone validate https://review.openstack.org/#/c/140343/ because I think it's blocking the tooz gate for over a week :( | 14:25 |
| AJaeger | fungi, ah, I didn't read all of the history. Thanks for the explanation! | 14:25 |
| *** zz_avozza is now known as avozza | 14:25 | |
| *** vigneshvar_ has quit IRC | 14:25 | |
| fungi | AJaeger: yeah, it was a lot of stabbing in the dark for a good chunk of the weekend to figure out a way forward | 14:26 |
| clarkb | oslo.db as released uses a version requirement string that new setuptools explodes on | 14:26 |
| clarkb | so need to release new versions that dont | 14:26 |
| *** nuritv has quit IRC | 14:26 | |
| fungi | i take it that was the only project we found still using that multi-range pattern? | 14:26 |
| AJaeger | fungi, clarkb: Could either of you write a summary mail to openstack-dev before more people ask, please? | 14:27 |
| clarkb | no a lot do but most have it overridden | 14:27 |
| fungi | i think we had been doing something similar for hacking transitions for a while, which well need to stop doing | 14:27 |
| *** aysyd has joined #openstack-infra | 14:27 | |
| *** dkliban has quit IRC | 14:27 | |
| sdague | hey, question | 14:27 |
| clarkb | so are less important in the fix now category. but with a pin it may not matter anyway | 14:27 |
| sdague | where is az3.clouds.archive.ubuntu.com coming from? | 14:27 |
| sdague | in devstack runs | 14:27 |
| *** dustins has joined #openstack-infra | 14:27 | |
| *** dkliban has joined #openstack-infra | 14:27 | |
| sdague | is that a mirror we set explicitly? | 14:27 |
| clarkb | sdague no thats hpcloud base image | 14:28 |
| * fungi loves that hpcloud gets packages from amazon's cloud | 14:28 | |
| *** dizquierdo has joined #openstack-infra | 14:28 | |
| AJaeger | rol | 14:28 |
| sdague | so all the failures on apt fetch failures seem to be to that host | 14:28 |
| fungi | i'll start drafting a setuptools redux/sotu in etherpad and others can jump in and correct any misstatements | 14:29 |
| AJaeger | sdague, could you review ACL normalization patch - just "cosmetics", please: https://review.openstack.org/#/c/140822/ | 14:29 |
| *** nuritv has joined #openstack-infra | 14:29 | |
| clarkb | AJaeger while a summary should be written that wont stop the asking :) | 14:29 |
| fungi | draft will be in https://etherpad.openstack.org/p/hFwpOHmMWX shortly | 14:29 |
| AJaeger | clarkb, yeah but it will avoid that I give wrong answers ;( | 14:30 |
| sdague | are we should that's actually an amazon server? | 14:30 |
| *** nuritv has quit IRC | 14:31 | |
| AJaeger | thanks, sdague ! | 14:31 |
| sdague | clarkb: do you know which file it's in? I'd like to try changing the mirror location and see if it helps | 14:31 |
| clarkb | /etc/apt/sources.list? | 14:32 |
| fungi | sdague: actually i guess it's not necessarily in amazon. seems to just be a dns round-robin between a bunch of canonical's ip addresses | 14:32 |
| sdague | fungi: yeh | 14:32 |
| sdague | clarkb: it might be there, or they might have put it in a .d/ file | 14:33 |
| sdague | mostly I wanted a live example to know | 14:33 |
| Guest91362 | have any of you looked into translating the infra manual? | 14:33 |
| Guest91362 | erg | 14:33 |
| *** Guest91362 is now known as annegentle | 14:33 | |
| openstackgerrit | Nikita Konovalov proposed openstack-infra/python-storyboardclient: Setting up base classes https://review.openstack.org/138092 | 14:33 |
| annegentle | I just noticed that there's a translated version of GerritWorkflow | 14:34 |
| sdague | because on every run which failed against az3, the rax mirrors actually seemed to work correctly | 14:34 |
| *** nuritv has joined #openstack-infra | 14:34 | |
| clarkb | they are different mirrors | 14:36 |
| sdague | clarkb: right, I know | 14:36 |
| openstackgerrit | Dmitry Teselkin proposed openstack-infra/zuul: Process event multiple times. https://review.openstack.org/136027 | 14:36 |
| sdague | but the point being, we actually have a mirror check bit in d-g | 14:36 |
| sdague | which checks connectivity to rax mirrors on each run | 14:36 |
| *** nuritv has quit IRC | 14:36 | |
| openstackgerrit | Merged openstack-infra/devstack-gate: Give a more specific error message when giving up in git_remote_update https://review.openstack.org/135656 | 14:36 |
| *** dizquierdo has quit IRC | 14:36 | |
| openstackgerrit | Merged openstack-infra/project-config: Normalize Gerrit ACLs https://review.openstack.org/140822 | 14:36 |
| sdague | which is passing when az3 connections fail | 14:37 |
| andreaf | AJaeger, fungi, sdague: thanks ^^^ | 14:39 |
| *** nuritv has joined #openstack-infra | 14:39 | |
| *** alaski_ is now known as alaski | 14:39 | |
| *** mriedem has joined #openstack-infra | 14:39 | |
| *** eharney has joined #openstack-infra | 14:41 | |
| *** eharney has quit IRC | 14:41 | |
| *** eharney has joined #openstack-infra | 14:41 | |
| *** nuritv has quit IRC | 14:41 | |
| *** sdake has joined #openstack-infra | 14:41 | |
| fungi | sdague: it's hit and miss though. a few months back the rackspace package mirrors were regularly unusable | 14:42 |
| fungi | they seem to trade places as to who's broken when | 14:42 |
| *** achanda has joined #openstack-infra | 14:42 | |
| *** nuritv has joined #openstack-infra | 14:43 | |
| fungi | clarkb: do we have any open lp bugs or sb tasks for any of the setuptools 8 stuff? mostly looking for good examples of the two classes of related failures | 14:43 |
| *** nikil has quit IRC | 14:43 | |
| *** nuritv has quit IRC | 14:44 | |
| clarkb | fungi for the first its the sqlalchemy issue second is pbr gitsha1 | 14:44 |
| *** _nadya_ has quit IRC | 14:45 | |
| AJaeger | fungi, odyssey4me mentioned https://bugs.launchpad.net/python-openstackclient/+bug/1402679 | 14:45 |
| uvirtbot | Launchpad bug 1402679 in python-openstackclient "setuptools 8 breaks python-openstackclient" [Undecided,New] | 14:45 |
| fungi | clarkb: right, i was looking for the error messages and tracebacks we were getting, but maybe that level of detail is unnecessary | 14:45 |
| clarkb | and there is that bug on openstackclient. I was going to file a bug on setuptools then realized this was intentional | 14:45 |
| sdague | fungi: sure | 14:46 |
| clarkb | fungi ya ruhe had links on saturday | 14:46 |
| *** bnemec has joined #openstack-infra | 14:46 | |
| fungi | on second thought, i'll just wave hands vigorously and omit examples for now | 14:46 |
| ttx | clarkb, fungi: Blocked on Swift 2.2.1.rc1 release due to the issue I posted on the -infra ML | 14:47 |
| clarkb | part of the issue is thia manifests not just in setuptools errorz eg ^ | 14:47 |
| clarkb | and heat template thing breaking and glance image uploads failing | 14:48 |
| clarkb | ttx yes known issue with setuptools release | 14:48 |
| *** dizquierdo has joined #openstack-infra | 14:48 | |
| clarkb | you are likely not going to want to use 'rc' anymore | 14:48 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/project-config: move oslo.version to attic https://review.openstack.org/141812 | 14:48 |
| *** sdake has quit IRC | 14:49 | |
| ttx | clarkb: ok, cool (or not) | 14:49 |
| fungi | ttx: pep 440 is the future, and the future is now | 14:49 |
| clarkb | ttx we can make an sdist with older setuptools for this release though | 14:49 |
| clarkb | maybe | 14:49 |
| *** nuritv has joined #openstack-infra | 14:49 | |
| annegentle | ttx: one other idea on the cross-project current states, is to add it to the cross-project meeting? | 14:49 |
| ttx | clarkb: that would be cool. I only need one | 14:49 |
| annegentle | ttx: I could do a short report this week on docs | 14:50 |
| ttx | annegentle: that would work for me, but the agenda is already a bit heavy for this week | 14:50 |
| annegentle | ttx: really I just want to find out more from everyone | 14:50 |
| ttx | annegentle: we can try to squeeze it | 14:50 |
| *** signed8bit has joined #openstack-infra | 14:50 | |
| *** wenlock has joined #openstack-infra | 14:50 | |
| annegentle | ttx: yeah I'm looking at the agenda now, the types of things I'm asking about are going to play into the "reform spec" line | 14:50 |
| annegentle | ttx: so I don't want to overload | 14:50 |
| *** spzala has joined #openstack-infra | 14:51 | |
| *** nuritv has quit IRC | 14:51 | |
| annegentle | ttx: and this is informal, also, sort of a "commiseration" session | 14:51 |
| ttx | clarkb: I need a swift rc1 today, next time I need some RC tag should be end of January, so gives us time to discuss what the future is | 14:51 |
| clarkb | ttx and you just need an sdist? | 14:51 |
| ttx | clarkb: yes, a swift-tarball job | 14:52 |
| clarkb | problem is we should be careful whereever possible to be compat with new setuptools | 14:52 |
| ttx | clarkb: so sdist + upload | 14:52 |
| ttx | (to tarballs.o.o/swift) | 14:52 |
| clarkb | so I dont just want to do a thing and have it not work sincs we already had enough of that tjis weekend | 14:52 |
| clarkb | but rc should be normalized | 14:53 |
| ttx | clarkb: alternative is to delete tag and push one that would be PEP440 futureproof instead | 14:53 |
| clarkb | oh I bet it spat out a normalized sdist name | 14:53 |
| ttx | I just need tag to match tarball | 14:53 |
| *** nuritv has joined #openstack-infra | 14:54 | |
| clarkb | which failed to upload because the name was wrong | 14:54 |
| ttx | it spit out a 2.2.1c1 | 14:54 |
| clarkb | ya | 14:54 |
| clarkb | woo | 14:54 |
| clarkb | so normalization is more aggressive than I hoped | 14:55 |
| ttx | it was uploaded alright, I just don't want to have a tag that says 2.2.1.rc1 and a tarball that says 2.2.1c1 | 14:55 |
| clarkb | yup | 14:55 |
| *** rkukura has quit IRC | 14:55 | |
| *** nuritv has quit IRC | 14:55 | |
| * ttx goes to read that crazy spec that can't handle rcs | 14:55 | |
| clarkb | I think in that case it is ok to rebuild with old setuptools | 14:56 |
| *** otter768 has joined #openstack-infra | 14:56 | |
| *** stevemar has joined #openstack-infra | 14:56 | |
| clarkb | fungi ^ make sense ? | 14:56 |
| ttx | clarkb: yes, final should be 2.2.1 and all happy | 14:56 |
| fungi | wow | 14:57 |
| *** ddieterly has quit IRC | 14:57 | |
| clarkb | also ^ breaks our ability to use dstuffts suggestion for trailing alphas | 14:57 |
| clarkb | so that is a nonstarter imo | 14:57 |
| *** nuritv has joined #openstack-infra | 14:57 | |
| fungi | so setuptools normalization is causing pbr to rewrite the version numbers generated when building an sdist | 14:57 |
| ttx | hmm, looks like c1 is the way of the future. I have seen the future, and it's ugly and not-self-explaniing. | 14:57 |
| clarkb | fungi well pbr says X and setuptools says X' | 14:58 |
| fungi | f(x) | 14:58 |
| * fungi nods | 14:58 | |
| ttx | weeird... the spec is pretty accepting of 'rc' as an equivalent of 'c' | 14:59 |
| fungi | yeah i think regenerating the sdist with setuptools 7.x should be okay for now | 14:59 |
| *** nuritv has quit IRC | 14:59 | |
| *** Sincler has joined #openstack-infra | 14:59 | |
| clarkb | hrm sdist build is in a tox venv though how did swift use ltest setuptools? | 14:59 |
| fungi | ttx: it accepts it in a version match, but doesn't emit it and i think pbr is letting setuptools decide what the version number should be when making the sdist | 15:00 |
| fungi | mordred: ^ new wrinkle | 15:00 |
| clarkb | rc is normalized to c | 15:00 |
| ttx | fungi: yeah. That's why we get the nice warning, followed by a quick kick in the nuts | 15:00 |
| clarkb | I think setuptools does that normalization whenever it does its thing | 15:00 |
| clarkb | tldr welcome to the new world order | 15:00 |
| *** otter768 has quit IRC | 15:00 | |
| clarkb | before we rebuild though we should figure out how sdist building used latest setuptools in the first place | 15:01 |
| fungi | so anyway, i'm guessing either that tag was pushed before we had setuptools pinning in place, or the pin isn't working for tarball jobs | 15:01 |
| *** rkukura has joined #openstack-infra | 15:01 | |
| ttx | clarkb: if retriggering swift-tarball to get a rc1 tarball is too awkward, we can probably push 2.2.1c1 instead... I'd just need you to delete the 2.2.1.rc1 tag from swift git repo | 15:02 |
| *** avozza is now known as zz_avozza | 15:02 | |
| *** nuritv has joined #openstack-infra | 15:02 | |
| clarkb | ttx well it used tox to build which should use old setuptools | 15:02 |
| clarkb | so I dont understand how this failed yet. and need to if a rebuild is going to work | 15:03 |
| *** sdake has joined #openstack-infra | 15:03 | |
| clarkb | fungi except its tox | 15:03 |
| clarkb | so should be unrelated to pinning | 15:03 |
| clarkb | unless | 15:03 |
| clarkb | does swift allow site packages? | 15:03 |
| *** nuritv has quit IRC | 15:04 | |
| ttx | sounds like a rhetorical question | 15:04 |
| clarkb | I dont think they do. so how did this fail? | 15:05 |
| fungi | clarkb: any chance something was installed into the virtualenv with -U and there was a gratuitous dependency on setuptools somewhere? | 15:05 |
| clarkb | oh maybe there is a -U | 15:05 |
| *** dmsimard_away is now known as dmsimard | 15:06 | |
| clarkb | so a rebuild likely wont fix :/ | 15:06 |
| clarkb | it will just update and get back on the pep440 party train | 15:06 |
| *** sdake has quit IRC | 15:07 | |
| *** nelsnelson has joined #openstack-infra | 15:07 | |
| *** sdake has joined #openstack-infra | 15:07 | |
| openstackgerrit | omri marcovitch proposed openstack-infra/system-config: Add Third Party FAQ paragraph https://review.openstack.org/141817 | 15:08 |
| *** dangers_away has quit IRC | 15:09 | |
| *** ihrachyshka has quit IRC | 15:09 | |
| *** nuritv has joined #openstack-infra | 15:09 | |
| openstackgerrit | omri marcovitch proposed openstack-infra/system-config: Add Third Party FAQ paragraph https://review.openstack.org/141817 | 15:10 |
| *** dangers_away has joined #openstack-infra | 15:10 | |
| *** superdan is now known as dansmith | 15:10 | |
| *** nuritv has quit IRC | 15:11 | |
| *** bknudson has quit IRC | 15:11 | |
| *** ihrachyshka has joined #openstack-infra | 15:12 | |
| *** esker has joined #openstack-infra | 15:14 | |
| *** nuritv has joined #openstack-infra | 15:14 | |
| fungi | clarkb: sdague: okay, rough stab at a message to the dev ml at https://etherpad.openstack.org/p/hFwpOHmMWX | 15:15 |
| fungi | please hack that up as necessary | 15:15 |
| *** nuritv has quit IRC | 15:16 | |
| clarkb | lgtm | 15:17 |
| openstackgerrit | Merged openstack-infra/devstack-gate: Add tooz in project list https://review.openstack.org/140343 | 15:17 |
| *** bknudson has joined #openstack-infra | 15:18 | |
| clarkb | wow that was fast | 15:18 |
| clarkb | fungi so for the swift thing you think new tag too? | 15:18 |
| *** nuritv has joined #openstack-infra | 15:19 | |
| fungi | clarkb: did you determine if there was an install -U happening in there? | 15:19 |
| * fungi finds the log | 15:20 | |
| clarkb | there is a -U in the tox.ini install command. I did not confidm if that is forcing setuptools to update (it is still really early here) | 15:20 |
| clarkb | fungi ttx put a link in his email | 15:20 |
| fungi | oh, i haven't gotten to e-mail yet | 15:20 |
| fungi | irc seemed to be on fire | 15:20 |
| *** nuritv has quit IRC | 15:20 | |
| ttx | fungi: -infra ml | 15:22 |
| *** nuritv has joined #openstack-infra | 15:23 | |
| *** zz_dimtruck is now known as dimtruck | 15:23 | |
| *** salv-orlando has quit IRC | 15:23 | |
| *** nuritv has quit IRC | 15:24 | |
| *** nuritv has joined #openstack-infra | 15:26 | |
| *** liusheng has quit IRC | 15:26 | |
| *** _nadya_ has joined #openstack-infra | 15:26 | |
| *** liusheng has joined #openstack-infra | 15:27 | |
| *** wenlock has quit IRC | 15:27 | |
| *** nuritv has quit IRC | 15:27 | |
| fungi | we should probably start archiving the tox logs from things like tarball jobs too | 15:30 |
| fungi | er, tox's pip logs | 15:30 |
| *** timcline has joined #openstack-infra | 15:30 | |
| *** annegent_ has joined #openstack-infra | 15:31 | |
| *** timcline has quit IRC | 15:31 | |
| *** timcline has joined #openstack-infra | 15:31 | |
| *** nuritv has joined #openstack-infra | 15:32 | |
| openstackgerrit | omri marcovitch proposed openstack-infra/system-config: Add Third Party FAQ paragraph https://review.openstack.org/141817 | 15:32 |
| *** teran has quit IRC | 15:32 | |
| openstackgerrit | Merged openstack-infra/project-config: Introduce Oslo-specific stable-maint team https://review.openstack.org/141769 | 15:33 |
| *** viktors has left #openstack-infra | 15:33 | |
| fungi | i'll get the group for that ^ fixed up shortly | 15:33 |
| *** teran has joined #openstack-infra | 15:33 | |
| *** nuritv has quit IRC | 15:34 | |
| *** teran_ has joined #openstack-infra | 15:34 | |
| clarkb | I guess I can ru that locally and checj tox log | 15:35 |
| fungi | yeah, the console log is silent about virtualenv setup | 15:36 |
| *** erikmwilson has quit IRC | 15:37 | |
| *** jerryz has quit IRC | 15:37 | |
| *** nuritv has joined #openstack-infra | 15:37 | |
| anteaya | there is a patch up to add more logs to swift: https://review.openstack.org/#/c/141668/ what is the feedback so far on logs in swift? | 15:37 |
| anteaya | more is good? | 15:37 |
| *** erikmwilson has joined #openstack-infra | 15:38 | |
| openstackgerrit | Julien Danjou proposed openstack/requirements: pymemcache and sysv_ipc for tooz https://review.openstack.org/140920 | 15:38 |
| clarkb | anteaya yes wanted to test it on devstack volume of logs and those jobs shoild only run on dg changes | 15:38 |
| *** achuprin_ has quit IRC | 15:38 | |
| anteaya | okay so +A would make us happy then | 15:38 |
| *** nuritv has quit IRC | 15:38 | |
| *** teran has quit IRC | 15:38 | |
| anteaya | just wanted to confirm | 15:38 |
| *** liusheng has quit IRC | 15:38 | |
| *** liusheng has joined #openstack-infra | 15:39 | |
| anteaya | boo, I have a nit about the name | 15:40 |
| clarkb | anteaya: name? | 15:40 |
| anteaya | underscores and a hyphen in the new name :( | 15:40 |
| anteaya | zuul_swift_devstack-logs | 15:40 |
| *** cdent has joined #openstack-infra | 15:41 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/project-config: move oslo.version to attic https://review.openstack.org/141812 | 15:42 |
| clarkb | ok I reproduced locally. Now to read logs | 15:42 |
| fungi | if nobody else has any additional corrections/suggestions/detail for https://etherpad.openstack.org/p/hFwpOHmMWX i'll send it to the dev ml in about 15 minutes | 15:43 |
| clarkb | fungi: confirmed that tox is installing 8.0.2 into the venv so a rebuild won't help unless we hobble its env a bit more | 15:44 |
| anteaya | fungi: I have read it, I have nothing to add | 15:44 |
| *** nuritv has joined #openstack-infra | 15:44 | |
| clarkb | send it | 15:44 |
| clarkb | fungi: ttx: I think that means simplest thing is a new tag | 15:44 |
| *** amitgandhinz has joined #openstack-infra | 15:44 | |
| *** e0ne_ has joined #openstack-infra | 15:44 | |
| fungi | re-tag 1.2.3.rc1 as 1.2.3c1? | 15:45 |
| clarkb | fungi: ya, then the tag version and sdist version will match each other | 15:45 |
| *** erikwilson has joined #openstack-infra | 15:45 | |
| *** erikmwilson has quit IRC | 15:45 | |
| clarkb | which seems to be ttx's biggest concern | 15:46 |
| *** nuritv has quit IRC | 15:46 | |
| *** erikwilson is now known as erikmwilson | 15:46 | |
| ttx | clarkb: well, I have to patch all my scripts too but I can survive that | 15:46 |
| *** wenlock has joined #openstack-infra | 15:47 | |
| *** e0ne has quit IRC | 15:47 | |
| ttx | if you guys remove the 2.2.1.rc1 tag I can retag 2.2.1c1 | 15:47 |
| fungi | ttx: any reason not to just have both tags on that commit? | 15:47 |
| * anteaya wonders if this is a case for y'all? | 15:47 | |
| *** nuritv has joined #openstack-infra | 15:48 | |
| ttx | fungi: none except noise and potential confusion | 15:48 |
| *** tonytan4ever has joined #openstack-infra | 15:48 | |
| fungi | i can push a null string to the old tag if you really want, though people/systems which have updated their remotes likely will have that old tag indefinitely cached | 15:48 |
| clarkb | dhellmann: before I go wanted ot make sure you saw that oslo.db could use a new release to help alleviate some of the new setuptools issues. Short story is the requirement on sqlalchemy in the current release is not parsed properly by setuptools | 15:48 |
| fungi | since tag removals don't convey on remote update | 15:48 |
| fungi | (though it will propagate to our replicas) | 15:48 |
| *** mjturek has quit IRC | 15:48 | |
| clarkb | dhellmann: the issue has been fixed on master but we consume it via releases in much of the testing now so a release would be awesome (similar thing likely needs to be done forthe stable/juno branch) | 15:49 |
| dhellmann | clarkb: I've had that reported. I am about to go into the oslo team meeting, and haven't had time to look carefully at what we'd be releasing, but we'll be addressing it after the meeting | 15:49 |
| clarkb | dhellmann: perfect. thank you | 15:49 |
| *** nuritv has quit IRC | 15:49 | |
| ttx | fungi: will existing http://tarballs.openstack.org/swift/swift-2.2.1c1.* be overwritten ? Should be a straight scp so ok, right ? | 15:49 |
| dhellmann | I know glance is blocked, but I don't want to break things worse by rushing out a relase | 15:50 |
| fungi | ttx: yes, it will get overwritten | 15:50 |
| clarkb | dhellmann: well at least as an upstream we should be fine with our pinning of setuptools | 15:50 |
| ttx | fungi: I guess I can already push the c1 tag | 15:50 |
| dhellmann | clarkb: so this problem was triggered by the setuptools release? | 15:50 |
| clarkb | dhellmann: so glance should be able to do things. HOwever anyone that hasn't pinned setuptools will need it and eventually we would like to remove the setuptools pin | 15:50 |
| dhellmann | sure, of course | 15:50 |
| clarkb | dhellmann: yes setuptools 8.0 broke the world in about 3 different ways | 15:51 |
| *** radez_g0n3 is now known as radez | 15:51 | |
| clarkb | dhellmann: the sqlalchemy thing is one of them :) | 15:51 |
| fungi | ttx: yep, and we can retrigger jobs for that new pep 440 normalized tag name if we find other subsequent issues | 15:51 |
| *** nuritv has joined #openstack-infra | 15:51 | |
| openstackgerrit | Merged stackforge/python-jenkins: Fix it so build_job triggers a build https://review.openstack.org/122462 | 15:51 |
| dhellmann | clarkb: Are we talking to them about helping with testing or anything? | 15:52 |
| *** bnemec has quit IRC | 15:52 | |
| clarkb | dhellmann: we have spoken with dstufft at lenght. Basically pep440 got approved and is now implemented for better or worse | 15:52 |
| *** salv-orlando has joined #openstack-infra | 15:52 | |
| * dhellmann adds pep440 to his reading list | 15:52 | |
| *** nuritv has quit IRC | 15:53 | |
| clarkb | mordred has a change to pbr too that will be helpful (it drops git sha info from versions and puts it in egg info metadata) | 15:53 |
| *** dizquierdo has quit IRC | 15:53 | |
| dhellmann | ok, I'll keep an eye out for that, too | 15:53 |
| ttx | fungi: so you'd rather keep the rc1 tag in the tree for everyone to sync ? | 15:53 |
| clarkb | I should write a quick patch to our pre release acceptance regex to reject all the stuff that gets normalized too | 15:53 |
| ttx | 2.2.1c1 pushed | 15:53 |
| *** che-arne has joined #openstack-infra | 15:55 | |
| fungi | ttx: well, it's not so much that i'd rather keep it around or not. i can see an argument that having both tag names on that same commit makes it more obvious we changed the tag naming pattern at that point | 15:55 |
| mriedem | devstack/tempest vm's are 8GB RAM / 8 VCPU, but how much disk is allocated to those VMs? | 15:55 |
| *** wenlock has left #openstack-infra | 15:55 | |
| clarkb | mriedem: it depends on cloud provider. its something like 80GB on rax and 200GB on hpcloud iirc | 15:55 |
| ttx | alrighty then | 15:55 |
| ttx | welcome to the future | 15:55 |
| *** bnemec has joined #openstack-infra | 15:55 | |
| mriedem | clarkb: wow, that seems huge | 15:55 |
| fungi | ttx: but i'm happy to delete it if you want it gone for cleanliness, just be aware that it will still be indefinitely cached in some places | 15:55 |
| openstackgerrit | Jan Klare proposed openstack-infra/project-config: Added gate-{name}-chef-rake job-template https://review.openstack.org/137134 | 15:55 |
| ttx | fungi: I'll let notmyname decide | 15:56 |
| fungi | sounds like a good plan | 15:56 |
| *** nuritv has joined #openstack-infra | 15:56 | |
| mriedem | clarkb: i guess it probably has to be big for grenade and the large cinder volumes created by devstack? | 15:56 |
| *** wenlock has joined #openstack-infra | 15:56 | |
| clarkb | mriedem: ya I think we give at least 24GB to cinder alone | 15:56 |
| mriedem | clarkb: ok. we were still doing 10GB cinder volume in our internal CI, but we're not running the same number of tests (or grenade for that matter) | 15:57 |
| *** nuritv has quit IRC | 15:58 | |
| mtreinish | mriedem: grenade doesn't use as much volume storage (less tests) the 24GB is probably overkill TBH, we should be able to get away with less | 15:58 |
| mriedem | but 200GB, wow | 15:58 |
| mriedem | per VM | 15:58 |
| mriedem | with up to max ~800 VMs? | 15:58 |
| mtreinish | mriedem: that's just because the hp flavor is big | 15:58 |
| *** amotoki has joined #openstack-infra | 15:58 | |
| mtreinish | it also has like 30GB of ram iirc | 15:58 |
| mriedem | mtreinish: oh i thought they were 8GB RAM/8 VCPU pretty standard | 15:59 |
| mriedem | or is that just min? | 15:59 |
| mtreinish | mriedem: it's 8GB because of a kernel param | 15:59 |
| *** bnemec has quit IRC | 15:59 | |
| *** mjturek has joined #openstack-infra | 15:59 | |
| mtreinish | that flavor was the only hp one with 8 vcpus to match rax nodes | 15:59 |
| clarkb | mtreinish: yup 30GB of ram artificially limited to 8GB | 15:59 |
| clarkb | mtreinish: well more importantly it was the smallest flavor that had us doing reasonable tempest run times | 16:00 |
| *** bnemec has joined #openstack-infra | 16:00 | |
| *** nuritv has joined #openstack-infra | 16:00 | |
| *** mjturek has quit IRC | 16:00 | |
| *** mjturek has joined #openstack-infra | 16:01 | |
| fungi | right, any smaller than that and hpcloud instances were way slower than rackspace, so it was an effort to get their test durations into the same ballpark | 16:01 |
| anteaya | omrim: collecting more information is what the third party meetings are for | 16:01 |
| *** nuritv has quit IRC | 16:02 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/project-config: Make prerelease and release pipelines match pep440 https://review.openstack.org/141831 | 16:02 |
| clarkb | fungi: ttx ^ | 16:02 |
| krotscheck | Storyboard meeting in #openstack-meeting-3 | 16:03 |
| clarkb | I used the normalized form grammar to base that on | 16:03 |
| omrim | anteaya: Sure, I just thought that it will be helpful to collect as more as we can frequent questions for the page... | 16:03 |
| mtreinish | clarkb: my regex sucks, so I'll just ask, is the tempest integer tagging effected by that? | 16:03 |
| mriedem | mtreinish: the volumes created by tempest are 1GB aren't they? | 16:03 |
| anteaya | omrim: it will be, and the meetings are your first stop | 16:03 |
| *** nuritv has joined #openstack-infra | 16:04 | |
| clarkb | mtreinish: no you have always been compliant (yay for simple versions) | 16:04 |
| mtreinish | mriedem: yes, they should all be 1GB | 16:04 |
| anteaya | omrim: this group doesn't tend to make good use of email space, so discussing it at the meeting is a good way to learn how to maximize email real estate | 16:04 |
| anteaya | omrim: otherwise you just make noise that other people ignore | 16:04 |
| mtreinish | mriedem: and I don't think we have 24 going at one time in the gate which is why I think we can drop that number | 16:04 |
| mtreinish | clarkb: ok cool | 16:05 |
| mtreinish | mriedem: to be fair IIRC there is a tempest config flag for default volume size if you wanted to make larger ones | 16:05 |
| *** ayoung has joined #openstack-infra | 16:05 | |
| *** nuritv has quit IRC | 16:05 | |
| krtaylor | omrim, are you talking about a FAQ? | 16:06 |
| omrim | krtaylor: Yes | 16:06 |
| mriedem | mtreinish: right, that's what i was thinking, with 8 VCPU you are just running 4 workers concurrently with testr right? | 16:06 |
| mriedem | so you could be creating 4 1GB volumes at once | 16:06 |
| fungi | clarkb: that looks right to me | 16:06 |
| mriedem | i guess the 24GB is buffer for volumes that were not cleaned up properly | 16:06 |
| mtreinish | mriedem: well it might be more than that, depending on how many volumes are created during a test class | 16:07 |
| mriedem | sure | 16:07 |
| gilliard | Which repo holds the sources for the nova API documentation? | 16:07 |
| mriedem | i don't think i've ever seen a test run more than 3 volumes though | 16:07 |
| mriedem | or 3 instances for that matter | 16:07 |
| clarkb | fungi: was your email sent yet? I may just be blind | 16:07 |
| fungi | clarkb: i'm sending it now--just amended it with a link to your zuul config change | 16:08 |
| clarkb | fungi: cool | 16:08 |
| mtreinish | mriedem: but you're right, the 24g is the is just a safety buffer. Before we did some cleanups on the volume creations we consistently jumped above 15GB which was the previous size iiirc | 16:08 |
| krtaylor | omrim, that would be a big help, maybe a section in th new third-party docs? | 16:08 |
| *** nuritv has joined #openstack-infra | 16:08 | |
| anteaya | krtaylor: do you read the logs of the other third party meetings? | 16:08 |
| *** bdpayne has joined #openstack-infra | 16:08 | |
| krtaylor | anteaya, yes | 16:08 |
| dteselkin | Hi, I have question about pip, again :) | 16:09 |
| dteselkin | Is it worth to implement https support on pypi node? | 16:09 |
| omrim | krtaylor: I already add a patch thank you.. | 16:09 |
| anteaya | then you would know what conversation took place already between omrim and myself | 16:09 |
| krtaylor | omrim, great! can you add that to the etherpad? | 16:09 |
| krtaylor | anteaya, yes, thats why I guessed FAQ | 16:09 |
| dteselkin | I made that in our infra lab, but don't know should I send a patch. | 16:09 |
| mtreinish | mriedem: https://review.openstack.org/#/c/35129/ | 16:09 |
| *** nuritv has quit IRC | 16:10 | |
| clarkb | dteselkin: we have avoided it because we have region specific mirrors and would either need to manage valid certs or have a CA of some sort | 16:10 |
| clarkb | dteselkin: I think if you pushed a patch that at least made it optional that would be great | 16:10 |
| mriedem | mtreinish: thanks | 16:10 |
| clarkb | dteselkin: then if/when we move to https it will be easy for us to do so | 16:10 |
| *** pcrews has joined #openstack-infra | 16:10 | |
| krtaylor | omrim, there is a Reviews section to the etherpad | 16:10 |
| mtreinish | I guess it was never 15G... :) | 16:11 |
| dteselkin | clarkb, ok, then I'll do it. Thanks! | 16:11 |
| mtreinish | mriedem: if you want to push a patch to try and drop it, I think I'd get behind that | 16:11 |
| clarkb | dteselkin: thank you | 16:11 |
| krtaylor | omrim, I was going to discuss it at todays meeting, but the current thought is to have a virtual sprint in the new year | 16:11 |
| mriedem | mtreinish: with 80GB-200GB of disk on the VMs, it's probably not necessary | 16:11 |
| anteaya | krtaylor: what is the link to this etherpad? | 16:11 |
| krtaylor | anteaya, in the email, let me get it one sec | 16:12 |
| *** dannywilson has joined #openstack-infra | 16:12 | |
| anteaya | yes posting it helps people discover it | 16:12 |
| omrim | krtaylor: If FAQ will be helpfull there I will be glad to add it..Anyway I am not familier with Etherpad | 16:12 |
| anteaya | if you want them to add to it | 16:12 |
| krtaylor | anteaya, it is also in the agenda for today's meeting | 16:12 |
| krtaylor | https://etherpad.openstack.org/p/third-party-ci-documentation | 16:12 |
| krtaylor | anteaya, it has been there for a couple of weeks, in addition to the email | 16:13 |
| krtaylor | asselin has signed up, and others have expressed interest | 16:13 |
| anteaya | good, thank you for posting it | 16:14 |
| krtaylor | sure, np | 16:14 |
| anteaya | and I hope it happens | 16:14 |
| krtaylor | yeah, it will, just tough this time of the year | 16:14 |
| *** nuritv has joined #openstack-infra | 16:15 | |
| *** yfried has quit IRC | 16:15 | |
| *** vigneshvar_ has joined #openstack-infra | 16:15 | |
| omrim | anteaya: Can we also discuss about it at tomorrow meeting? (8:00UTC) | 16:16 |
| *** ala_ has quit IRC | 16:16 | |
| *** nuritv has quit IRC | 16:16 | |
| anteaya | omrim: of course | 16:17 |
| anteaya | that is the point of the meetings | 16:18 |
| *** nuritv has joined #openstack-infra | 16:18 | |
| fungi | and so it begins... https://mail.python.org/pipermail/distutils-sig/2014-December/025371.html | 16:18 |
| omrim | anteaya: Thanks! | 16:18 |
| *** mjturek has left #openstack-infra | 16:19 | |
| *** mjturek has joined #openstack-infra | 16:20 | |
| anteaya | fungi: glad the soaps don't fail to deliver | 16:20 |
| *** nuritv has quit IRC | 16:20 | |
| *** jpich has quit IRC | 16:20 | |
| *** yamamoto has quit IRC | 16:20 | |
| *** skolekonov has quit IRC | 16:22 | |
| adam_g | clarkb, im here now, did ninjas sort everything out? | 16:22 |
| clarkb | adam_g: yes the mtreinsh ninja had you covered | 16:23 |
| *** nuritv has joined #openstack-infra | 16:23 | |
| adam_g | mtreinish, thanks for that | 16:23 |
| *** yamamoto has joined #openstack-infra | 16:23 | |
| *** nuritv has quit IRC | 16:24 | |
| *** nuritv has joined #openstack-infra | 16:27 | |
| clarkb | looks like nova is broken too in other ways | 16:27 |
| *** nuritv has quit IRC | 16:29 | |
| *** markmcclain has joined #openstack-infra | 16:30 | |
| *** ddieterly has joined #openstack-infra | 16:31 | |
| *** packet has joined #openstack-infra | 16:31 | |
| *** ddieterly has quit IRC | 16:31 | |
| *** ddieterly has joined #openstack-infra | 16:32 | |
| fungi | oh neat | 16:32 |
| *** annegent_ has quit IRC | 16:33 | |
| *** nuritv has joined #openstack-infra | 16:33 | |
| *** annegent_ has joined #openstack-infra | 16:33 | |
| *** david-lyle_afk is now known as david-lyle | 16:34 | |
| clarkb | adam_g reports that 8.0.2 should fix nova | 16:34 |
| clarkb | so likely an out of date mirror | 16:34 |
| *** yamamoto has quit IRC | 16:34 | |
| fungi | checking now | 16:34 |
| adam_g | according to https://review.openstack.org/#/c/141654/ at least | 16:35 |
| *** atiwari has joined #openstack-infra | 16:35 | |
| *** nuritv has quit IRC | 16:35 | |
| fungi | yeah, i see mirrors have only 8.0.1 | 16:35 |
| fungi | no 8.0.2 yet | 16:35 |
| fungi | looking into why | 16:36 |
| *** liusheng has quit IRC | 16:36 | |
| fungi | oh, dfw and ord are missing it but iad has it. fun! | 16:36 |
| *** liusheng has joined #openstack-infra | 16:37 | |
| *** dimtruck is now known as zz_dimtruck | 16:37 | |
| fungi | also region-b.geo-1 has 8.0.2 | 16:37 |
| fungi | time to look at logs | 16:37 |
| *** achanda has quit IRC | 16:37 | |
| *** nuritv has joined #openstack-infra | 16:38 | |
| *** achanda has joined #openstack-infra | 16:38 | |
| *** yamamoto has joined #openstack-infra | 16:38 | |
| *** reed has joined #openstack-infra | 16:39 | |
| *** nuritv has quit IRC | 16:39 | |
| fungi | looks like maybe the refresh on dfw and ord got prematurely aborted and now the cron job is trying to complete it but falling prey to the 30-minute timeout | 16:40 |
| fungi | i'll get those sync'd up | 16:40 |
| openstackgerrit | Matt Riedemann proposed openstack-infra/elastic-recheck: Add query for setuptools 8/glance bug 1402747 https://review.openstack.org/141835 | 16:40 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [Undecided,New] https://launchpad.net/bugs/1402747 | 16:40 |
| *** nuritv has joined #openstack-infra | 16:40 | |
| jeblair | fungi: (not that it's soon enough to help here, but i made some progress on bandersnatch + afs friday) | 16:41 |
| fungi | excellent! | 16:41 |
| *** achanda has quit IRC | 16:42 | |
| *** nuritv has quit IRC | 16:42 | |
| *** achanda has joined #openstack-infra | 16:42 | |
| *** liusheng has quit IRC | 16:43 | |
| *** mugsie has quit IRC | 16:43 | |
| *** liusheng has joined #openstack-infra | 16:43 | |
| *** KanagarajM has joined #openstack-infra | 16:44 | |
| *** yfried has joined #openstack-infra | 16:46 | |
| *** yaguang has quit IRC | 16:46 | |
| *** yamamoto has quit IRC | 16:47 | |
| openstackgerrit | Merged openstack-infra/elastic-recheck: Add query for setuptools 8/glance bug 1402747 https://review.openstack.org/141835 | 16:47 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [Undecided,New] https://launchpad.net/bugs/1402747 | 16:47 |
| openstackgerrit | Matt Riedemann proposed openstack-infra/elastic-recheck: Add query for nova setuptools 8 bug 1402751 https://review.openstack.org/141840 | 16:48 |
| uvirtbot | Launchpad bug 1402751 in nova "gate-nova-python27 fails with "TypeError: 'SetuptoolsVersion' object does not support indexing" after setuptools 8" [Undecided,New] https://launchpad.net/bugs/1402751 | 16:48 |
| *** yamamoto has joined #openstack-infra | 16:49 | |
| pleia2 | good morning | 16:50 |
| *** mugsie has joined #openstack-infra | 16:50 | |
| *** _sweston is now known as sweston | 16:51 | |
| bookwar | how do you version JJB templates? let's say there was a job for icehouse made from template, then there is a new job for Juno made from the same template, and then you need to make certain change in this Juno job, which is incompatible with icehouse. Do you "freeze" icehouse jobs after release somehow? | 16:51 |
| *** dtantsur is now known as dtantsur|afk | 16:53 | |
| fungi | bookwar: we take extra effort to make sure that job templates are applicable across multiple branches, and implement conditional switches in them where needed to perform branch-specific steps | 16:53 |
| *** armax has joined #openstack-infra | 16:53 | |
| fungi | bookwar: since the branch is passed into the job through an environment variable, you can match on that to identify which commands should be run | 16:53 |
| bookwar | fungi: i see, thanks | 16:54 |
| dstufft | fungi: was the bandersnatch thing a bandersnatch problem? | 16:54 |
| dstufft | or something else | 16:54 |
| dims | dstufft: thanks for the quick response on the Version/index problem over the weekend | 16:54 |
| *** yamamoto has quit IRC | 16:54 | |
| *** nuritv has joined #openstack-infra | 16:55 | |
| dstufft | dims: no problem, Jason made the release I just wrote the patch :) | 16:55 |
| dims | :) | 16:55 |
| dstufft | one benefit to releasing over the weekend is that less people means less people yelling at us while we fix any issues :D | 16:55 |
| dims | hehe | 16:55 |
| fungi | dstufft: i think the bandersnatch thing was possibly me leaving the sync running under a timeout (at least i haven't ruled out the possibility) | 16:56 |
| *** arxcruz has quit IRC | 16:56 | |
| openstackgerrit | Merged openstack-infra/elastic-recheck: Add query for nova setuptools 8 bug 1402751 https://review.openstack.org/141840 | 16:56 |
| uvirtbot | Launchpad bug 1402751 in nova "gate-nova-python27 fails with "TypeError: 'SetuptoolsVersion' object does not support indexing" after setuptools 8" [Critical,Confirmed] https://launchpad.net/bugs/1402751 | 16:56 |
| *** KanagarajM has quit IRC | 16:56 | |
| *** Masahiro has joined #openstack-infra | 16:56 | |
| *** nuritv has quit IRC | 16:57 | |
| dstufft | fungi: ok | 16:57 |
| *** otter768 has joined #openstack-infra | 16:57 | |
| *** pblaho has quit IRC | 16:57 | |
| dstufft | fungi: just wanted to make sure we didn't break something else :) | 16:57 |
| fungi | dstufft: i don't think so | 16:58 |
| * ttx blames dstufft for the world being on fire and hits him with a trout | 16:58 | |
| fungi | a fire extinguishing trout! | 16:59 |
| *** k4n0 has quit IRC | 16:59 | |
| *** sarob has joined #openstack-infra | 16:59 | |
| *** barra204_ has quit IRC | 16:59 | |
| *** nuritv has joined #openstack-infra | 16:59 | |
| dstufft | https://www.youtube.com/watch?v=n9D6GNzpmkM | 17:00 |
| *** MaxV has quit IRC | 17:00 | |
| *** MaxV has joined #openstack-infra | 17:01 | |
| *** nuritv has quit IRC | 17:01 | |
| *** Masahiro has quit IRC | 17:01 | |
| *** zz_dimtruck is now known as dimtruck | 17:01 | |
| *** otter768 has quit IRC | 17:01 | |
| *** nuritv has joined #openstack-infra | 17:03 | |
| *** shakamunyi has quit IRC | 17:03 | |
| *** kmartin has joined #openstack-infra | 17:04 | |
| anteaya | morning pleia2 | 17:04 |
| *** nuritv has quit IRC | 17:05 | |
| *** MaxV has quit IRC | 17:05 | |
| *** gyee has joined #openstack-infra | 17:06 | |
| *** kmartin has quit IRC | 17:08 | |
| *** nuritv has joined #openstack-infra | 17:08 | |
| *** annegent_ has quit IRC | 17:09 | |
| openstackgerrit | Matt Riedemann proposed openstack-infra/elastic-recheck: Add query for pbr bug 1402759 https://review.openstack.org/141848 | 17:09 |
| uvirtbot | Launchpad bug 1402759 in pbr "gate-neutron-python27 fails with "ValueError: ("Expected ',' or end-of-list in", u'neutron==2015.1"" [Critical,Fix released] https://launchpad.net/bugs/1402759 | 17:09 |
| *** zz_avozza is now known as avozza | 17:09 | |
| *** jgallard_ has quit IRC | 17:10 | |
| *** nuritv has quit IRC | 17:10 | |
| *** mmaglana has joined #openstack-infra | 17:11 | |
| notmyname | ttx: just got in and catching up. what's the current state of things? | 17:11 |
| anteaya | dstufft: I'm reading https://caremad.io/2013/07/setup-vs-requirement/ is it still accurate? | 17:12 |
| *** ihrachyshka has quit IRC | 17:13 | |
| ttx | notmyname: that email I sent one hour ago is pretty accurate | 17:13 |
| notmyname | ttx: ok, thanks. just getting to email now (IRC first!) | 17:13 |
| *** hdd has joined #openstack-infra | 17:13 | |
| ttx | about to leave, so let me know if that email is unclear. Or I can parrot it now :) | 17:13 |
| *** nuritv has joined #openstack-infra | 17:14 | |
| notmyname | ttx: just read it. the question is what to do with the rc tag | 17:14 |
| dstufft | anteaya: it's accurate as to how pip treats the two different files yes, and my thinking on the matter | 17:14 |
| notmyname | ttx: meh. leave it. no harm done with it there | 17:14 |
| dstufft | anteaya: obviously openstack doesn't agree with me/us since it parses the requirements.txt files ;) | 17:15 |
| notmyname | ttx: also, I'll send out an announce shortly to -dev | 17:15 |
| ttx | notmyname: alright! | 17:15 |
| notmyname | ttx: thanks for chasing that | 17:15 |
| ttx | notmyname: welcome to the c1 new world order! | 17:15 |
| notmyname | ttx: fungi: final question is if "rc" is done for all python things or just openstack things | 17:16 |
| *** nuritv has quit IRC | 17:16 | |
| clarkb | anything that uses setuptools | 17:16 |
| notmyname | clarkb: ok | 17:16 |
| dstufft | setuptools normalizes | 17:16 |
| ttx | notmyname: yes, it's PEP440 | 17:17 |
| clarkb | dstufft yes thats the issue | 17:17 |
| clarkb | dstufft we tag one thibg setuptools spits out another | 17:17 |
| dstufft | clarkb: sorry I'm running on a few hours of asleep and I'm tired so I'm a little slow today | 17:17 |
| clarkb | so in an effort to reduce confusion we use c1 | 17:17 |
| fungi | notmyname: setuptools 8 and anything else implementing pep 440 no longer sees "rc" as normalized | 17:17 |
| notmyname | ok | 17:17 |
| *** penguinRaider has quit IRC | 17:17 | |
| clarkb | dsufft we want ag version to match sdist version. just requires us to use c | 17:18 |
| clarkb | *tag version | 17:18 |
| fungi | dstufft: right, the new issue we identified is that since we rely on setuptools parsing to determine sdist tarball and wheel names, we need to use already-normalized versions in tags | 17:18 |
| jd__ | tooz is still failing for an unknown reason, hint appreciated http://logs.openstack.org/14/141314/3/check//gate-tempest-dsvm-neutron-src-tooz/5c3aee9/ :( | 17:19 |
| jd__ | cc dhellmann who might have a clue | 17:19 |
| *** armax has quit IRC | 17:19 | |
| *** kmartin has joined #openstack-infra | 17:20 | |
| *** amotoki has quit IRC | 17:20 | |
| dstufft | fungi: clarkb ah ok, sorry about that | 17:20 |
| fungi | jd__: "'pymemcache' is not in global-requirements.txt" http://logs.openstack.org/14/141314/3/check/gate-tempest-dsvm-neutron-src-tooz/5c3aee9/logs/devstacklog.txt.gz#_2014-12-15_16_18_14_298 | 17:20 |
| *** sputnik13 has joined #openstack-infra | 17:20 | |
| *** armax has joined #openstack-infra | 17:20 | |
| jd__ | fungi: -_- | 17:20 |
| *** armax has quit IRC | 17:21 | |
| jd__ | that will teach me not checking the log before pasting an error | 17:21 |
| anteaya | dstufft: very good, thank you | 17:21 |
| jd__ | we used to have a different error but I guess we ended up fixing it with the devstack-gate patch and I thought it will solve everything, it turns out that no, we still miss a patch | 17:21 |
| jd__ | :) | 17:21 |
| dstufft | anteaya: for the record, I think openstack should just add the requirements to setup.cfg to match what they're already doing... but I don't have the time to argue for or make that change myself so :) | 17:22 |
| *** AJaeger has quit IRC | 17:22 | |
| *** nuritv has joined #openstack-infra | 17:22 | |
| *** armax has joined #openstack-infra | 17:22 | |
| openstackgerrit | Matt Riedemann proposed openstack-infra/elastic-recheck: Update query for bug 1402747 https://review.openstack.org/141852 | 17:22 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [High,Confirmed] https://launchpad.net/bugs/1402747 | 17:23 |
| *** markmcclain has quit IRC | 17:23 | |
| *** markmcclain has joined #openstack-infra | 17:23 | |
| anteaya | dstufft: yes, the reason I ask is go get my head around requirements myself, it is a bit of work indeed | 17:23 |
| fungi | dstufft: can we do multiple sections that way for situation-specific dependencies? (test-only requirements, interpreter-version-specific requirements, et cetera)? | 17:24 |
| *** cnesa2 has quit IRC | 17:24 | |
| *** nuritv has quit IRC | 17:25 | |
| sweston | anteaya: Good morning! I am looking over the latest iteration of comments on the third-party dashboard, and could really use your input. https://review.openstack.org/#/c/135170 | 17:25 |
| *** armax has quit IRC | 17:25 | |
| *** hashar has quit IRC | 17:25 | |
| *** SumitNaiksatam has quit IRC | 17:26 | |
| dstufft | fungi: well pbr would be the one translating it into setup.py key arguments like you're doing now, so you can do it however you want | 17:26 |
| zaro | morning | 17:26 |
| *** ZZelle is now known as ZZelle__ | 17:26 | |
| anteaya | morning zaro | 17:26 |
| *** armax has joined #openstack-infra | 17:26 | |
| *** markvan has quit IRC | 17:26 | |
| *** nuritv has joined #openstack-infra | 17:27 | |
| *** asselin has joined #openstack-infra | 17:27 | |
| fungi | dstufft: good point. so in theory yes we could enumerate multiple requirements scenarios in one setup.cfg and drop the requirements files | 17:27 |
| *** Ryan_Lane has joined #openstack-infra | 17:27 | |
| *** markvan has joined #openstack-infra | 17:27 | |
| dstufft | fungi: you'd of course be limited to whatever setuptools supports, but you're limited to that right onw anyways | 17:27 |
| anteaya | sweston: I've rechecked for a fresh set of draft docs | 17:27 |
| *** markmcclain has quit IRC | 17:27 | |
| *** nuritv has quit IRC | 17:28 | |
| *** avozza is now known as zz_avozza | 17:28 | |
| *** zz_avozza is now known as avozza | 17:29 | |
| zaro | pelix: i think you might be interested in this to speeed up jjb: https://review.openstack.org/#/c/138886/7/jenkins/__init__.py | 17:29 |
| sweston | anteaya: ok. where are the docs produced? | 17:29 |
| anteaya | there is a link in the returned jenkins results | 17:29 |
| *** tonytan4ever has quit IRC | 17:30 | |
| anteaya | but they only last a certain period of time (7 days? 10 days? I don't know) | 17:30 |
| dstufft | fungi: when setuptools 8 is working in openstack and you're unpinned, it'd be useful I think to test pip dev branch against openstack before we release pip to see if there's anything in there that's going to break the world for y'all before it gets released | 17:30 |
| anteaya | so running recheck gets a fresh set of them | 17:30 |
| *** nuritv has joined #openstack-infra | 17:30 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/infra-specs: Update the test-metrics-db spec to reflect recent changes https://review.openstack.org/141855 | 17:31 |
| sweston | I see. clicking on the link right now produces a not found. does it take some time to produce them? | 17:31 |
| fungi | dstufft: "test pip dev branch" is i think still not something we entirely understand how to accomplish without building an entirely parallel duplicate of our infrastructure | 17:31 |
| dstufft | fungi: ok | 17:31 |
| openstackgerrit | Merged openstack-infra/elastic-recheck: Add query for pbr bug 1402759 https://review.openstack.org/141848 | 17:31 |
| uvirtbot | Launchpad bug 1402759 in pbr "gate-neutron-python27 fails with "ValueError: ("Expected ',' or end-of-list in", u'neutron==2015.1"" [Critical,Fix released] https://launchpad.net/bugs/1402759 | 17:31 |
| clarkb | we can monkey patch it with devstack easily enough | 17:31 |
| clarkb | just have devstack grab prerelease pip | 17:31 |
| fungi | dstufft: there are probably some things we can test against it, but its large and interdependent nature means we won't be able to thoroughly confirm it everywhere | 17:32 |
| *** e0ne_ is now known as e0ne | 17:32 | |
| *** nuritv has quit IRC | 17:32 | |
| anteaya | sweston: yes, jenkins jobs need to finish running on the recheck | 17:32 |
| anteaya | sweston: http://status.openstack.org/zuul/ filter on 135170 | 17:33 |
| *** jkraj has joined #openstack-infra | 17:33 | |
| anteaya | sweston: you will see the progress of the jenkins jobs | 17:34 |
| sweston | anteaya: thank you | 17:34 |
| anteaya | welcome | 17:34 |
| dstufft | fungi: yea well, it's packaging tooling, the only real test is to release it i've found :D but yea throwing it at devstack would at least give an indication I hope | 17:34 |
| anteaya | do share that information please | 17:34 |
| anteaya | sweston: I didn't know you didn't know that | 17:34 |
| anteaya | it is helpful information to have | 17:34 |
| mattfarina | can someone point me to what happened to pypi.openstack.org? | 17:34 |
| fungi | mattfarina: we stopped using it and took it down, why? | 17:35 |
| *** AJaeger has joined #openstack-infra | 17:35 | |
| sweston | it certainly is! is this information in the docs anywhere? | 17:35 |
| *** AJaeger has quit IRC | 17:35 | |
| *** AJaeger has joined #openstack-infra | 17:35 | |
| mattfarina | fungi i just found a couple referenes to it and i was curious what happened. that's all. thanks | 17:35 |
| *** dkranz has quit IRC | 17:35 | |
| fungi | mattfarina: we run separate pypi mirrors local to each provider/region where we run jobs now | 17:35 |
| *** nuritv has joined #openstack-infra | 17:35 | |
| *** dkranz has joined #openstack-infra | 17:36 | |
| *** markmcclain has joined #openstack-infra | 17:36 | |
| anteaya | sweston: if not it should be, I smell an opportunity for you to write a patch :D | 17:36 |
| *** Ryan_Lane has quit IRC | 17:37 | |
| dstufft | fungi: clarkb one thing that I know will affect you, is pip 6 deprecates (but doesn't remove!) accessing non HTTPS indexes without passing a flag (techincally non secure origins as defined by Chrome), it won't be removed until pip 7 or 8 and there's a flag (and env var, and config file) you can set to allow it | 17:37 |
| *** shashankhegde has joined #openstack-infra | 17:37 | |
| sweston | anteaya: hehe, yes 😃 I will look into it today. | 17:38 |
| fungi | dstufft: is there a good way to add trusted snakeoil server certs pip will accept? | 17:38 |
| *** nuritv has quit IRC | 17:38 | |
| fungi | dstufft: server cert pinning basically? | 17:38 |
| dstufft | fungi: every version of pip that verifies TLS has a config option for an alternate CA bundle, so as most pip options goes there is a config file, env var, cli flag | 17:39 |
| sweston | anteaya: what do you think of having the third party systems check into the dashboard, as jhesketh suggests in the review? | 17:39 |
| dstufft | fungi: pip 6+ also now will default to using the system CA bundle if it can find one and only falls back to it's bundled one if it can't, so adding it to the system should work too on pip 6+ | 17:39 |
| fungi | dstufft: does that work okay for self-signed certs? | 17:39 |
| dstufft | fungi: yes | 17:39 |
| dstufft | there's nothing special about CA signed certs except they are in the default CA bundle | 17:39 |
| *** signed8bit has quit IRC | 17:40 | |
| dstufft | fungi: oh yea, pip 6+ also adds a system wide config file in /etc/, and a virtualenv specific config file | 17:40 |
| dstufft | not sur eif that's useful for y'all | 17:40 |
| krtaylor | sweston, duncant, mriedem and asselin would be good for input on the dashboard also | 17:41 |
| anteaya | sweston: let me get to that place in the review and then either comment or get back to you | 17:41 |
| sweston | anteaya: ok. | 17:41 |
| sweston | krtaylor: thank you | 17:42 |
| fungi | dstufft: thanks--i'll see about getting snakeoil https going on our mirrors in preparation | 17:42 |
| dstufft | fungi: pip 6+ also adds a flag that just lets you disable TLS verification too | 17:42 |
| dstufft | though maybe that should be wrapped into the flag that lets you specify a certain host doesn't need TLS | 17:42 |
| sweston | pleia2: ^ might have some comments as well | 17:43 |
| *** jcoufal_ has quit IRC | 17:43 | |
| *** sputnik13 has quit IRC | 17:43 | |
| *** _nadya_ has quit IRC | 17:44 | |
| *** wenlock1 has joined #openstack-infra | 17:44 | |
| *** nuritv has joined #openstack-infra | 17:44 | |
| *** _nadya_ has joined #openstack-infra | 17:44 | |
| krtaylor | sweston, I can add them as reviewers if you wish | 17:44 |
| *** jp_at_hp has quit IRC | 17:45 | |
| *** harlowja_at_home has joined #openstack-infra | 17:45 | |
| *** achanda has quit IRC | 17:45 | |
| sweston | krtaylor: yes, please | 17:45 |
| *** rkukura_ has joined #openstack-infra | 17:45 | |
| clarkb | will pip6 accept + ? | 17:45 |
| *** achanda has joined #openstack-infra | 17:45 | |
| openstackgerrit | Merged openstack-infra/elastic-recheck: Update query for bug 1402747 https://review.openstack.org/141852 | 17:45 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [High,Confirmed] https://launchpad.net/bugs/1402747 | 17:45 |
| *** nuritv has quit IRC | 17:46 | |
| dstufft | clarkb: in a version? | 17:46 |
| *** achanda has quit IRC | 17:47 | |
| *** achanda has joined #openstack-infra | 17:47 | |
| *** rkukura has quit IRC | 17:47 | |
| *** rkukura_ is now known as rkukura | 17:47 | |
| *** _nadya_ has quit IRC | 17:49 | |
| *** sputnik13 has joined #openstack-infra | 17:49 | |
| *** _nadya_ has joined #openstack-infra | 17:50 | |
| *** sputnik13 has quit IRC | 17:50 | |
| clarkb | ya | 17:50 |
| *** bhunter71 has joined #openstack-infra | 17:50 | |
| clarkb | since current doesnt via vendored setuptools aiui | 17:50 |
| pelix | zaro: speed up jjb? looks like https://review.openstack.org/#/c/138886/7 is about correct encoding for spaces in job names? | 17:51 |
| *** wenlock1 has quit IRC | 17:51 | |
| zaro | pelix: take a look at the comment/suggestion by kevin. | 17:51 |
| dstufft | clarkb: yea, whatever the latest setuptools is when 6.0 is released is what will be bundled into pip | 17:51 |
| dstufft | clarkb: which means 8.0.2 right now, so + should be supported | 17:52 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config: Normalize ACLs https://review.openstack.org/141860 | 17:52 |
| *** hdd has quit IRC | 17:52 | |
| dstufft | clarkb: if for whatever reason something in PEP 440 isn't supported in pip we'd probably fix it and issue a new release asap | 17:52 |
| *** harlowja_at_home has quit IRC | 17:53 | |
| openstackgerrit | Khai Do proposed openstack-infra/jenkins-job-builder: deprecate postbuildscript onsuccess and onfailure parameter names https://review.openstack.org/139257 | 17:53 |
| *** sweston is now known as _sweston | 17:54 | |
| pelix | zaro: what he needs is for us to get https://review.openstack.org/#/c/75514/ sorted and landed | 17:54 |
| *** nuritv has joined #openstack-infra | 17:55 | |
| pelix | zaro: or https://review.openstack.org/#/c/45371/ either should do the job | 17:55 |
| *** yamamoto has joined #openstack-infra | 17:55 | |
| pleia2 | _sweston: thanks for the ping, I held off on further review of that patch after jhesketh's comment because I'm really not sure | 17:56 |
| *** sputnik13 has joined #openstack-infra | 17:56 | |
| *** yfried is now known as yfried|afk | 17:57 | |
| pleia2 | anteaya: can you look at https://review.openstack.org/#/c/141817/ when you have a chance? the patch needs a lot of work, but I'm not sure what to suggest re: formatting (and if a FAQ is appropriate at all), it seems something a lot of operators struggle with, so it's probably useful in some form | 17:57 |
| *** nuritv has quit IRC | 17:57 | |
| *** e0ne has quit IRC | 17:58 | |
| zaro | pelix: so parrallel would be nice but might not be needed if python-jenkins library was faster? | 17:58 |
| *** sputnik13 has quit IRC | 17:58 | |
| *** nuritv has joined #openstack-infra | 17:59 | |
| *** fandi has quit IRC | 17:59 | |
| pelix | zaro: unlikely, the main delay is latency of requests | 17:59 |
| zaro | pelix: it seems we should make the fix to the underlying library first to see how much fater it will work? | 17:59 |
| *** yamamoto has quit IRC | 18:00 | |
| zaro | pelix: so kevin's suggestion will reduce the number of requests by what looks like quite a bit. | 18:00 |
| *** nuritv has quit IRC | 18:01 | |
| pelix | it'll help, but not much once you have lots of jobs think +2000 | 18:01 |
| *** _sweston is now known as sweston | 18:01 | |
| *** shashankhegde has quit IRC | 18:01 | |
| pelix | his only helps for the rename, one of the biggest delays is the attempt to retrieve the XML from the server if the job exists | 18:01 |
| *** signed8bit has joined #openstack-infra | 18:02 | |
| dstufft | fungi: so far one thread on distutils-sig about setuptools 8 with only two people other than me. MAybe popcorn won't be needed *knocks on wood* | 18:02 |
| *** _nadya_ has quit IRC | 18:02 | |
| pelix | think both of the parallel approaches discard this with the realization that if the cache is empty, simply uploading the job and letting jenkins work out if the job is the same or not is much quicker | 18:03 |
| anteaya | pleia2: sure that came out of last tuesdays third party meeting if you wanted to skim the logs for context, I'll review after I finish reviewing sweston's patch | 18:03 |
| sweston | pleia2: ok, I am responding to it today. am providing an update in the third party meeting soon if you want to discuss it further. | 18:03 |
| fungi | dstufft: very promising! i think it's gone surprisingly smoothly, really | 18:03 |
| pelix | zaro: difference between 2hrs+ and 3-4 minutes to upload 2000 jobs | 18:03 |
| fungi | dstufft: thanks in no small part to your assistance all weekend. thanks again | 18:04 |
| *** derekh has quit IRC | 18:04 | |
| pleia2 | sweston: thanks | 18:04 |
| dstufft | fungi: considering we changes some very old and very core code to setuptools that a lot of people had cause to apply some sort of work around over the years, I'm pretty happy with how it's gone so far | 18:04 |
| *** cnesa has joined #openstack-infra | 18:04 | |
| stevemar | is there a bug that is tracking the setuptools<0.8 work? | 18:04 |
| dstufft | fungi: thanks :) least I could do heh after breaking things ;) | 18:04 |
| clarkb | mriedem has a few now for specific behaviors | 18:05 |
| clarkb | stevemar ^ | 18:05 |
| fungi | stevemar: it's really lots of different issues, just all exposed by a common event, but being worked more or less separately in parallel | 18:06 |
| *** whayutin_ has joined #openstack-infra | 18:06 | |
| *** weshay has quit IRC | 18:06 | |
| *** nuritv has joined #openstack-infra | 18:06 | |
| *** subscope has quit IRC | 18:07 | |
| *** annegent_ has joined #openstack-infra | 18:07 | |
| clarkb | at this point we should try to get the pbr change in that drops git shas from versions, release pbr, then propose revert of setuptools pin | 18:07 |
| clarkb | as oslo.db is released now iirc | 18:07 |
| dstufft | yea I saw dhellmann say that on the ML | 18:07 |
| *** sputnik13 has joined #openstack-infra | 18:08 | |
| clarkb | the revert of the pin should self test | 18:08 |
| *** nuritv has quit IRC | 18:08 | |
| clarkb | I would help but need to afk for much of the day now. fwiw I think mordreds pbr change looked good just was unable to review and test properly | 18:08 |
| anteaya | omrim: so Signed-off-by is used in the commit message for https://review.openstack.org/#/c/141817/ yet you are the owner | 18:09 |
| dstufft | then we'll quick release pip 6.0 and break it all over again! (just kidding, I don't suspect anything in there is going to break anything further thans etuptools pep 440ization did) | 18:09 |
| anteaya | omrim: what is meant by your use of Signed-off-by? | 18:09 |
| *** rustlebee is now known as russellb | 18:09 | |
| *** sputnik13 has quit IRC | 18:09 | |
| clarkb | well we need to get https dealt with but that sounds straightforward | 18:09 |
| clarkb | anteaya so that the info is in git not gerrit | 18:10 |
| *** patrickeast has joined #openstack-infra | 18:10 | |
| *** hdd has joined #openstack-infra | 18:10 | |
| dstufft | clarkb: that's not changed in 6 other than 6 warns that in the future it'll break | 18:10 |
| dstufft | so you have time for that | 18:10 |
| *** tjones1 has joined #openstack-infra | 18:11 | |
| clarkb | ah I misread I thought behavior changed with toggle to go back | 18:11 |
| *** nuritv has joined #openstack-infra | 18:11 | |
| clarkb | thats even better | 18:11 |
| fungi | clarkb: the bandersnatch mirror syncs i have going in dfw and ord need to complete before we see full coverage from the oslo.db release | 18:11 |
| *** sputnik13 has joined #openstack-infra | 18:11 | |
| zaro | pelix: i see your point about latency on retreiving the job xml but the check for existing jobs seems to on every operation so i think it would affect more than just renaming of jobs. | 18:11 |
| clarkb | fungi will need that for the pbr release too | 18:12 |
| dstufft | clarkb: yea, the toggle exists in 6.0, but all it does it shut up the warning since currently there is only a warning | 18:12 |
| clarkb | gotcha | 18:12 |
| anteaya | clarkb: I read the words but I still am lacking context I'll read the gerrit commit message wikipage | 18:12 |
| dstufft | things that the end user needs to change is way easier to deprecate than thing that package author needs to change :( | 18:12 |
| anteaya | pleia2: I agree with all your points, nice review | 18:13 |
| *** nuritv has quit IRC | 18:13 | |
| *** sputnik13 has quit IRC | 18:13 | |
| clarkb | anteaya basically signed off by has nothing to do with gerrit as we use it | 18:13 |
| clarkb | looks like pbr change needs pep8 fixes | 18:14 |
| clarkb | mordred are you aroynd to do that? | 18:14 |
| anteaya | found it it has to do with the dco: https://wiki.openstack.org/wiki/OpenStackAndItsCLA#The_Developer_Certificate_of_Origin | 18:15 |
| anteaya | which to the best of my knowledge we are not using | 18:15 |
| anteaya | perhaps he is just planning ahead | 18:15 |
| anteaya | so no harm to leave it in the commit message then I guess | 18:15 |
| *** marcusvrn has quit IRC | 18:15 | |
| *** sputnik13 has joined #openstack-infra | 18:15 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/infra-specs: Update the test-metrics-db spec to reflect recent changes https://review.openstack.org/141855 | 18:17 |
| *** yfried|afk is now known as yfried | 18:17 | |
| *** tjones1 has left #openstack-infra | 18:18 | |
| *** marcusvrn has joined #openstack-infra | 18:18 | |
| *** nuritv has joined #openstack-infra | 18:18 | |
| fungi | anteaya: some contributors who also contribute to dco-using projects likely have git commit hooks already installed to add that header automatically | 18:19 |
| zaro | pelix: also i think that jjb creating lots of non-existing jobs is the slowness and since python-jenkins will check whether the job already exists before creating a job that's 2 request for every job instead of just 1. | 18:19 |
| *** melwitt has joined #openstack-infra | 18:19 | |
| *** nuritv has quit IRC | 18:20 | |
| anteaya | fungi: ah cool | 18:21 |
| pelix | zaro: I'm guessing we might get a 33% increase in throughput by avoiding that extra round trip, though it might be better to have the jenkins object store the list of jobs internally and for assert_job_exists to retrieve the information from this by default unless passed 'refresh=True'? | 18:22 |
| *** Hal_ has quit IRC | 18:22 | |
| *** nuritv has joined #openstack-infra | 18:22 | |
| *** HeOS has quit IRC | 18:22 | |
| *** Ryan_Lane has joined #openstack-infra | 18:23 | |
| anteaya | sweston: I suggest you talk to emagama in neutron, DuncanT and junglejoyj in cinder, morgainfainberg in keystone and jogo in nova about your third party dashboard spec | 18:23 |
| *** BobBall is now known as BobBall_AWOL | 18:23 | |
| * morganfainberg 's ears perk up. | 18:23 | |
| anteaya | sweston: if you hope to get buy in from the various programs to use it, you will need to talk to them about it | 18:24 |
| anteaya | morganfainberg: hello | 18:24 |
| morganfainberg | anteaya, Hiya! | 18:24 |
| anteaya | sweston: oh and also in ironic, though I don't have a contact name yet | 18:24 |
| *** nuritv has quit IRC | 18:24 | |
| sweston | anteaya: ok, that is excellent advice. | 18:24 |
| anteaya | sweston: you have a nice start | 18:24 |
| anteaya | it will take some revisions and a lot of conversations | 18:25 |
| anteaya | but it would nice to see it happen | 18:25 |
| anteaya | morganfainberg meet sweston | 18:25 |
| anteaya | sweston, morganfainberg | 18:25 |
| morganfainberg | sweston, hi there! | 18:25 |
| sweston | anteaya: thank you. what is the best way to start these discussions? | 18:25 |
| anteaya | attend their meetings | 18:25 |
| sweston | morganfainberg: hello | 18:25 |
| morganfainberg | sweston, yep attend the meetings! | 18:25 |
| morganfainberg | best way to get all the project cores eyes / input on your ideas | 18:26 |
| anteaya | invite them to any third party meeting you plan on attending (don't be surprised if they don't show up) | 18:26 |
| anteaya | you will have better luck if you go to them | 18:26 |
| sweston | morganfainberg: ok | 18:26 |
| morganfainberg | sweston, our meeting (keystone) is tomorrow... | 18:26 |
| jroll | anteaya: what contact do you need from ironic? | 18:26 |
| sweston | should I update the agenda items for this? | 18:26 |
| anteaya | jroll: ah yes | 18:26 |
| morganfainberg | sweston, https://wiki.openstack.org/wiki/Meetings/KeystoneMeeting | 18:27 |
| clarkb | one line docstring needs punctuation? | 18:27 |
| *** prad has joined #openstack-infra | 18:27 | |
| fungi | possibly also getting it onto an upcoming cross-project weekly meeting agenda could help too if this is something which needs cross-project input | 18:27 |
| anteaya | jroll: which ironic core reviewer wants to be point person for third party matters? | 18:27 |
| morganfainberg | sweston, you can put something on the agenda /me checks to make sure it's accurate | 18:27 |
| clarkb | if H402 isnt dead yet it should go on the list | 18:27 |
| anteaya | fungi: good idea | 18:27 |
| sweston | morganfainberg: ok, will do | 18:27 |
| morganfainberg | that agenda might be out-dated | 18:27 |
| *** nuritv has joined #openstack-infra | 18:27 | |
| morganfainberg | we skipped last week's meeting | 18:27 |
| anteaya | sweston: so consider the cross-project meeting as well | 18:27 |
| zaro | pelix: is that a suggestion for have a python-jenkins cache? | 18:27 |
| jroll | anteaya: great question, does it have to be a core? adam_g is our usual testing guru but not core. what are the responsibilities? | 18:27 |
| sweston | fungi: ok, yes thanks | 18:27 |
| mriedem | clarkb: stevemar: bug 1402751 - bug 1402747 - bug 1402759 | 18:27 |
| uvirtbot | Launchpad bug 1402751 in nova "gate-nova-python27 fails with "TypeError: 'SetuptoolsVersion' object does not support indexing" after setuptools 8" [Critical,Confirmed] https://launchpad.net/bugs/1402751 | 18:27 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [High,Confirmed] https://launchpad.net/bugs/1402747 | 18:27 |
| morganfainberg | hm | 18:27 |
| uvirtbot | Launchpad bug 1402759 in pbr "gate-neutron-python27 fails with "ValueError: ("Expected ',' or end-of-list in", u'neutron==2015.1"" [Critical,Fix released] https://launchpad.net/bugs/1402759 | 18:27 |
| anteaya | jroll: the responsibilites is to be responsible | 18:27 |
| sweston | I will attend cross-project as well | 18:27 |
| zaro | pelix: would make sense to me. | 18:28 |
| morganfainberg | looks like it's mostly up to date | 18:28 |
| pelix | zaro: I believe JJB gets the full list of jobs on the server to switch between update versus create calls to python-jenkins | 18:28 |
| jroll | ha | 18:28 |
| anteaya | jroll: doesn't have to be core, if the other cores will listen and back them up | 18:28 |
| morganfainberg | so feel free to add to the end | 18:28 |
| morganfainberg | sweston,^ | 18:28 |
| anteaya | if adam_g wants it, I think other people will listen to him in ironic | 18:28 |
| sweston | morganfainberg: ok, thank you! | 18:28 |
| jroll | adam_g: can I throw you under anteaya's bus? :) | 18:28 |
| jroll | yeah, agree | 18:28 |
| anteaya | jroll: now you are getting the picture | 18:28 |
| adam_g | jroll, sure! :) | 18:28 |
| anteaya | adam_g: congratulations and my condolences | 18:28 |
| jroll | perfect | 18:29 |
| jroll | thanks y'all | 18:29 |
| anteaya | sweston: talk to adam_g about ironic's input on the spec | 18:29 |
| adam_g | anteaya, is there a wiki describing the role i just signed up for? :) | 18:29 |
| *** nuritv has quit IRC | 18:29 | |
| anteaya | not really | 18:29 |
| *** _nadya_ has joined #openstack-infra | 18:29 | |
| pelix | zaro: exactly just the first initialization to retrieve the list of jobs from the server using the 'get_jobs()' method and have assert_job_exists go through that by default unless we need to explicitly refresh. Which should only be true for rename and copy | 18:30 |
| anteaya | since I wanted to use peer pressure and the soft sell | 18:30 |
| clarkb | I will update mordreds patch shortly then afk for real | 18:30 |
| anteaya | adam_g: a wiki will just scare folks away :) | 18:30 |
| sweston | anteaya: neutron meeting is today, should we get some input from mestery as well? | 18:30 |
| *** ivar-lazzaro has joined #openstack-infra | 18:30 | |
| anteaya | sweston: my suggestion is talk to edgar first | 18:30 |
| pelix | zaro: that leaves the ability to error early in place and only adds extra round trips where needed | 18:30 |
| anteaya | he will give you the obvious suggestions and relay to kyle | 18:30 |
| *** e0ne has joined #openstack-infra | 18:31 | |
| sweston | anteaya: ok, will do | 18:31 |
| *** ihrachyshka has joined #openstack-infra | 18:31 | |
| anteaya | kyle is busy and edgar can let him know once the spec suits edgar's needs | 18:31 |
| anteaya | so yes having kyle's input is valuable | 18:31 |
| sweston | anteaya: gotcha | 18:31 |
| adam_g | sweston, whats URL for the spec? | 18:31 |
| anteaya | getting it the right way is suggested | 18:31 |
| *** dimtruck is now known as zz_dimtruck | 18:31 | |
| *** _nadya_ has quit IRC | 18:31 | |
| sweston | adam_g: https://review.openstack.org/#/c/135170/ | 18:31 |
| adam_g | sweston, thanks | 18:32 |
| *** doug-fish has quit IRC | 18:32 | |
| sweston | adam_g: you bet! | 18:32 |
| *** shashankhegde has joined #openstack-infra | 18:32 | |
| *** markmcclain has quit IRC | 18:32 | |
| *** lttrl has quit IRC | 18:32 | |
| sweston | anteaya: agreed. | 18:32 |
| *** koolhead17 has quit IRC | 18:32 | |
| *** harlowja has joined #openstack-infra | 18:33 | |
| *** nuritv has joined #openstack-infra | 18:33 | |
| *** marun has joined #openstack-infra | 18:33 | |
| *** markmcclain has joined #openstack-infra | 18:33 | |
| sweston | the more eyes we can get on this from the beginning, the better | 18:33 |
| *** sdake has quit IRC | 18:33 | |
| *** sarob has quit IRC | 18:33 | |
| *** dpaterson has quit IRC | 18:34 | |
| sweston | anteaya: regarding the stackforge repo, I have sent an email to mikal but so far no response | 18:34 |
| *** nuritv has quit IRC | 18:34 | |
| anteaya | sweston: hmmmmm | 18:35 |
| *** _nadya_ has joined #openstack-infra | 18:35 | |
| anteaya | sweston: okay I can follow up with jhesketh | 18:35 |
| anteaya | hopefully he will be at the tuesday 0800 utc meeting | 18:35 |
| anteaya | thanks for letting me know | 18:35 |
| *** sdake has joined #openstack-infra | 18:36 | |
| *** sdake has quit IRC | 18:36 | |
| *** sdake has joined #openstack-infra | 18:36 | |
| *** armax has quit IRC | 18:36 | |
| sweston | anteaya: okay, that would be helpful, thank you. I'll be at the point this week at which I will be ready to start integrating some of the repo, but most of it will have to be reworked | 18:36 |
| *** armax has joined #openstack-infra | 18:36 | |
| anteaya | understood | 18:36 |
| anteaya | let's see if we can get the radar repo opened up for you to contribute to | 18:37 |
| *** _nadya_ has quit IRC | 18:37 | |
| anteaya | I'll do my best | 18:37 |
| clarkb | fungi: dhellmann dstufft mordred jeblair SergeyLukjanov jhesketh https://review.openstack.org/#/c/141667/ that should be ready now | 18:37 |
| zaro | pelix: yeah, makes sense. would you like to make that suggestion on the review? | 18:37 |
| sweston | anteaya: ok, sounds good. | 18:38 |
| clarkb | and with that I need to run | 18:38 |
| *** nuritv has joined #openstack-infra | 18:39 | |
| *** Ryan_Lane has quit IRC | 18:41 | |
| *** nuritv has quit IRC | 18:41 | |
| anteaya | enjoy your run | 18:41 |
| anteaya | I think I am going to go for a walk | 18:41 |
| *** liusheng has quit IRC | 18:42 | |
| *** gyee_ has joined #openstack-infra | 18:42 | |
| *** liusheng has joined #openstack-infra | 18:42 | |
| *** gyee_ has quit IRC | 18:43 | |
| *** andreykurilin_ has joined #openstack-infra | 18:43 | |
| *** nuritv has joined #openstack-infra | 18:43 | |
| *** nuritv has quit IRC | 18:45 | |
| fungi | clarkb: all our mirrors have setuptools 8.0.3 now | 18:45 |
| fungi | and presumably new oslo.db as well | 18:45 |
| *** Masahiro has joined #openstack-infra | 18:45 | |
| *** ayoung has quit IRC | 18:46 | |
| jeblair | 2014-12-15 18:46:28,713 INFO: Downloading: https://pypi.python.org/packages/3.4/s/setuptools/setuptools-8.0.3-py2.py3-none-any.whl | 18:46 |
| jeblair | fungi: 8.0.3 was just released, right? | 18:47 |
| *** nuritv has joined #openstack-infra | 18:47 | |
| *** groknix has joined #openstack-infra | 18:48 | |
| fungi | jeblair: yep | 18:48 |
| fungi | like, minutes ago | 18:49 |
| fungi | so everything seems to be back in sync now | 18:49 |
| jeblair | thought so -- just saw that scroll by in my mirror testing | 18:49 |
| fungi | and running with the bandersnatch version released over the weekend | 18:49 |
| *** nuritv has quit IRC | 18:49 | |
| fungi | jeblair: note new bandersnatch release yesterday with another fix for raciness, in case you missed it | 18:49 |
| *** Masahiro has quit IRC | 18:50 | |
| jeblair | fungi: ah, thanks. i'm working from master from like friday, so i probably don't have that, but close. | 18:51 |
| *** lttrl has joined #openstack-infra | 18:51 | |
| fungi | heh | 18:51 |
| jeblair | i'll be sure to update as soon as i figure out how to do that with hg | 18:51 |
| fungi | hg clone uhdfuhiklzdfvuuhikuhiscrew it i'm importing this into git | 18:51 |
| dstufft | import all the things into git | 18:52 |
| *** Ryan_Lane has joined #openstack-infra | 18:52 | |
| *** nuritv has joined #openstack-infra | 18:52 | |
| *** wuhg has quit IRC | 18:53 | |
| *** nuritv has quit IRC | 18:54 | |
| *** zz_dimtruck is now known as dimtruck | 18:56 | |
| dhellmann | clarkb, mordred : I think I'm going to have to start asking for more comments in the pbr code that runs these git commands. _gert_version_from_git() o_O | 18:56 |
| lifeless | clarkb: hows the setuptools thing going? | 18:57 |
| *** dprince has quit IRC | 18:57 | |
| fungi | lifeless: mostly wrangled now. see http://lists.openstack.org/pipermail/openstack-dev/2014-December/052985.html | 18:57 |
| *** doug-fish has joined #openstack-infra | 18:57 | |
| *** otter768 has joined #openstack-infra | 18:58 | |
| *** nuritv has joined #openstack-infra | 18:58 | |
| *** cnesa2 has joined #openstack-infra | 18:58 | |
| *** cnesa has quit IRC | 18:58 | |
| mestery | sweston: neutron meeting is tomorrow | 18:58 |
| mestery | sweston: https://wiki.openstack.org/wiki/Network/Meetings | 18:58 |
| *** groknix has quit IRC | 18:59 | |
| *** nuritv has quit IRC | 18:59 | |
| *** rockyg has joined #openstack-infra | 18:59 | |
| lifeless | fungi: so we're going to sit pinned for a while, and then 8.0.2 will fix us? | 18:59 |
| *** ayoung has joined #openstack-infra | 19:00 | |
| clarkb | lifeless: lolno | 19:00 |
| *** whayutin_ has quit IRC | 19:00 | |
| lifeless | dhellmann: I may be able to shed light - I had to internalise the whole thing when I was doing changes there | 19:00 |
| clarkb | lifeless: we are going to drop git shas from versions and include them in egg info metadata instead | 19:00 |
| lifeless | dhellmann: what in particular are you looking at? | 19:00 |
| dhellmann | lifeless: https://review.openstack.org/#/c/141667/ | 19:00 |
| lifeless | clarkb: do we log that during test jobs? | 19:00 |
| fungi | lifeless: we can start | 19:01 |
| clarkb | lifeless: yes (via zuul) | 19:01 |
| clarkb | fungi: lifeless we always have | 19:01 |
| fungi | oh, right that part | 19:01 |
| lifeless | clarkb: the egg metadata? cool. I never saw that. | 19:01 |
| clarkb | lifeless: no | 19:02 |
| fungi | lifeless: but also mordred mentioned there that we should start running 'pbr freeze' in the jobs too so we have a record of the git shas for the versions being used | 19:02 |
| clarkb | we log it differently in the tests | 19:02 |
| clarkb | but we will also have pbr freeze which we can use too | 19:02 |
| sweston | mestery: yes, apologies .. need to update my calendar :-) | 19:02 |
| clarkb | so I think we have the use case pretty well covered | 19:02 |
| *** dustins_ has joined #openstack-infra | 19:02 | |
| *** andreykurilin_ has quit IRC | 19:02 | |
| *** andreykurilin_ has joined #openstack-infra | 19:03 | |
| *** otter768 has quit IRC | 19:03 | |
| lifeless | we'll still put the git versions in the debian package versions etc? | 19:03 |
| lifeless | actually what I'm asking is this - has this been considered with the same amount of input from operators that the semver changes got in the first place? | 19:04 |
| clarkb | lifeless: I think we can, I haven't looked at the change atop master | 19:04 |
| *** nuritv has joined #openstack-infra | 19:04 | |
| clarkb | lifeless: honestly no because it doesn't matter if setuptools doesn't work | 19:04 |
| lifeless | clarkb: well dstufft seemed to think we had a way to do it and work with setuptools, no ? | 19:05 |
| clarkb | you can't install anything properly without versions it groks | 19:05 |
| clarkb | lifeless: he was wrong | 19:05 |
| *** dustins has quit IRC | 19:05 | |
| lifeless | clarkb: hahahahahahahahah | 19:05 |
| lifeless | dstufft: why can't we have nice things? | 19:05 |
| dstufft | wait a moment, what did I say? | 19:06 |
| clarkb | dstufft: the +g thing | 19:06 |
| *** pblaho has joined #openstack-infra | 19:06 | |
| *** nuritv has quit IRC | 19:06 | |
| clarkb | it didn't work because old setuptools explodes on the + | 19:06 |
| clarkb | so we can't sanely have the gitsha in a version and expect it to work with 8.0 and 7.0 | 19:06 |
| fungi | lifeless: in this case it's somewhat complicated by the fact that pbr relying on setuptools emitting version details means we need to use already-pep-440-normalized versions strings in tarball/wheel names and egg-info | 19:06 |
| *** ociuhandu has joined #openstack-infra | 19:07 | |
| clarkb | that broke the +g thing then ^ broke the idea of supporting trailing alpha things | 19:07 |
| *** _nadya_ has joined #openstack-infra | 19:07 | |
| clarkb | because we would essentially be requiring everyong to have setuptool>=8.0 | 19:07 |
| dstufft | clarkb: right, and I was going to see about adjusting the spec so it allowed a trailing alpha numeric, and mordred said the pbr change will handle it instead | 19:08 |
| dstufft | so I didn't bother to run the numbers to see about that | 19:08 |
| clarkb | dstufft: ya its fine, it wouldn't have worked anyways | 19:08 |
| clarkb | so I htink mordreds approach is the best one | 19:08 |
| fungi | well, at least wouldn't have worked without a more significant overhaul of how we're doing things | 19:09 |
| annegent_ | hey reed, around? | 19:09 |
| dstufft | clarkb: I can possibly add a thing that'll tell setuptools _not_ to emit normalized versions | 19:09 |
| dstufft | like a flag | 19:09 |
| reed | annegent_, one ear yes | 19:10 |
| lifeless | clarkb: fungi: I think I have captured this correctly in the distutils sig thread | 19:10 |
| openstackgerrit | Ramy Asselin proposed openstack-infra/project-config: Prepare project-config for puppet module split #1 https://review.openstack.org/140523 | 19:10 |
| openstackgerrit | Ramy Asselin proposed openstack-infra/project-config: Prepare project-config for puppet module split #2 https://review.openstack.org/140548 | 19:10 |
| *** sarob has joined #openstack-infra | 19:10 | |
| * fungi catches up | 19:10 | |
| *** pblaho has quit IRC | 19:11 | |
| fungi | honestly, i think getting rid of the non-monotonic revision control noise in dev version strings is a cleaner way forward anyway | 19:11 |
| *** nuritv has joined #openstack-infra | 19:12 | |
| *** tonytan4ever has joined #openstack-infra | 19:12 | |
| clarkb | fungi: I disagree :) | 19:12 |
| clarkb | fungi: its particularly important for the CD crowd | 19:12 |
| clarkb | of which lifeless is a participant | 19:12 |
| clarkb | but that use case is relatively well dealt with using mordreds latest pbr changes so I think we will survive | 19:12 |
| openstackgerrit | Ramy Asselin proposed openstack-infra/project-config: Prepare project-config for puppet module split #2 https://review.openstack.org/140548 | 19:12 |
| dstufft | wouldn't they be getting a different devN if they are running off master anyways for each commit? | 19:13 |
| *** whayutin_ has joined #openstack-infra | 19:13 | |
| dstufft | or running off the same branch anyways | 19:13 |
| lifeless | fungi: +1 on non-monotonic; CD wants monotonic, but we *can't* avoid non-monotonic with a DVCS | 19:13 |
| *** nuritv has quit IRC | 19:13 | |
| fungi | i don't disagree that having the git sha mapping easily obtained from the software/package is good, just that it doesn't necessarily belong embedded in version numbers | 19:13 |
| dstufft | the git hash is only useful if you're mixing branches afaik | 19:13 |
| clarkb | dstufft: no that is not true, because git | 19:13 |
| lifeless | fungi: so -1 on removing the signal: if version A==version B, the contents of A must == the contents of B. | 19:13 |
| lifeless | the more I think about this, the worse it is | 19:13 |
| clarkb | dstufft: no no the way gerrit works nova has hundreds of commits with the same devN | 19:14 |
| dstufft | clarkb: but different branches | 19:14 |
| clarkb | dstufft: because everything is run and tested against master | 19:14 |
| *** teran_ has quit IRC | 19:14 | |
| clarkb | dstufft: no | 19:14 |
| openstackgerrit | Ramy Asselin proposed openstack-infra/project-config: Prepare project-config for puppet module split #1 https://review.openstack.org/140523 | 19:14 |
| clarkb | dstufft: everything is differnt applied to master | 19:14 |
| lifeless | clarkb: they are branches in the DAG | 19:14 |
| fungi | dstufft: ephemeral commits which never merge. so in theory different branches yes (thousands of ephemeral branches) | 19:14 |
| clarkb | lifeless: yes but that isn't reflected in the version data either | 19:14 |
| lifeless | clarkb: it is via the git sha | 19:14 |
| clarkb | lifeless: here "branch" is is the last tag different | 19:15 |
| dstufft | clarkb: and the CI people are applying random builds from gerrit instead of tracking master? | 19:15 |
| clarkb | lifeless: which we are dropping | 19:15 |
| lifeless | clarkb: and which I've just -1'd. | 19:15 |
| *** sarob has quit IRC | 19:15 | |
| *** nuritv has joined #openstack-infra | 19:15 | |
| clarkb | lifeless: sure its a fair thing to -1 just understand there is no other way to fix it given the state of the world right now | 19:15 |
| *** jkraj has quit IRC | 19:15 | |
| lifeless | dstufft: no, they are building their own branch because of local fixes etc | 19:15 |
| clarkb | lifeless: and your email seems to capture that relatively well | 19:16 |
| lifeless | clarkb: setuptools reverting would fix it | 19:16 |
| *** _nadya_ has quit IRC | 19:16 | |
| dstufft | setuptools reverting only fixes it until we re-land PEP 440 | 19:16 |
| jeblair | note that while it is helpful to know that version 1.0+abc != 1.0+def, i have still taken to the habit of uninstalling installed versions of software with those hashes because of the sorting problem (i never know whether a different 1.0+hash will sort > or < the installed version) | 19:17 |
| *** dims has quit IRC | 19:17 | |
| fungi | it seems like setuptools sees this as ripping off a bandage, pain expected, hopefully short-lived as the python package community solves all the issues around it | 19:17 |
| jeblair | so while it is a useful signal that some package is not the same as what's installed, the only thing i can do with that signal is "go clean up the mess" | 19:17 |
| *** nuritv has quit IRC | 19:17 | |
| clarkb | jeblair: right, which is why I think mordreds change is a good one | 19:17 |
| *** dims has joined #openstack-infra | 19:17 | |
| mmedvede | asselin: there is a typo in the topic of your recent patches | 19:17 |
| clarkb | and as was pointed out it puts the sha1 in every version regardless of whether or not it is a tag | 19:18 |
| clarkb | which is also handy | 19:18 |
| asselin | mmedvede, thanks. I'll check and fix | 19:18 |
| clarkb | that said the commit hash is part of the version and it would be nice if we could have that info there | 19:18 |
| clarkb | if we can't then oh well | 19:18 |
| clarkb | anothe rthought I had was the convert the hash to base 10 and put that in the version | 19:18 |
| fungi | dstufft: and some ci systems (such as ours) _are_ installing random commits from gerrit because they need to test whether or not they function | 19:18 |
| clarkb | since that would be valid under pep440 | 19:18 |
| lifeless | jeblair: the rule used to be that they sorted == | 19:18 |
| clarkb | but that would make sorting even more confusing | 19:19 |
| sdague | if it's not there we're going to need to do a bunch more dump of things | 19:19 |
| *** smcginnis has joined #openstack-infra | 19:19 | |
| lifeless | jeblair: e.g. installing X+g1 over X+g0 would replace it rather than going 'oh, installed already' | 19:19 |
| sdague | because pip-freeze having that is extremely important in understanding what's coming from where | 19:19 |
| clarkb | sdague: mordred's change comes with a pbr freeze | 19:19 |
| clarkb | sdague: we would just use that | 19:19 |
| sdague | ok | 19:19 |
| *** sputnik13 has quit IRC | 19:19 | |
| fungi | sdague: which will have the added benefit that you get the shas associated with the tagged version numbers too | 19:20 |
| lifeless | clarkb: which then we're telling operators to do as well | 19:20 |
| *** MaxV has joined #openstack-infra | 19:20 | |
| asselin | mmedvede, done | 19:20 |
| dstufft | So like I offered over the weekend, I can check on the feasibility of adding a normalization rule to PEP 440 so that a single trailing .alphanumeric gets converted to a +alphanumeric | 19:20 |
| clarkb | lifeless: yes if deploying from pbr I think that would be the expectation | 19:20 |
| dstufft | I just didn't bother to do that because mordred told me not to | 19:21 |
| mmedvede | asselin: thx | 19:21 |
| clarkb | dstufft: keep in mind that would require changing the emit behavior too | 19:21 |
| fungi | dstufft: however, to take advantage of that, we'd need to stop relying on normalized versions from setuptools in naming our sdists/wheels | 19:21 |
| *** MaxV has quit IRC | 19:21 | |
| fungi | er, what clarkb just said | 19:21 |
| *** dims has quit IRC | 19:21 | |
| dstufft | do you check against that tag somewhere? I thought that was jsut in released versions | 19:22 |
| clarkb | humans do. Its a sanity thing | 19:22 |
| clarkb | there should be a 1:1 relationship between those strings that doens't include normalization rules | 19:22 |
| clarkb | especially since if we don't change emit behavior that means the version on pypi will have a + | 19:23 |
| fungi | dstufft: also that's where the continuous deployment crowd gets bitten. they're installing non-release versions, so likely need to be able to generate tarballs/wheels | 19:23 |
| dstufft | PyPI won't allow anything with a + in it to be uploaded | 19:23 |
| *** sputnik13 has joined #openstack-infra | 19:23 | |
| *** nuritv has joined #openstack-infra | 19:23 | |
| clarkb | dstufft: sorry let me rephrase | 19:23 |
| clarkb | dstufft: local pypi type things like mirrors or devpi could have + in them | 19:23 |
| clarkb | since that is again how CDers deploy their stuff | 19:24 |
| dstufft | yes | 19:24 |
| clarkb | and if you have old setuptools that breaks ou | 19:24 |
| *** markmcclain has quit IRC | 19:24 | |
| clarkb | so we could just require new setuptools and use the + | 19:24 |
| clarkb | instead of adding new normalization rules that will in practice not add any benefit | 19:24 |
| *** sputnik13 has quit IRC | 19:25 | |
| *** nuritv has quit IRC | 19:25 | |
| clarkb | anyways I really really need to go now or will be late for things | 19:25 |
| dstufft | well it'd only require new setuptools if you're installing a dev version from some sort of index, not in general. Presumably since PyPI won't allow uploads like that someone doing that is doing it internally and has some semblance of control over what version of setuptools they are using. Alternatively I can probably add a flag to disable the "emit normalized" behavior | 19:26 |
| lifeless | clarkb: what happens with trunk pbr? it already has + - and worked on setuptools 7, so what breaks if we just release trunk ? | 19:26 |
| *** sputnik13 has joined #openstack-infra | 19:26 | |
| clarkb | lifeless old setup tools explodes on + | 19:26 |
| lifeless | how old? | 19:27 |
| lifeless | ah, no- trunk is .g | 19:27 |
| *** Swanson has joined #openstack-infra | 19:27 | |
| *** nuritv has joined #openstack-infra | 19:27 | |
| lifeless | it does +g for debian serialisation | 19:27 |
| clarkb | release before 8.0 | 19:27 |
| dstufft | it explodes on + depending on at what place you input the version into setuptools, because setuptools (apparently) has multiple places it parses versions | 19:27 |
| fungi | right, we tried (briefly) running with a pbr which used +g instead of .g and it went badly | 19:28 |
| dstufft | with multiple different things it allows depending on what place you're sticking a version | 19:28 |
| *** whayutin_ has quit IRC | 19:28 | |
| *** sputnik13 has quit IRC | 19:28 | |
| lifeless | trunk will be no different there | 19:28 |
| *** smcginnis has left #openstack-infra | 19:28 | |
| lifeless | but FWIW the patch is a one liner | 19:28 |
| lifeless | (release_string) - to output it | 19:28 |
| *** nuritv has quit IRC | 19:29 | |
| *** markmcclain has joined #openstack-infra | 19:30 | |
| openstackgerrit | Ramy Asselin proposed openstack-infra/project-config: Add puppet-recheckwatch as split out module https://review.openstack.org/140430 | 19:32 |
| dhellmann | clarkb: I'm looking into a 1.2.1 release of oslo.db for stable/juno -- is the requirements spec for sqlalchemy in the stable/juno branch of the requirements list correct? | 19:34 |
| *** sputnik13 has joined #openstack-infra | 19:34 | |
| dstufft | dhellmann: it is similar in scope to the old one | 19:34 |
| dstufft | allows the same versions | 19:34 |
| fungi | dhellmann: since https://review.openstack.org/141584 merged, yes | 19:34 |
| dhellmann | dstufft: ok | 19:34 |
| dstufft | could theortically allow some extra ones if SQLAlchemy went backwards and released them | 19:35 |
| dstufft | but it shouldn't _not_ allow something that the old one did | 19:35 |
| *** sputnik13 has quit IRC | 19:36 | |
| dhellmann | hrm | 19:36 |
| dhellmann | the 1.0.x series is stable/juno, but we don't want to set a cap lower than we have to | 19:36 |
| dstufft | (by that I mean, it does like !=0.9.3, if SQLA released a 0.9.3.1 or something it'd allow that) | 19:36 |
| dhellmann | I suppose I could create a 1.2.0 branch and release 1.2.1 from there | 19:36 |
| *** nuritv has joined #openstack-infra | 19:37 | |
| *** sputnik13 has joined #openstack-infra | 19:37 | |
| dhellmann | why did we pull sqlalchemy 0.8.x out of the allowed range in master? | 19:37 |
| dstufft | dhellmann: someone suggested to me that we should | 19:37 |
| fungi | dhellmann: you'll want to name it either stable/juno or feature/1.2 depending on how you plan to go about it | 19:37 |
| dstufft | I forget who it was | 19:37 |
| dhellmann | dstufft: I think we can convince zzzeek not to do that | 19:37 |
| dstufft | someone in here | 19:37 |
| dstufft | sot hat's what I did | 19:37 |
| *** dims has joined #openstack-infra | 19:37 | |
| dstufft | I think it was like a | 19:37 |
| dhellmann | fungi: we already have a stable/juno branch started from 1.0.0, which is why this is confusing | 19:37 |
| fungi | dhellmann: ahh | 19:38 |
| dstufft | "well we can probably drop 0.8 in master" sort of thing | 19:38 |
| dstufft | if i remember correctly | 19:38 |
| *** sputnik13 has quit IRC | 19:39 | |
| dhellmann | ok, I'm sure the oslo.db team will be happy to be able to drop some of the compat code they're maintaining | 19:39 |
| *** david-lyle is now known as david-lyle_lunch | 19:39 | |
| *** nuritv has quit IRC | 19:39 | |
| dstufft | So if you want the sha in the version tag I see five ways forward -> A) Use + and mandate that everyone moves to setuptools 8+ B) Use . and mandate that everyone stays on setuptools < 8 C) I check to see if we can normalize .alphanum to +alphanum and CD operators need to be on setuptools >= 8 D) I check to see if we can normalize .alphanum to +alphanum and make it so pbr can turn off emiting normalized E) Use + and see about releasing | 19:40 |
| dstufft | patched copies of the older setuptools that won't explode | 19:40 |
| *** sputnik13 has joined #openstack-infra | 19:40 | |
| *** nuritv has joined #openstack-infra | 19:40 | |
| sdague | dhellmann: while you have a stable juno branch, because it's not capped in requirements, stable/* are totally allowed to use latest oslo.db (that's what their requirements say) | 19:40 |
| dhellmann | sdague: right, I'm just trying to figure out where to put a change to the sqlalchemy version used by oslo.db for stable/juno, and what version to tag it as | 19:41 |
| sdague | dhellmann: ah, gotcha | 19:41 |
| dhellmann | if we're going to make a 1.2.1 and cap juno to that, then I need a new branch | 19:41 |
| dhellmann | but it's not clear that's the right fix, either | 19:41 |
| sdague | so, previously stable/juno was < 1.2 | 19:41 |
| dhellmann | yeah, I guess we want stable/juno to be <=1.2.1 and kilo to be >=1.2.1 | 19:42 |
| sdague | sure | 19:42 |
| *** nuritv has quit IRC | 19:42 | |
| sdague | yeh, that would be a transitional bridge | 19:42 |
| *** whayutin_ has joined #openstack-infra | 19:42 | |
| dhellmann | right now master is >=1.1.0, but with the sqlalchemy change that's no longer correct I think | 19:42 |
| dhellmann | so now what do we call this branch? "feature/1.2" seems wrong | 19:44 |
| fungi | dhellmann: i'm unconvinced we should >=1.2.1 in master. openstack still works with earlier versions, just recent setuptools causes a problem | 19:44 |
| dhellmann | fungi: ok, fair point | 19:44 |
| dhellmann | so just the cap to <=1.2.1 in stable/juno then | 19:45 |
| fungi | we won't be able to run tests in master for older oslo.db with newer setuptools, but we'd be testing with the latest version allowed by the requirements list regardless | 19:45 |
| *** rwsu has joined #openstack-infra | 19:45 | |
| lifeless | dstufft: C D and E all make a lot of sense to me | 19:45 |
| dhellmann | yeah, I think it's safe to assume if you end up with something older installed it came from a package somewhere anyway | 19:45 |
| *** Ryan_Lane has quit IRC | 19:46 | |
| lifeless | dstufft: B isn't long term tenanble | 19:46 |
| lifeless | dstufft: and A seems likely to be disruptive | 19:46 |
| dhellmann | fungi: what name should we give this branch? "feature/1.2" feels odd because it's not really a feature, but is "stable/1.2" going to cause trouble with some of our automation tools? | 19:46 |
| *** armax has quit IRC | 19:46 | |
| dstufft | lifeless: I don't believe so either, I was jsut being exhaustive as to what I could think of :) | 19:46 |
| lifeless | :) | 19:47 |
| fungi | dhellmann: stable/1.2 will fail because of the devstack feature matrix regarding it as an invalid stable branch name (i got to go through this exercise with pbr on saturday) | 19:47 |
| lifeless | what TZ is mordred in at the moment? | 19:47 |
| dstufft | and of course there's the other option of not including the sha sums in the version number which mordred's pbr change does (though I think the change in general is a good one even if you keep the sha sums in the version numbers, it'd nice to know what git hash a released version came from too) | 19:47 |
| dhellmann | fungi: ok, I'll take the benefit of your experience then -- would you please make me a feature/1.2 branch for oslo.db from changeset 4ad43a40611ef8cba715aa4d392dda40c079894a | 19:47 |
| fungi | dhellmann: we settled on feature/0.10 for the pbr backport releases, but the plan was to delete that branch as soon as it got tagged | 19:47 |
| harlowja | hmm, qq, i assume the juno gate stuff is being adjusted due to this setuptools stuffs, i'm seeing 'pkg_resources.VersionConflict: SQLAlchemy 0.8.4 is installed but SQLAlchemy>=0.9.7,<=0.9.99 is required by ['oslo.db']' | 19:47 |
| dstufft | I'm happy to do whatever I need to do on my side for whatever option seems best | 19:47 |
| harlowja | * http://logs.openstack.org/80/121280/67/check//gate-tempest-dsvm-neutron-src-taskflow-juno/4b4ed02/logs/devstacklog.txt.gz | 19:47 |
| *** nuritv has joined #openstack-infra | 19:48 | |
| lifeless | dstufft: yes, I support putting the git info in the metadata files; separate from removing it in dev builds :) | 19:48 |
| dhellmann | fungi: we'll need to keep this branch in case we have to make more 1.2 backports | 19:48 |
| dhellmann | harlowja: yeah, we're working out the fix right now | 19:48 |
| harlowja | thx dhellmann | 19:48 |
| lifeless | dstufft: I can't speak for mordred, but I'd be super happy if you investigate the C/D/E options | 19:48 |
| fungi | dhellmann: well, we can always re-branch from the 1.2.1 tag, but sure it can just be left there too | 19:48 |
| *** annegent_ has quit IRC | 19:49 | |
| dstufft | lifeless: okies | 19:49 |
| *** nuritv has quit IRC | 19:49 | |
| dhellmann | fungi: yeah, this is going to mess with how we backport oslo.db fixes to juno, too -- I wonder if we should move the stable/juno branch instead of making a new one | 19:50 |
| dhellmann | that feels dangerous though | 19:50 |
| fungi | dhellmann: okay, now an openstack/oslo.db feature/1.2 branch exists, starting from the 1.2.0 tag (commit a73d499) | 19:50 |
| lifeless | argh | 19:50 |
| lifeless | https://www.python.org/dev/peps/pep-0440/#changing-the-version-scheme | 19:50 |
| lifeless | is going to bite us separately | 19:50 |
| *** nuritv has joined #openstack-infra | 19:50 | |
| *** spzala has quit IRC | 19:51 | |
| fungi | dhellmann: that's certainly an option... fast-forwarding it to the 1.2.0 commit should be fine assuming there are no other changes in that branch deviating from master | 19:51 |
| lifeless | clarkb: fungi: ^ this *also* means that the current pbr trunk versions are as correct as we can make them, modulo the +. | 19:51 |
| *** spzala has joined #openstack-infra | 19:51 | |
| dhellmann | fungi: we have some cherry picks for existing backports | 19:52 |
| *** nuritv has quit IRC | 19:52 | |
| fungi | dhellmann: you may need a merge from the 1.2.0 tag then to keep the branch ff-only compatible | 19:52 |
| dstufft | lifeless: what's going to bite you about that? | 19:52 |
| *** david-lyle_lunch is now known as david-lyle | 19:53 | |
| lifeless | dstufft: CD deployers deploying off of master | 19:53 |
| *** sputnik13 has quit IRC | 19:53 | |
| *** HeOS has joined #openstack-infra | 19:53 | |
| dstufft | lifeless: not sure I understand | 19:53 |
| lifeless | dstufft: we start with a release | 19:53 |
| lifeless | version 1.0 | 19:53 |
| lifeless | we do one commit | 19:53 |
| dhellmann | fungi: I've never done that through gerrit before, how does that work? create a merge commit and then review it? | 19:53 |
| lifeless | version 1.0.dev1 | 19:53 |
| lifeless | we do a second commit | 19:53 |
| *** david-lyle is now known as david-lyle_t | 19:53 | |
| lifeless | 1.0.dev2 | 19:53 |
| lifeless | we tag that | 19:53 |
| *** david-lyle_t is now known as david-lyle | 19:53 | |
| lifeless | 1.0a1 | 19:53 |
| lifeless | which has the same content as 1.0.dev2 | 19:54 |
| dstufft | ok? | 19:54 |
| fungi | dhellmann: i'm thinking through potential gotchas there. for starters the acl isn't going to allow that with a stable-named branch, but there are other linearity issues to think about too | 19:54 |
| lifeless | we wanted to then do | 19:54 |
| lifeless | we do another commit | 19:54 |
| lifeless | we wanted to make that be 1.0.a1.dev1 | 19:54 |
| lifeless | but 1.0.dev2 will sort before it. | 19:54 |
| lifeless | So it will never install. | 19:55 |
| fungi | dhellmann: the acl thing is tractible, but i'm slightly worried about branching from a 1.2.0 which isn't actually 1.2.0 (because it involves a merge) | 19:55 |
| fungi | clarkb: ideas there? ^ | 19:55 |
| *** sputnik13 has joined #openstack-infra | 19:55 | |
| *** zz_sabari is now known as zz_zz_sabari | 19:55 | |
| dhellmann | fungi: yeah, and I think dragging the 1.2.0 requirements back into stable/juno is going to open another can of worms | 19:55 |
| lifeless | so | 19:55 |
| *** rockyg has quit IRC | 19:55 | |
| dstufft | lifeless: sort before means "less than" | 19:55 |
| dstufft | 1.0.dev2 is < 1.0.a1 | 19:56 |
| lifeless | oh | 19:56 |
| *** jseiler has joined #openstack-infra | 19:56 | |
| lifeless | ok so this is the reason we can't release pbr trunk | 19:56 |
| dhellmann | fungi: I'll go with the feature/1.2 branch for now, release a 1.2.1, and write this up for the mailing list so the oslo.db team knows about handling backports differently | 19:56 |
| lifeless | but we can fix I think | 19:56 |
| lifeless | dstufft: so 'ahead of' means 'lower than' ? | 19:56 |
| dstufft | lifeless: it's kind of confusing I guess, the PEP sort order talks in terms of sort order in an ascending fashion, though often times people think descending when they think versions | 19:56 |
| dstufft | essentially the PEP means what order does sorted() put it in if they are in a list | 19:57 |
| dstufft | sorted([1, 2, 3, 4]) => [1, 2, 3, 4] | 19:57 |
| dstufft | 1 sorts ahead of 2 | 19:57 |
| dstufft | 2 sorts lower than 3 | 19:57 |
| *** yamamoto has joined #openstack-infra | 19:57 | |
| dstufft | the _old_ PEP, PEP 386 had the behavior you were talking about | 19:57 |
| lifeless | would you accept a patch to make it use lesser and greater consistently ? | 19:57 |
| nelsnelson | Greetings, channel. Could I please request another review of this submission? https://review.openstack.org/#/c/134348/ Thanks for your attention. | 19:57 |
| dstufft | .dev was _higher_ than an alpha and beta and rc and was sorted between rc and the release | 19:58 |
| lifeless | dstufft: ok so this means we *have* to do 1.0a1.dev2 | 19:58 |
| lifeless | or dev3 will never be installed | 19:58 |
| lifeless | dstufft: that will work, right ? | 19:58 |
| dstufft | the change sin PEP 440 started out as me arguing that made no sense and it was a brea kin compat in the order that setuptools put them in | 19:58 |
| dstufft | lifeless: and yes we accept clarification patches always | 19:59 |
| *** doug-fish1 has joined #openstack-infra | 19:59 | |
| *** esp has joined #openstack-infra | 19:59 | |
| dstufft | lifeless: 1.0.dev0, 1.0.dev1, 1.0.a1, 1.0.a1.dev0, 1.0.a1.dev1, 1.0 | 19:59 |
| dstufft | terms of "oldest" to "newest" | 20:00 |
| dstufft | (maybe older and newer is better terminology than lower/higher) | 20:00 |
| dstufft | er | 20:00 |
| *** whayutin_ has quit IRC | 20:00 | |
| dstufft | sorry that's wrong | 20:00 |
| dstufft | 1.0.dev0, 1.0.dev1, 1.0.a1.dev0, 1.0.a1.dev1, 1.0.a1, 1.0 | 20:00 |
| dstufft | that's right | 20:00 |
| lifeless | so thats a problem | 20:00 |
| *** sputnik13 has quit IRC | 20:01 | |
| *** yolanda has quit IRC | 20:01 | |
| lifeless | dstufft: whats the next version number after 1.0.a1 ? | 20:01 |
| lifeless | 1.0.a2.dev1 | 20:01 |
| lifeless | sorted | 20:01 |
| lifeless | I need to wake up :) | 20:01 |
| *** yamamoto has quit IRC | 20:01 | |
| lifeless | this is why the semver code has the stuff in it it has ;) | 20:02 |
| lifeless | righto, time to go - later folk | 20:02 |
| *** doug-fish has quit IRC | 20:02 | |
| dstufft | (just double checked my answer, and that sort order is for pre 8 and post 8 setuptools) | 20:02 |
| dstufft | lifeless: a2 is after a1, or .postN | 20:02 |
| dstufft | a2 is probably what you want though | 20:02 |
| dstufft | a2.devN | 20:02 |
| *** tkelsey has joined #openstack-infra | 20:03 | |
| *** doug-fish1 has quit IRC | 20:04 | |
| *** annegent_ has joined #openstack-infra | 20:04 | |
| *** subscope has joined #openstack-infra | 20:05 | |
| *** stpierre has joined #openstack-infra | 20:05 | |
| stpierre | can anyone help me figure out what's happening with tempest here? http://logs.openstack.org/51/120451/3/check//check-grenade-dsvm/ef5f14f/logs/grenade.sh.txt.gz (or, failing that, help me find the correct IRC channel in which to ask?) thanks. | 20:05 |
| fungi | just tested out mordred's 141666 for pbr master with tip of gertty master, and here's the (sorted) diff in output between pip freeze and pbr freeze: http://paste.openstack.org/show/151391/ | 20:06 |
| *** doug-fish has joined #openstack-infra | 20:06 | |
| fungi | i like that it includes python interpreter version, pip and setuptools too (sor of like pip list) | 20:06 |
| *** bitblt has joined #openstack-infra | 20:07 | |
| jeblair | i have made a commit; now i need to figure out how to push that to bitbucket... | 20:07 |
| *** bitblt has left #openstack-infra | 20:07 | |
| *** kevinbenton has joined #openstack-infra | 20:07 | |
| dhellmann | fungi: https://review.openstack.org/141893 https://review.openstack.org/141894 | 20:07 |
| dhellmann | harlowja: ^^ | 20:07 |
| harlowja | cools | 20:08 |
| harlowja | durn setuptools | 20:08 |
| jeblair | fungi: oh that's a git sha in the comment? | 20:08 |
| anteaya | hi asselin I had gone for a walk | 20:08 |
| fungi | jeblair: yep | 20:08 |
| anteaya | asselin: on the mailing list I stated what I am doing | 20:08 |
| dstufft | might make sense to make that like | 20:08 |
| dstufft | git: <hash> | 20:08 |
| dstufft | er, # git: <hash> | 20:08 |
| dhellmann | fungi: oh, interesting, those patches don't seem to have triggered zuul | 20:08 |
| anteaya | asselin: and the times I am available to answer questions during third party meetings | 20:08 |
| fungi | jeblair: that's the proposed alternative to making it a subcomponent of the version string itself | 20:08 |
| jeblair | fungi: *nod* | 20:09 |
| harlowja | someone just needs to implement a parser for that version that follows how people think it should be; some simple DSL :-P | 20:09 |
| anteaya | asselin: anyone is welcome to join me to ask questions and share ideas during those times I have stated I will be available | 20:09 |
| lifeless | clarkb: fungi: - I have to go - my opinion is on https://review.openstack.org/#/c/141667/ now | 20:09 |
| *** jedimike has quit IRC | 20:09 | |
| *** subscope has quit IRC | 20:09 | |
| fungi | harlowja: all the versioning standards are broken in some way. what we need is a new standard to replace them... http://xkcd.com/927/ | 20:10 |
| lifeless | If I had the time to commit to being around to discuss it, I would -2 this | 20:10 |
| lifeless | But I'm on leave and life is short. | 20:10 |
| harlowja | fungi exactly! | 20:10 |
| fungi | lifeless: thanks for checking back in on your travels | 20:10 |
| harlowja | i was just thinking; u know something like the following DSL "(<=0.84) or (>=0.99)" | 20:10 |
| harlowja | something simple with just 'or' 'and ' | 20:11 |
| harlowja | and be done with it, ha | 20:11 |
| dstufft | harlowja: my plan for the future is to do just that | 20:11 |
| *** Ryan_Lane has joined #openstack-infra | 20:11 | |
| harlowja | woot | 20:11 |
| dstufft | although it'll sit alongside this, so the , will just be an alias for AND | 20:11 |
| harlowja | that will match what a SAT solver needs also | 20:12 |
| harlowja | so that will help making a good dependency solver | 20:12 |
| harlowja | someday | 20:12 |
| dstufft | yea, that's a future task for me too, first pip needs more refactorings to make the code base sane so we can add one at all | 20:12 |
| dstufft | pip is very spaghetti | 20:12 |
| *** whayutin_ has joined #openstack-infra | 20:13 | |
| fungi | dhellmann: zuul may just be busy handling its events/results queues. i'm looking into it | 20:13 |
| dstufft | part of what I'm doing here is taking features from setuptools and pip, standardizing, and putting them into https://github.com/pypa/packaging which then gets used instead inside of pip/setuptools | 20:13 |
| harlowja | cool | 20:13 |
| dhellmann | fungi: ok, I wasn't sure if the branch name had anything to do with it | 20:13 |
| *** dprince has joined #openstack-infra | 20:13 | |
| fungi | dhellmann: i think you caught it right as a gate reset or two happened | 20:14 |
| dhellmann | fungi: ok | 20:14 |
| *** amuller has quit IRC | 20:14 | |
| dstufft | PEP 440 is probably the most painful one of those :/ though we tried to make it less painful overall | 20:14 |
| *** ociuhandu has quit IRC | 20:14 | |
| fungi | dhellmann: yep, they're in check now | 20:17 |
| dhellmann | fungi: I'm spoiled; usually zuul shows the patch immediately :-) | 20:17 |
| *** shashankhegde has quit IRC | 20:18 | |
| *** ihrachyshka has quit IRC | 20:20 | |
| *** prad has quit IRC | 20:21 | |
| dhellmann | fungi, dstufft, sdague: https://review.openstack.org/#/c/141896/ caps oslo.db in juno | 20:22 |
| *** annegen__ has joined #openstack-infra | 20:22 | |
| *** viglesias has quit IRC | 20:22 | |
| *** annegent_ has quit IRC | 20:22 | |
| *** prad has joined #openstack-infra | 20:22 | |
| *** ihrachyshka has joined #openstack-infra | 20:22 | |
| *** baoli has quit IRC | 20:23 | |
| *** vhoward has left #openstack-infra | 20:23 | |
| *** viglesias has joined #openstack-infra | 20:25 | |
| anteaya | EmilienM: okay so the puppet-beaker-rspec job runs on puppet-storyboard for instance: http://git.openstack.org/cgit/openstack-infra/puppet-storyboard/tree/ | 20:26 |
| anteaya | which does have a Rakefile, yay off to a good start | 20:27 |
| *** avozza is now known as zz_avozza | 20:27 | |
| anteaya | however the commit message states "For modules that have unit tests in the spec/ directory" | 20:27 |
| anteaya | I don't see that puppet-storyboard has a spec/ directory | 20:28 |
| anteaya | EmilienM: what am I not understanding? | 20:28 |
| nibalizer | anteaya: those tests run on the stackforge/puppet-* projects as well | 20:28 |
| *** zz_avozza is now known as avozza | 20:28 | |
| anteaya | nibalizer: okay | 20:28 |
| nibalizer | and those projects do have unit tests | 20:28 |
| nibalizer | so it worked great in the case of storyboard because it did not have unit tests, but against a more mature module it fell short | 20:29 |
| anteaya | so if we are changing a job, I need to see how the functionality will affect the projects running that job | 20:29 |
| anteaya | and puppet-storyboard runs that job and doesn't have a spec/ directory | 20:29 |
| anteaya | so perhaps I don't yet understand the commit message | 20:29 |
| *** eharney has quit IRC | 20:30 | |
| *** nithyag_ has quit IRC | 20:30 | |
| *** hamzy_vacation has quit IRC | 20:30 | |
| *** redrobot has quit IRC | 20:30 | |
| *** jcooley has quit IRC | 20:30 | |
| *** spiffxp has quit IRC | 20:30 | |
| *** leifmadsen has quit IRC | 20:30 | |
| *** timfreund has quit IRC | 20:30 | |
| *** sdague has quit IRC | 20:30 | |
| *** crinkle has quit IRC | 20:30 | |
| *** cinerama has quit IRC | 20:30 | |
| *** jogo has quit IRC | 20:30 | |
| *** grantbow has quit IRC | 20:30 | |
| *** Guest20522 has quit IRC | 20:30 | |
| *** niedbalski has quit IRC | 20:30 | |
| *** avozza is now known as zz_avozza | 20:30 | |
| anteaya | because after reading it I was expecting the job being edited to run on repos with spec/ directories only | 20:30 |
| anteaya | which according to the first repo I found, is not the case | 20:30 |
| *** tonytan4ever has quit IRC | 20:30 | |
| anteaya | I'm willing to say that I don't understand | 20:30 |
| anteaya | which is probably the case | 20:30 |
| anteaya | but I would like to | 20:30 |
| *** zz_avozza is now known as avozza | 20:31 | |
| *** prad has quit IRC | 20:31 | |
| *** eharney has joined #openstack-infra | 20:31 | |
| *** nithyag_ has joined #openstack-infra | 20:31 | |
| *** hamzy_vacation has joined #openstack-infra | 20:31 | |
| *** redrobot has joined #openstack-infra | 20:31 | |
| *** jcooley has joined #openstack-infra | 20:31 | |
| *** spiffxp has joined #openstack-infra | 20:31 | |
| *** leifmadsen has joined #openstack-infra | 20:31 | |
| *** timfreund has joined #openstack-infra | 20:31 | |
| *** sdague has joined #openstack-infra | 20:31 | |
| *** crinkle has joined #openstack-infra | 20:31 | |
| *** cinerama has joined #openstack-infra | 20:31 | |
| *** jogo has joined #openstack-infra | 20:31 | |
| *** grantbow has joined #openstack-infra | 20:31 | |
| *** Guest20522 has joined #openstack-infra | 20:31 | |
| *** niedbalski has joined #openstack-infra | 20:31 | |
| nibalizer | okay let me try to explain | 20:32 |
| * anteaya listens | 20:33 | |
| *** _nadya_ has joined #openstack-infra | 20:33 | |
| nibalizer | the end state we want all modules to be in is they have beaker tests and rspec-puppet tests, those are sometimes called acceptance tests and unit tests, respectively | 20:33 |
| nibalizer | now i made that job so that i could run and develop acceptance tests for storyboard (and other modules, but i started on storyboard) | 20:33 |
| anteaya | I support the end state | 20:34 |
| nibalizer | and storyboard doesn't have unit tests because i haven't gotten there yet | 20:34 |
| nibalizer | so i iterated until it worked | 20:34 |
| anteaya | until what worked? | 20:34 |
| nibalizer | puppet storyboard + acceptance tests + beaker testing | 20:34 |
| *** Masahiro has joined #openstack-infra | 20:34 | |
| anteaya | having beaker run and not fall over? | 20:34 |
| nibalizer | ya | 20:34 |
| nibalizer | and everything was good until someone tried to use that test on a module that had unit tests as well | 20:35 |
| anteaya | okay | 20:35 |
| anteaya | where are the acceptance tests for puppet-storyboard? | 20:35 |
| nibalizer | turns out the syntax i was using in that test was wrong, and it was kindof a fluke that it worked at all | 20:35 |
| *** sarob has joined #openstack-infra | 20:35 | |
| anteaya | I couldn't find them | 20:35 |
| sdague | dhellmann: actually, you want to make it < 1.3 ? | 20:35 |
| nibalizer | in a review that hasn't landed yet | 20:35 |
| anteaya | url? | 20:35 |
| sdague | because that allows for possible future fixes without updating the world | 20:35 |
| dhellmann | sdague: oh, good point | 20:35 |
| nibalizer | anteaya: https://review.openstack.org/#/c/126086/ | 20:36 |
| anteaya | nibalizer: while I look this over, can we add this and its significance to the commit message for https://review.openstack.org/#/c/141728/1 | 20:36 |
| *** sarob has quit IRC | 20:36 | |
| dhellmann | sdague: updated | 20:37 |
| sdague | dhellmann: +2 | 20:37 |
| anteaya | nibalizer: because so far I am feeling that 141728 depends on 126086 | 20:37 |
| dhellmann | sdague: I'll let the tests pass and then push that through | 20:37 |
| *** Masahiro has quit IRC | 20:39 | |
| nibalizer | anteaya: no if anything 126086 depends on 141728 | 20:39 |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/elastic-recheck: Bugs _in_ this project go in StoryBoard now https://review.openstack.org/139120 | 20:39 |
| nibalizer | 141728 is about improving the testing infra and 126086 just takes adavantage of those improvements | 20:40 |
| anteaya | nibalizer: okay can we track that relationship then in the commit message? | 20:40 |
| *** amcrn has joined #openstack-infra | 20:41 | |
| anteaya | I'm lost on how we improve the testing structure that requires a directory of tests in order to run that don't yet exist in the repo | 20:41 |
| * anteaya goes back to listening | 20:41 | |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/system-config: Elastic recheck has two kinds of bugs https://review.openstack.org/139139 | 20:42 |
| nibalizer | so the puppet-storyboard stuff all works fine under the current state of project-config | 20:42 |
| nibalizer | what doesn't work fine is the stackforge modules | 20:42 |
| nibalizer | https://review.openstack.org/#/c/140844/ is an example | 20:42 |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/system-config: Elastic recheck bug tracking is now in storyboard https://review.openstack.org/139139 | 20:43 |
| anteaya | Don't know how to build task 'spec/acceptance' | 20:43 |
| anteaya | so rake is upset and you want to use rspec | 20:44 |
| nibalizer | ya, basically | 20:44 |
| anteaya | so can you show me a paste of output of what rspec does on puppet-storyboard in its current state? | 20:45 |
| nibalizer | calling 'rake spec' should cause the unit tests to be run, the fact that 'rake spec spec/acceptance' did the beaker tests is wild | 20:45 |
| anteaya | well especially since there were no spec/acceptance tests in the first place | 20:45 |
| *** MarkAtwood has joined #openstack-infra | 20:46 | |
| anteaya | btw I have never worked with beaker | 20:46 |
| *** baoli has joined #openstack-infra | 20:46 | |
| *** sputnik13 has joined #openstack-infra | 20:46 | |
| nibalizer | well this is kindof what you asked for http://paste.ubuntu.com/9532723/ | 20:47 |
| nibalizer | these acceptance tests haven't been merged yet | 20:47 |
| *** ihrachyshka has quit IRC | 20:47 | |
| nibalizer | and there are some lingering issues around rabbitmq | 20:47 |
| *** baoli has quit IRC | 20:47 | |
| *** shashankhegde has joined #openstack-infra | 20:47 | |
| nibalizer | but yea if you use the syntax that colleen suggested you do get beaker tests out the other end | 20:47 |
| *** baoli has joined #openstack-infra | 20:47 | |
| anteaya | here is my problem: Failed examples: | 20:48 |
| anteaya | rspec ./spec/acceptance/class_spec.rb:7 # storyboard class default parameters should work with no errors | 20:48 |
| anteaya | the repo you are testing has a file in spec/acceptance | 20:48 |
| anteaya | so the job ran but the test failed, fine | 20:48 |
| anteaya | puppet-storyboard has no spec/ directory | 20:48 |
| openstackgerrit | Colleen Murphy proposed openstack-infra/project-config: Run acceptance tests without rake https://review.openstack.org/141728 | 20:49 |
| anteaya | can you show me an example of what happens if you run the rspec command on a repo with no spec/ directory? | 20:49 |
| crinkle | anteaya: I've attempted to make the commit message more descriptive | 20:49 |
| nibalizer | my change adds the spec directory | 20:49 |
| anteaya | crinkle: thanks | 20:49 |
| *** teran has joined #openstack-infra | 20:49 | |
| anteaya | 141728 adds a spec directory? | 20:50 |
| * anteaya looks again | 20:50 | |
| crinkle | 141728 corrects the command | 20:50 |
| nibalizer | http://paste.ubuntu.com/9532745/ | 20:50 |
| crinkle | running rspec spec/acceptance without that directory will fail, but this is a nonvoting job and that is expected | 20:51 |
| nibalizer | it fails, but the test is nonvoting so thats not a big deal | 20:51 |
| anteaya | ah ha | 20:51 |
| anteaya | that was the missing piece | 20:51 |
| crinkle | we intend to keep it nonvoting until we can update the modules to have acceptance tests, but we need the infrastructure in place first | 20:51 |
| anteaya | please state in the commit message that the expected outcome of this patch is that the job will fail | 20:52 |
| crinkle | okay | 20:52 |
| anteaya | and btw I'm adverse to merging jobs that will fail | 20:52 |
| anteaya | so now you have to convince me why I should merge a patch that will cause a job to fail | 20:52 |
| *** vigneshvar_ has quit IRC | 20:52 | |
| *** teran_ has joined #openstack-infra | 20:52 | |
| anteaya | or sorry, +2 a patch | 20:53 |
| nibalizer | well the job alraedy fails? | 20:53 |
| crinkle | this was causing jobs to fail anyway, see https://review.openstack.org/#/c/140194/ | 20:53 |
| anteaya | crinkle says running rspec spec/acceptance without a spec directory will cause the job to fail | 20:53 |
| anteaya | which was my suspicion | 20:53 |
| crinkle | it was failing with rake spec spec/acceptance too | 20:53 |
| *** erikwilson has joined #openstack-infra | 20:54 | |
| nibalizer | are you saying you would like some logic to be added so that it 'passes' the test if there are no test files to run? | 20:54 |
| crinkle | the purpose of 141728 is so that when the acceptance tests are in place, the correct tests will be run | 20:54 |
| crinkle | otherwise the tests won't get run at all | 20:54 |
| anteaya | well that would be my utopia | 20:54 |
| crinkle | we don't really want acceptance tests to pass if there are no acceptance tests | 20:55 |
| *** teran has quit IRC | 20:55 | |
| anteaya | fair point | 20:55 |
| *** dprince has quit IRC | 20:55 | |
| anteaya | any way of adding a message to the failure stating that the reason it failed is due to absence of the expected directory of tests? | 20:55 |
| crinkle | in nibalizer's paste it says "cannot load such file" | 20:56 |
| crinkle | which seems clear to me | 20:56 |
| anteaya | and that loaderror comes from rspec? | 20:57 |
| *** mtanino has joined #openstack-infra | 20:57 | |
| crinkle | yes | 20:57 |
| *** sputnik13 has quit IRC | 20:57 | |
| anteaya | okay thanks | 20:57 |
| jogo | fungi: there is something wrong with status.openstack.org/elastic-recheck/ it hasn't been updating | 20:57 |
| jogo | I am running some of the scripts locally to see if anything is broke on that side | 20:58 |
| *** kgiusti has quit IRC | 20:58 | |
| openstackgerrit | Weidong Shao proposed openstack-infra/project-config: Grant push tag privilege to compass-core group https://review.openstack.org/141903 | 20:58 |
| jogo | but haven't found anything yet (just finished running the unclassified tool) | 20:58 |
| jogo | is it possible to get a copy of the logs | 20:58 |
| fungi | jogo: takign a look now | 20:58 |
| jogo | fungi: thanks | 20:58 |
| *** prad has joined #openstack-infra | 20:58 | |
| *** otter768 has joined #openstack-infra | 20:59 | |
| *** doug-fish has quit IRC | 20:59 | |
| fungi | jogo: hrm... 102 processes stretching back 6 hours or more | 21:00 |
| fungi | jogo: i think the flock isn't right | 21:00 |
| openstackgerrit | Colleen Murphy proposed openstack-infra/project-config: Run acceptance tests without rake https://review.openstack.org/141728 | 21:00 |
| jogo | flock? | 21:00 |
| *** Ng has quit IRC | 21:00 | |
| *** david-ly_ has joined #openstack-infra | 21:00 | |
| fungi | jogo: the update is run under a flock on /var/lib/elastic-recheck/er_safe_run.lock | 21:01 |
| fungi | file lock | 21:01 |
| *** dmellado has quit IRC | 21:01 | |
| *** ddieterl_ has joined #openstack-infra | 21:01 | |
| *** baoli has quit IRC | 21:01 | |
| *** dangers_away has quit IRC | 21:01 | |
| *** marcusvrn has quit IRC | 21:02 | |
| *** esker has quit IRC | 21:02 | |
| *** adalbas has quit IRC | 21:02 | |
| *** Swanson has quit IRC | 21:02 | |
| *** grue_pm_ has quit IRC | 21:02 | |
| *** lifeless has quit IRC | 21:02 | |
| *** gingerjiang has quit IRC | 21:02 | |
| *** vipul has quit IRC | 21:02 | |
| *** ekarlso- has quit IRC | 21:02 | |
| fungi | jogo: the mail spool is full of tracebacks about timeouts querying logstash too | 21:02 |
| *** ddieterly has quit IRC | 21:02 | |
| *** david-lyle has quit IRC | 21:02 | |
| *** ctlaugh has quit IRC | 21:02 | |
| *** yjiang5 has joined #openstack-infra | 21:02 | |
| anteaya | nibalizer crinkle what would happen if you merged the patch that creates the spec/ directory first? | 21:02 |
| *** tkelsey has quit IRC | 21:03 | |
| jogo | fungi: ahh maybe we need some sort of timeout for that job? | 21:03 |
| *** ctlaugh has joined #openstack-infra | 21:03 | |
| *** dangers_away has joined #openstack-infra | 21:03 | |
| jogo | say an hour? | 21:03 |
| openstackgerrit | Wayne Warren proposed openstack-infra/jenkins-job-builder: Add plugins_info to module registry object. https://review.openstack.org/132927 | 21:03 |
| openstackgerrit | Wayne Warren proposed openstack-infra/jenkins-job-builder: Update 'timeout' wrapper module https://review.openstack.org/129467 | 21:03 |
| *** baoli has joined #openstack-infra | 21:03 | |
| fungi | jogo: well, not sure yet. i'm going to clean up the process pileup first and then see what's actually broken | 21:03 |
| *** otter768 has quit IRC | 21:03 | |
| *** Longgeek has quit IRC | 21:03 | |
| *** erikmwilson has quit IRC | 21:03 | |
| dhellmann | fungi: looks like picking a different branch name introduced an issue with the requirements check: https://jenkins06.openstack.org/job/gate-oslo.db-requirements/45/console | 21:03 |
| crinkle | anteaya: well we wouldn't be able to verify that the test works, since it would be running a different test between when we merge that and when we merge the -config patch | 21:04 |
| *** grue_pm has joined #openstack-infra | 21:04 | |
| *** lifeless has joined #openstack-infra | 21:04 | |
| crinkle | but that's not my patch so nibalizer ^ | 21:04 |
| jogo | fungi: cool, local tests didn't find anything failing | 21:04 |
| anteaya | crinkle: okay I wondered if that was the case, but thought I'd ask | 21:05 |
| anteaya | thanks | 21:05 |
| *** erikwilson has quit IRC | 21:06 | |
| dhellmann | fungi: I don't think we want a feature/1.2 branch on the global requirements list; I wonder if we should go back to the idea of bringing stable/juno up to date to 1.2.0 | 21:07 |
| fungi | dhellmann: oh, that's a good point about the requirements check jobs | 21:08 |
| dhellmann | how did you get around this over the weekend? | 21:08 |
| fungi | clarkb: do you have any thoughts on how we could safely update oslo.db's stab;e/juno branch from 1.0 plus backported patches to 1.2.0? | 21:09 |
| clarkb | merge 1.2.0 into it | 21:09 |
| nibalizer | anteaya: hrm? | 21:09 |
| fungi | dhellmann: the fixes i was trying to backport in pbr were not reqs changes, and i was only rewinding to the most recent release tag | 21:09 |
| clarkb | like a feature branch | 21:09 |
| fungi | clarkb: yeah, that's what i figured we'd need to do to retain ff-only safeness | 21:09 |
| *** vipul has joined #openstack-infra | 21:10 | |
| asselin | mmedvede, when do you think this will merge? https://review.openstack.org/#/c/137156/ | 21:10 |
| fungi | clarkb: however, that means that we then release a 1.2.1 which is based on a 1.2.0-like history with a merge commit | 21:10 |
| anteaya | reviewed and commented | 21:10 |
| fungi | clarkb: so not entirely sold on whether that's a clean process to follow, but maybe so | 21:10 |
| anteaya | nibalizer: crinkle answered me | 21:10 |
| clarkb | fungi I think the merge commit is required for fast forwarding | 21:11 |
| nibalizer | okay | 21:11 |
| *** sputnik13 has joined #openstack-infra | 21:11 | |
| fungi | clarkb: the other question is, if this is likely to crop up for other libraries in the future, how do we accommodate it repeatably? | 21:11 |
| jeblair | clarkb, fungi: https://bitbucket.org/pypa/bandersnatch/pull-request/12/add-option-to-dir-hash-index-files/diff | 21:11 |
| *** erikmwilson has joined #openstack-infra | 21:11 | |
| mattoliverau | Morning | 21:12 |
| clarkb | fungi I think we dont stable/juno instead we stable/x | 21:12 |
| dhellmann | fungi, clarkb : the only conflicts I get when I merge are in the requirements lists. Do I need to make this a special kind of merge, or can i just "git merge 1.2.0" make the fixes, then commit? | 21:12 |
| clarkb | maybe but that is hard for other reasons | 21:12 |
| fungi | jeblair: is that the proposed solution to the afs node count limit? | 21:12 |
| *** erikmwilson has quit IRC | 21:12 | |
| *** adalbas has joined #openstack-infra | 21:12 | |
| jeblair | fungi: yep. ran that on afstest and 'vos release', so what's there is the output | 21:13 |
| dhellmann | clarkb: like stable/1.2? I like the idea of that | 21:13 |
| *** amcrn has quit IRC | 21:13 | |
| dhellmann | we could call it something other than stable, too, if we need to do that | 21:13 |
| clarkb | dhellmann ya I need to think about that more but I think thats better | 21:13 |
| fungi | clarkb: stable/foo with foo not in icehouse,juno right now breaks the devstack branch matrix, yeah? | 21:13 |
| clarkb | fungi yes so would need some work | 21:13 |
| fungi | k | 21:13 |
| *** dpaterson has joined #openstack-infra | 21:13 | |
| fungi | was pretty sure the feature matrix barfed when i tried that with pbr over the weekend | 21:14 |
| clarkb | yup | 21:14 |
| *** erikwilson has joined #openstack-infra | 21:14 | |
| *** erikwilson is now known as erikmwilson | 21:15 | |
| dhellmann | we still have the problem with the requirements compat job though if we don't call the branch stable/juno -- because the job assumes the branch of the lib and the branch of the requirements list are the same | 21:15 |
| *** ekarlso- has joined #openstack-infra | 21:15 | |
| dhellmann | fungi, clarkb : the only conflicts I get when I merge are in the requirements lists. Do I need to make this a special kind of merge, or can i just "git merge 1.2.0" make the fixes, then commit? | 21:15 |
| dhellmann | I haven't pushed a merge commit to gerrit before | 21:15 |
| clarkb | dhellmann I think making fixes then committing is what you want | 21:15 |
| *** Swanson has joined #openstack-infra | 21:16 | |
| mmedvede | asselin: the https://review.openstack.org/#/c/137156/ is ready to go in whenever. jeblair, are you still willing to shepherd the elasticsearch module split? It has a couple of +2's now. | 21:16 |
| anteaya | mattoliverau: morning | 21:16 |
| *** esker has joined #openstack-infra | 21:16 | |
| *** Sukhdev has joined #openstack-infra | 21:17 | |
| *** marcusvrn has joined #openstack-infra | 21:17 | |
| asselin | mmedvede, yes, lgtm. I'll rebase my changes on top of that. | 21:18 |
| *** Ng has joined #openstack-infra | 21:19 | |
| jeblair | mmedvede: yep, aprvd | 21:19 |
| *** baoli has quit IRC | 21:20 | |
| mmedvede | jeblair: thank you. I have rechecked, the seed repository is in sync with system-config | 21:21 |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: DO NOT MERGE: testing aiopcpu with tempest full https://review.openstack.org/136504 | 21:24 |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: Set custom cpu_model for live_migrate https://review.openstack.org/141530 | 21:24 |
| jeblair | fungi, clarkb, dhellmann: i believe we always intended to create stable/version-number branches for library projects with no stable/releasename branches | 21:24 |
| jeblair | fungi, clarkb, dhellmann: in the case that we needed them to do a point release | 21:25 |
| jesusaurus | im have some trouble with my nodepool service, it doesn't seem to be getting the events that jenkins is using a node (the node never leaves the "ready" state despite tests being run). how can i best troubleshoot this pipeline? | 21:25 |
| jeblair | fungi, clarkb, dhellmann: we may have forgotten about that when reviewing the feature matrix change | 21:25 |
| dhellmann | jeblair: we've been using stable/name branches for that so far | 21:25 |
| dhellmann | in this case, we need to jump several releases up, though | 21:26 |
| jeblair | dhellmann: yeah, i meant for projects with no stable/releasename branch... | 21:26 |
| jeblair | dhellmann: for projects with one -- why do you need a 1.2 branch that is not juno? | 21:26 |
| dhellmann | jeblair: ok. The requirements job breaks on those because there's no feature/1.2 branch in the requirements project | 21:26 |
| dhellmann | the juno version of oslo.db was 1.0 and we have had a couple of point releases | 21:27 |
| dhellmann | we need a version of the lib with the requirements for stable/juno's version of sqlalchemy | 21:27 |
| dhellmann | since master and stable/juno don't agree on those requirments, it needs to be a version < 1.3.0 | 21:27 |
| dhellmann | 1.2.0 was the last release that worked with juno, so... | 21:28 |
| *** sarob has joined #openstack-infra | 21:28 | |
| jeblair | so master of oslo.db is 1.3, juno is 1.0, however, juno versions of projects are actually using 1.2? | 21:29 |
| dhellmann | jeblair: 1.2 does work with juno versions of project (we just released 1.3 today) | 21:29 |
| fungi | jeblair: the "stable/juno" branch was used to backport fixes to 1.0 which was the most recent version at the time of the juno release | 21:29 |
| dhellmann | right | 21:29 |
| *** atiwari has quit IRC | 21:29 | |
| fungi | but apparently there is a desire for the stable/juno branch to track master until master ceases to work with stable/juno | 21:30 |
| fungi | except that we also want to backport patches to it that we haven't released from master yet? | 21:30 |
| dhellmann | fungi, clarkb : git review says "remote rejected] HEAD -> refs/publish/stable/juno/bug/1394298 (you are not allowed to upload merges)" | 21:30 |
| fungi | that's the part i've been unable to reconcile with the use case | 21:30 |
| fungi | dhellmann: right, that will need an acl adjustment | 21:31 |
| jeblair | is this another situation that would suggest that we cap or freeze dependencies in stable branches when they are cut? | 21:31 |
| dhellmann | fungi: I'm not sure what you mean about backporting unreleased fixes? | 21:31 |
| dhellmann | jeblair: we did try that, and that prevents grenade from upgrading a deployment so we undid it | 21:31 |
| *** jedimike has joined #openstack-infra | 21:31 | |
| fungi | dhellmann: you said you backported fixes (from somewhere) to the stable/juno branch of oslo.db, but that oslo.db should work with master (up until it didn't as of 1.3.0) | 21:32 |
| dhellmann | jeblair: now, we could have tried a higher cap, but neither sdague nor I had the time to work out what those numbers should be | 21:32 |
| *** ihrachyshka has joined #openstack-infra | 21:32 | |
| fungi | er, juno should work with oslo.db master up until 1.3.0 i mean | 21:32 |
| jeblair | because the straightforward resolution would seem to me to be to release 1.0.X to fix juno, but that is complicated by the fact that juno is really using 1.2, though unspecified. | 21:32 |
| dhellmann | fungi: we follow the usual backport process, though, so those changes have to be in oslo.db master before they go into a branch | 21:32 |
| jedimike | clarkb, hi, we're having a problem with jenkins and the zmq publisher. Jenkins is running, jobs are running and passing/failing, but it's not opening port 8888 for nodepool to connect to. Any ideas where we should start looking? | 21:33 |
| dhellmann | jeblair: right, and master is now >=1.1 and so we can't have a cap in juno lower than that | 21:33 |
| fungi | dhellmann: right, i think the multiple use cases being expressed may be irreconcilible. still trying to think through the interactions there | 21:33 |
| dhellmann | k | 21:33 |
| jeblair | jedimike, jesusaurus: we might have seen the zmq plugin in jenkins fail; try disable/re-enable zmq publishing in jenkins, and if that doesn't work, restart jenkins | 21:34 |
| jeblair | jedimike, jesusaurus: we don't really have any data on whether or how often this happens because it is hard to debug | 21:34 |
| jedimike | jeblair, ok we'll try enable/disable, we've just restarted and the publisher didn't come back up | 21:34 |
| jesusaurus | jeblair: thanks | 21:34 |
| jeblair | jedimike, jesusaurus: also try "zmq-stream" in nodepool/tools to help debug | 21:35 |
| fungi | dhellmann: i think what we want is to be able to branch from any release of a library to backport fixes. having a rolling branch to aggregate those is going to get messy | 21:35 |
| dhellmann | I tried to summarize things in http://lists.openstack.org/pipermail/openstack-dev/2014-December/053021.html | 21:35 |
| anteaya | dhellmann: so just so this doesn't get lost, this is dougwig's script for splitting out advanced services from neutron last week, based on your script for oslo: https://github.com/dougwig/split-repo/blob/master/service-split.sh | 21:35 |
| dhellmann | fungi: I agree | 21:35 |
| fungi | dhellmann: however, that means that we can no longer di strict branch name mapping in requirements checks | 21:35 |
| dhellmann | fungi: right - and interestingly we need oslo.db 1.2 to be compatible with the requirements of juno, but it might not be :-/ | 21:36 |
| fungi | and i'm not immediately seeing how we can continue doing global-requirements enforcement on changes to requirements files in libraries | 21:36 |
| dhellmann | oh, wait, I guess it has to be so far because it has been working -- forget that | 21:37 |
| fungi | unless we somehow embed additional metadata in the release about what version of openstack it should be tested against | 21:37 |
| dhellmann | yeah, for a 1.2 branch, do we check against juno or kilo? | 21:37 |
| fungi | er, in the branch from which we're going to release the backport point versions | 21:37 |
| *** dkliban is now known as dkliban_bbr | 21:37 | |
| *** dkliban_bbr is now known as dkliban_brb | 21:38 | |
| *** doug-fish has joined #openstack-infra | 21:38 | |
| dhellmann | yeah, we could easily put a file in the branch and have the job read it -- but what would the file say in this case? | 21:38 |
| *** teran has joined #openstack-infra | 21:38 | |
| fungi | something like "stable/juno" i think | 21:38 |
| dhellmann | so that means we would have to change all of the requirements of 1.2 to match the requirements of stable/juno to get the job to pass | 21:38 |
| jedimike | jeblair, thanks, we're doing the disable/enable thing now | 21:38 |
| dhellmann | because for 1.2 we have already bumped several requirements | 21:38 |
| fungi | we would presumably want to treat the stable/1.2 branch like part of juno for testing purposes | 21:38 |
| dhellmann | yeah | 21:39 |
| fungi | which we can't retroactively solve at this point, right | 21:39 |
| jeblair | it seems like 1.2 is part of juno... | 21:39 |
| dhellmann | semver would tell us that release number should be 1.3, since it has requirements changes :-| | 21:39 |
| jeblair | because juno says >1.0 and 1.2 matches. | 21:39 |
| fungi | jeblair: except it sounds like 1.2 has requirements which are newer than the stable/juno requirements list | 21:39 |
| jeblair | sorry >=1.0 | 21:39 |
| dhellmann | well, 1.3 is > 1.0 but 1.3 doesn't work | 21:40 |
| dhellmann | 1.2 does have newer requirements -- I don't know if the older ones would work, I only updated sqlalchemy | 21:40 |
| fungi | if we'd already had this "should work with juno" flag in place somewhere that we could match on, that would allow us to force it to be tested against the juno reqs list | 21:40 |
| *** _nadya_ has quit IRC | 21:41 | |
| fungi | and refuse addition of post-juno requirements until the developers had decided to move on to a breaking version | 21:41 |
| * dhellmann wonders how different the 1.1 requirements are from juno | 21:41 | |
| *** teran_ has quit IRC | 21:41 | |
| fungi | but this of course means that we need to plan in advance when we will start developing things which won't work with the last release, which i suspect we've been slack about before now | 21:41 |
| jeblair | dhellmann: how does putting a version cap on the old side prevent an upgrade? | 21:41 |
| jogo | fungi: thanks for fixing http://status.openstack.org/elastic-recheck | 21:42 |
| *** baoli has joined #openstack-infra | 21:42 | |
| dhellmann | jeblair: grenade fails, but I forget the exact scenario | 21:42 |
| fungi | jeblair: it breaks upgrading _if_ the new side also declares a minimum version greater than the maximum version supported by the previous release | 21:42 |
| dhellmann | yeah, I don't know if grenade is trying to update the libraries first and then the apps, or one app at a time, or what | 21:43 |
| openstackgerrit | Merged openstack-infra/project-config: Add puppet-elasticsearch as split out module https://review.openstack.org/137156 | 21:43 |
| jeblair | fungi: that's confusing. :( | 21:43 |
| fungi | jogo: all i did was kill a bunch of copies of e-r query update processes on status.o.o and let cron start new ones | 21:43 |
| jeblair | dhellmann: but yeah, i could see how that approach would break | 21:43 |
| dhellmann | jeblair: yeah, I was disappointed since I thought we had worked this out at the summit :-( | 21:44 |
| jogo | fungi: do you think it makes sense to add a timeout to those jobs? if so how would I go about adding it? | 21:44 |
| fungi | jeblair: the upgrade problem arises if your consecutive server releases have incompatible requirements, since you would potentially have to upgrade multiple services at the same time | 21:44 |
| dhellmann | jeblair: it has something to do with the way we test rolling upgrades by only upgrading part of the stack and then trying to do things with services | 21:44 |
| *** dmellado has joined #openstack-infra | 21:44 | |
| dhellmann | it looks like oslo.db 1.1 has some requirements bumps, too | 21:45 |
| jeblair | i sort of doubt that is going to work unless we split services into venvs | 21:45 |
| fungi | jogo: https://git.openstack.org/cgit/openstack-infra/system-config/tree/modules/openstack_project/manifests/pypi_mirror.pp#n64 | 21:46 |
| dhellmann | jeblair: yeah, that's what we did at Dreamhost because we didn't want to fight this fight | 21:46 |
| jeblair | (which i think strong requirements enforcement lets us consider doing since we can guarantee installability with one set of packages with that instead of the side effect of a devstack install) | 21:47 |
| *** atiwari has joined #openstack-infra | 21:47 | |
| jhesketh | Morning | 21:47 |
| *** armax has joined #openstack-infra | 21:47 | |
| anteaya | morning jhesketh | 21:48 |
| jogo | fungi: thanks | 21:48 |
| fungi | jogo: i would recommend moving the flock out of the wrapper script and just call it directly from cron like i did for bandersnatch | 21:48 |
| *** ryanpetrello has joined #openstack-infra | 21:48 | |
| jeblair | dhellmann: something i'm still not groking though -- why not release 1.0.X and cap stable/juno to that? | 21:49 |
| dhellmann | jeblair: master needs >=1.1 at this point | 21:50 |
| dhellmann | that means there would be no overlapping compatible requirement, so grenade fails to upgrade us | 21:50 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/elastic-recheck: Add another failure pattern for 1402747 https://review.openstack.org/141912 | 21:50 |
| *** openstackgerrit has quit IRC | 21:50 | |
| *** openstackgerrit has joined #openstack-infra | 21:50 | |
| dhellmann | because when we created stable/juno we didn't foresee needing to release a new version with a change in requirements, only a bug fiix | 21:50 |
| *** dhp has joined #openstack-infra | 21:50 | |
| dhellmann | jeblair: can we cap setuptools in the stable/juno test environment? | 21:51 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/elastic-recheck: Add another failure pattern for bug 1402747 https://review.openstack.org/141912 | 21:51 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [High,Confirmed] https://launchpad.net/bugs/1402747 | 21:51 |
| dstufft | it doesn't generally work to cap the version of the installer you want to be installed with | 21:52 |
| *** jlibosva has quit IRC | 21:52 | |
| dstufft | because if someone has setuptools 8 installed, and you add a cap on setuptools < 8, they'll install everything with setuptools 8 and as part of that downgrade their setuptools | 21:52 |
| jeblair | dhellmann, fungi: i'm kind of stuck on "how does grenade ever work if it assumes that there will be compatible overlapping versions of all python libraries" | 21:53 |
| fungi | dhellmann: we explicitly don't have a requirement on setuptools because there's a chicken-and-egg problem with declaring you need an older version than you're running | 21:53 |
| dhellmann | jeblair: I don't know. sdague may | 21:54 |
| jeblair | that just seems like a requirement that i would expect to be impossible to satisfy | 21:54 |
| fungi | dhellmann: when we "pinned" it on our workers, we did so by changing the version we preinstall | 21:54 |
| openstackgerrit | Ramy Asselin proposed openstack-infra/project-config: Prepare project-config for puppet module split #2 https://review.openstack.org/140548 | 21:54 |
| dhellmann | fungi: and I guess we use the same workers for juno and master? | 21:54 |
| fungi | dhellmann: yes | 21:54 |
| *** emagana has joined #openstack-infra | 21:55 | |
| jogo | fungi: ack | 21:56 |
| fungi | dhellmann: also it's only feasible for the moment because we're running with virtualenv/pip which haven't updated to embed newer versions of setuptools yet | 21:56 |
| jeblair | sdague: what am i missing regarding grenade ^ ? | 21:56 |
| sdague | jeblair: until recently we had compatible overlaps | 21:57 |
| sdague | because we kept all of stable unpinned | 21:57 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/elastic-recheck: Add another failure pattern for 1402747 https://review.openstack.org/141912 | 21:58 |
| jeblair | sdague: but doesn't that assume that we will never have a third-party dependency that breaks backwards compat between a release we use in juno and a release we use in master? | 21:58 |
| *** jlibosva has joined #openstack-infra | 21:58 | |
| fungi | sdague: which as it turns out is misleading, because we test changes to the libraries with requirements master but expect stable release branches which don't follow the same global requirements set to continue working with them | 21:58 |
| dhellmann | jeblair, fungi, sdague: would it help us to revert dstufft's patch changing the version of sqlalchemy we use in master to match what we use in juno? http://git.openstack.org/cgit/openstack/requirements/commit/?id=f02eb22ad8a5a18835f5755c619a25fbf24a2453 | 21:58 |
| dhellmann | because then I could release a version of 1.2 that tests against master requirements | 21:59 |
| sdague | jeblair: possibly, but that's the system we had. I will in no way say it's the best system, but it caught a bunch of bugs in OpenStack | 21:59 |
| jeblair | sdague: it seems like a normal grenade job should be able to upgrade from capped stable requirements to uncapped master requirements no matter the leap. and perhaps the rolling-upgrade jobs need to do something different | 22:01 |
| sdague | jeblair: so... sorry, rebuilding context | 22:01 |
| jeblair | sdague: like install into a virtualenv so that they can similarly make the leap, but on a service-by-service basis | 22:01 |
| sdague | a normal grenade can do that | 22:01 |
| sdague | except.... | 22:01 |
| *** yamamoto has joined #openstack-infra | 22:01 | |
| cody-somerville | zaro: ping | 22:01 |
| sdague | things like ceilometer install into other packages paste pipelines | 22:02 |
| sdague | so our normal offline upgrade model breaks, because we assumed that we could shut everything down | 22:02 |
| sdague | upgrade and start services in a reasonable order | 22:02 |
| *** mrunge has quit IRC | 22:02 | |
| jeblair | would a process of 'shut all down; upgrade all; start all' address that? | 22:03 |
| sdague | but when one software package installs into the execution space of another, then there is a coupling that wasn't known until recently (there is an ML thread on this) | 22:03 |
| sdague | yep | 22:03 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/elastic-recheck: Expand failure pattern for bug 1402747 https://review.openstack.org/141912 | 22:03 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [High,Confirmed] https://launchpad.net/bugs/1402747 | 22:03 |
| *** mattfarina has quit IRC | 22:03 | |
| sdague | except, that's a whole bunch of rework | 22:03 |
| jeblair | sdague: is that path agreed upon? | 22:04 |
| *** dkliban_brb is now known as dkliban | 22:04 | |
| *** ildikov has quit IRC | 22:04 | |
| sdague | I don't know that there is any agreed on path here | 22:04 |
| sdague | also, got to end my day, call of the family | 22:05 |
| sdague | talk to folks later | 22:05 |
| fungi | it seems like this is partly an atrifact of our previous limitation of needing to be able to test upgrades on an all-in-one server environment, and what we really want to be able to test is that different versions of the software can work together successfully. so the need for coinstallability of versions from different releases is more of a symptom of our testing design? | 22:05 |
| dhellmann | that sounds right | 22:06 |
| *** sarob has quit IRC | 22:06 | |
| jeblair | fungi: well, that's what i was getting at earlier, but the ceilometer case still makes it different | 22:06 |
| dhellmann | ceilometer case? | 22:06 |
| *** mrmartin has quit IRC | 22:06 | |
| jeblair | what sdague just described -- ceilometer installing itself into other projects execution space | 22:06 |
| dhellmann | oh, I missed that | 22:06 |
| dhellmann | yeah, that makes sense | 22:06 |
| fungi | rather than interacting through well-defined interfaces | 22:06 |
| *** bswartz has quit IRC | 22:07 | |
| dhellmann | well, swift's interface for emitting notification was "add something to the wsgi pipeline" | 22:07 |
| dhellmann | because they won't add it directly to swift | 22:07 |
| jeblair | at any rate, it seems like the described method of altering grenade would solve the problem | 22:07 |
| *** ZZelle has joined #openstack-infra | 22:07 | |
| jeblair | i'm sort of sad that we don't know if anyone is working on it | 22:07 |
| dhellmann | at least, that's my understanding | 22:07 |
| *** stpierre has quit IRC | 22:07 | |
| dhellmann | all of that sounds like a long term fix, so here's a short term idea: | 22:08 |
| jeblair | because this seems like a pretty big design flaw that is certain to make things very unecessarily difficult | 22:08 |
| *** pelix has quit IRC | 22:08 | |
| jeblair | we're just "lucky" that it broke on libraries that we manage first | 22:08 |
| openstackgerrit | Wayne Warren proposed openstack-infra/jenkins-job-builder: Update 'timeout' wrapper module https://review.openstack.org/129467 | 22:09 |
| zaro | cody-somerville: yo! | 22:09 |
| *** emagana has quit IRC | 22:09 | |
| dhellmann | change the requirements check job to look at a file in the repo being tested to find the branch of the requirements repo to use; set up the feature/1.2 branch of oslo.db to refer to master in that file; change the master requirements for sqlalchemy to match stable/juno | 22:09 |
| dhellmann | (in the global list); then change feature/1.2 requirements to that same string; release 1.2.1 from that | 22:10 |
| dhellmann | jeblair: yes, indeed | 22:10 |
| jeblair | dtroyer: do you know if there are plans to rework grenade to stop all/upgrade all/start all ? | 22:10 |
| fungi | to put the current paradox in concrete terms, we're saying we expect stable/juno servers to work reliably with oslo.db 1.1.0 and 1.2.0 but we didn't actually test those oslo.db releases with the stable/juno requirements list (and instead declare moving-target requirements from master in those oslo.db releases) | 22:11 |
| *** david-ly_ is now known as david-lyle | 22:11 | |
| dhellmann | fungi, jeblair : let me know what you think of that short-term fix, and if you like it I'll work on patches in the morning | 22:11 |
| *** sputnik1_ has joined #openstack-infra | 22:11 | |
| jeblair | dtroyer: ^ see conversation with sdague above for more context | 22:12 |
| *** sputnik13 has quit IRC | 22:12 | |
| *** kmartin has quit IRC | 22:12 | |
| jeblair | dhellmann: is all of that in service of being able to land changes to feature/1.2 of oslo.db? | 22:12 |
| *** emagana has joined #openstack-infra | 22:12 | |
| fungi | dhellmann: if you change the requirements enforcement so that the relevant branch can be declared from a file in the project being tested, we'll probably also want to adjust the requirements sync in devstack to use that too | 22:12 |
| *** asselin has quit IRC | 22:12 | |
| *** kmartin has joined #openstack-infra | 22:13 | |
| *** mfink has quit IRC | 22:13 | |
| *** asselin has joined #openstack-infra | 22:13 | |
| dhellmann | jeblair: well, yes, so that we can have a 1.2.x that has requirements that allows it to be installed for juno tests so that we have a version of oslo.db that works with both juno and master so we can do grenade updates | 22:13 |
| dhellmann | jeblair: oh, I left out the cap oslo.db in juno to < 1.3 | 22:13 |
| jogo | fungi: man flock doesn't explain what "-k" does | 22:13 |
| jogo | oh thats because its the command timeout ... | 22:14 |
| dhellmann | fungi: oh, wow, yeah, probably | 22:14 |
| fungi | dhellmann: since down the road you'll want to be able to set oslo.db's stable/1.3 branch to test as if it were stable/kilo (so that new non-kilo-compatible requirements don't get added) | 22:14 |
| *** jlibosva has quit IRC | 22:14 | |
| jeblair | dhellmann: yeah, i'm just trying to figure out why you want to change the requirements check job -- what is it blocking? | 22:14 |
| *** ddieterl_ has quit IRC | 22:14 | |
| dhellmann | jeblair: http://logs.openstack.org/94/141894/1/check//gate-oslo.db-requirements/0a860bb/console.html | 22:14 |
| dhellmann | "warning: Could not find remote branch feature/1.2 to clone" | 22:15 |
| dtroyer | jeblair: no long term plans, but that is close to how it was orignally written…start all, stop all, upgrade one at a time and restart. some of that may have changed recently with the sideways jobs and whatnot | 22:15 |
| *** ryanpetrello has quit IRC | 22:15 | |
| jeblair | dhellmann: right, so it's blocking changes to the feature/1.2 branch of oslo.db | 22:15 |
| *** ryanpetrello_ has joined #openstack-infra | 22:16 | |
| dhellmann | jeblair: to any branch of any project that does not have a similar branch in openstack/requirements | 22:16 |
| dhellmann | but yeah, immediately it's to oslo.db | 22:16 |
| *** dustins_ has quit IRC | 22:16 | |
| *** ddieterly has joined #openstack-infra | 22:16 | |
| *** ryanpetrello_ is now known as ryanpetrello | 22:16 | |
| dhellmann | it means I couldn't have a feature branch that changed requirements, though | 22:16 |
| *** ihrachyshka has quit IRC | 22:17 | |
| jeblair | dhellmann: feature branches should probably fall back on master requirements | 22:17 |
| dhellmann | I don't know. Short term I'd be happy to cap oslo.db<1.1 in stable/juno but I don't think that's a long term fix. | 22:17 |
| dhellmann | ok, we could make the script that smart I guess | 22:17 |
| *** yfried is now known as yfried|afk | 22:17 | |
| dhellmann | rather than specifying it in a file, if the branch isn't found and the name starts with "feature" use master? | 22:17 |
| *** doug-fish has quit IRC | 22:18 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/elastic-recheck: Expand failure pattern for bug 1402747 https://review.openstack.org/141912 | 22:18 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [High,Confirmed] https://launchpad.net/bugs/1402747 | 22:18 |
| jeblair | dhellmann: yeah, in devstack-gate for instance we fall back on cross-testing with master | 22:18 |
| *** annegen__ has quit IRC | 22:18 | |
| *** sputnik1_ is now known as sputnik13 | 22:19 | |
| jeblair | dtroyer: so i think the fact that we have to have compatible versions of dependencies across major openstack versions is a harmful artificial restriction | 22:19 |
| dhellmann | ok, that would be a simpler change | 22:19 |
| jeblair | dtroyer: i don't think it was or should be the intended effect of grenade, but it seems like it is in the status quo | 22:19 |
| dtroyer | it is an artifact of the case of running with mixed releases of things, specifically n-cpu on a separate upgrad path introduced that | 22:20 |
| jeblair | dtroyer: if that's the only hangup, then i think we need to go back to the drawing board on how that test is implemented | 22:20 |
| *** ddieterl_ has joined #openstack-infra | 22:20 | |
| dtroyer | So it sounds like it is time to start planning son-of-devstack-and-grenade to split the tasks up more clearly, and get as much into venvs from the start as possible | 22:20 |
| *** ddieterly has quit IRC | 22:20 | |
| *** ildikov has joined #openstack-infra | 22:21 | |
| *** ddieterl_ has quit IRC | 22:21 | |
| jeblair | dtroyer: yeah, i think if we want to have isolated environment upgrades, we'll have to do something like that. that way each environment can do any upgrade it needs (including jumping major incompatible versions) | 22:21 |
| *** teran_ has joined #openstack-infra | 22:22 | |
| *** ddieterly has joined #openstack-infra | 22:22 | |
| jeblair | dtroyer: i think that the requirements jobs leave us in a position where we can still say that at the end of the day, the projcets will all converge on the same set of dependencies at the end of the upgrade | 22:22 |
| *** ddieterly has joined #openstack-infra | 22:22 | |
| jeblair | dhellmann: also, removing the requirements check job from the feature/1.2 branch of oslo.db is an option | 22:23 |
| *** Masahiro has joined #openstack-infra | 22:23 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/elastic-recheck: Expand failure pattern for bug 1402747 https://review.openstack.org/141912 | 22:23 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [High,Confirmed] https://launchpad.net/bugs/1402747 | 22:23 |
| jeblair | dhellmann: basically, i don't want to make the rube-goldberg requirements machinery any more complex than it already is; i feel like in the long run, if oslo.db has a stable/juno branch, that's where we ought to be able to make these kinds of fixes, so i'd like the long-term solutions to involve making that happen | 22:24 |
| dhellmann | jeblair, dtroyer : mordred did pitch the idea of a single job that installed everything and ran some basic tests, which would ensure that those requirements sort of work together | 22:24 |
| jeblair | ie, fixing grenade so that we can do that. | 22:24 |
| jeblair | dhellmann: ++ | 22:24 |
| *** teran has quit IRC | 22:25 | |
| jeblair | dhellmann: (though i do still believe that feature branches should fall back to master in requirements check jobs anyway) | 22:25 |
| jeblair | (regardless of _this_ problem) | 22:25 |
| dhellmann | jeblair: yeah, unfortunately our stable/juno branch is pretty far out of date by now. We did discuss merging 1.2 back into stable/juno, but that opens another can of worms with requirements changes like this because then I have to have 1.2.1 use different requirements from 1.2.0, which is not only a semver violation but also confusing | 22:25 |
| dhellmann | yeah, falling back to master totally makes sense | 22:25 |
| *** dkranz has quit IRC | 22:26 | |
| dhellmann | turning off the check job for that branch is ok with me | 22:26 |
| dhellmann | it means I need to think harder about what requirements update to make, though | 22:26 |
| openstackgerrit | Joe Gordon proposed openstack-infra/system-config: Improve er_safe_run https://review.openstack.org/141923 | 22:26 |
| jogo | fungi: ^ | 22:26 |
| *** ChuckC has quit IRC | 22:26 | |
| dhellmann | jeblair: I think we want 1.2.1 to have the same requirements as master | 22:26 |
| dhellmann | jeblair: but what I have up for review now is the same requirements as juno, so I need to spend more time with a fresh brain on that | 22:27 |
| jogo | fungi: I left the script in because I don't think that was the issue. I think you saw a bunch of them due to flock blocking by default | 22:27 |
| jeblair | so my thought process is: fixing grenade lets us cap stable branches which would make stable/juno branches of libraries useful and thus avoid this problem. | 22:27 |
| *** jedimike has quit IRC | 22:27 | |
| jeblair | (for the long term fix) | 22:27 |
| *** Masahiro has quit IRC | 22:28 | |
| dhellmann | yes | 22:28 |
| * dhellmann looks around for someone with time to rewrite grenade | 22:28 | |
| *** bhunter71 has quit IRC | 22:29 | |
| jeblair | i'm hoping that jogo will work with dtroyer on that since jogo did a lot of the n-cpu upgrade work | 22:29 |
| mtreinish | jeblair: heh, good luck with that :) | 22:29 |
| *** erlon has quit IRC | 22:29 | |
| jeblair | dhellmann: so if you want the reqs check job to run to continue to provide safety for your 1.2 work, then maybe doing the feature fallback to master is the way to go | 22:30 |
| jeblair | mtreinish: why? | 22:30 |
| anteaya | jeblair: I do believe capping stable branches is something ttx wants, inasmuch as I have agreed to work on a spec for it | 22:30 |
| * jogo reads backlog | 22:30 | |
| jeblair | jogo: would you be willing to help out with thatL | 22:30 |
| anteaya | re-writing grenade will have to be part of the spec? | 22:30 |
| dhellmann | jeblair: well, what I really need is to know that devstack-gate works with those requirements, so we can turn off the requirements job | 22:30 |
| mtreinish | jeblair: take a look at the grenade git logs, commits are far and few between | 22:31 |
| anteaya | sweston: do check your email for a response from mikal about radar | 22:31 |
| jogo | jeblair: help out with what? | 22:31 |
| jeblair | mtreinish: yeah, but what we have is grenade being used in a way it was note designed for, and the unintended side effect is that we are going to _seriously_ shoot ourselves in the foot with that. | 22:31 |
| jeblair | mtreinish: we _might_ be able to dig ourselves out of it this time because we control all the parts involved. we may not be as lucky next time | 22:32 |
| jogo | jeblair: and it sounds like I should say yes to whatever it is | 22:33 |
| mtreinish | jeblair: I'm not disagreeing, I'm just saying that people with grenade knowledge are often busy with 10 other things, so someone spending the time to do a rewrite like that is going to be a long way off | 22:33 |
| mtreinish | hopefully I'm wrong though | 22:33 |
| * anteaya notes jogo just said yes | 22:33 | |
| jeblair | jogo: so grenade currently requires that there be a version of every dependency that can work with both the old and new sides of the upgrade | 22:33 |
| jogo | jeblair: right, we don't re-install deps in grenade | 22:34 |
| jeblair | jogo: so that means deps can never change, which is clearly not what we desire | 22:34 |
| dhellmann | jogo: yeah, with the new setuptools and subsequent changes to the sqlalchemy requirements specification, that's now causing a problem for things that use oslo.db | 22:34 |
| *** yfried|afk is now known as yfried | 22:35 | |
| mtreinish | jeblair: they can just in N+1 intervals. You just have to have overlap between releases. So you slowly ratchet up the min version | 22:35 |
| jogo | jeblair: agreed, I think its reasonable to tell people to upgrade python deps at the same time as upgrading an openstack service | 22:35 |
| dhellmann | jeblair: hrm, what if we make master sqlalchemy the same as juno, and then release oslo.db 1.4 that uses those settings, then update juno to force it to skip 1.1, 1.2, and 1.3? | 22:36 |
| *** dizquierdo has joined #openstack-infra | 22:36 | |
| dhellmann | jeblair: because oslo.db will work with those versions of sqlalchemy; I don't know why we changed the settings over the weekend | 22:36 |
| jogo | jeblair: so what is the new desired grenade behavior? install new version deps during the pull up phase of grenade | 22:36 |
| dhellmann | jogo: if the point of grenade is to test that individual services can be updated separately, those services could be installed into isolated virtualenvs to allow the dependencies to be updated, too | 22:37 |
| jogo | dhellmann: well originally the point of grenade was upgrades AFAIK we didn't clearly specify individual services | 22:38 |
| jogo | dhellmann: venvs may be an issue for distros | 22:38 |
| jeblair | mtreinish: problem is that an N+1 interval might be completely incomptabile. it's completely reasonable to say that juno openstack works with foo<1.0 and kilo openstack works with bar>=1.0. if the maintainer of foo releases something like that, we have to be able to handle it. | 22:39 |
| dhellmann | jogo: do we upgrade all services at once in grenade? | 22:39 |
| dhellmann | jogo: or do we update one, test, update next, test, etc. | 22:39 |
| mtreinish | jogo: well devstack is probably going to move towards per service venvs eventually, we can discuss doing that in grenade once that's done | 22:39 |
| jogo | dhellmann: all at once (but in partial n-cpu, we do all but n-cpu) | 22:39 |
| *** mfink has joined #openstack-infra | 22:39 | |
| dhellmann | jogo: ok. I guess we don't update the dependencies to simulate being able to update the apps separately | 22:40 |
| jogo | mtreinish: while I am a big fan of going to per service venvs I think there will be distro backlash (which I am ok wit) | 22:40 |
| mtreinish | jogo: why would they care? devstack/grenade don't use packaged python things (or at least they shouldn't) | 22:41 |
| *** cnesa2 has quit IRC | 22:41 | |
| dhellmann | jogo: we would still be testing that everything can be installed together, but this changes the upgrade test from "one service at a time" to "one container at a time" | 22:41 |
| jogo | dhellmann: as long as we test both I think distros will be happy | 22:42 |
| dhellmann | right, we'd need to keep both scenarios | 22:42 |
| jogo | dhellmann: so your proposing we migrate grenade to do service venvs? | 22:42 |
| dhellmann | whatever sort of container makes you happy, but venvs are easy | 22:42 |
| jogo | dhellmann: and leave the default config for devstack to not use them? | 22:42 |
| dhellmann | yeah, I think that makes sense | 22:42 |
| jogo | dhellmann: makes sense to me too | 22:42 |
| dhellmann | and also have grenade update dependencies | 22:43 |
| *** atiwari has quit IRC | 22:43 | |
| jogo | dhellmann: per $container | 22:43 |
| dhellmann | jogo: right | 22:43 |
| jogo | dhellmann: would making grenade update dependencies without service $containers be sufficient? | 22:44 |
| *** doug-fish has joined #openstack-infra | 22:44 | |
| dhellmann | jogo: it would fix my problem, but it would no longer tell us if we could update systems without "forklifting" the whole thing | 22:44 |
| dhellmann | so I don't think that's what we really want, long term | 22:44 |
| jeblair | yeah, that would break the ncpu test right? | 22:44 |
| jogo | jeblair: well I was thinking for ncpu, we could just restart n-cpu with new deps | 22:45 |
| mtreinish | it could if there were incompatible reqs between releases on n-cpu | 22:45 |
| jeblair | dhellmann: re 1.4: will juno work with oslo.db 1.4? i thought an incompatability had been introduced? or is that incompatibility just the sqla version you're talking about changing? | 22:45 |
| jogo | mtreinish: exactly | 22:45 |
| dhellmann | jogo: that won't work if the dependencies don't overlap, because you won't be able to start the other services then | 22:46 |
| *** ociuhandu has joined #openstack-infra | 22:46 | |
| dhellmann | jeblair: the incompatibility is the sqla version setting | 22:46 |
| jeblair | mtreinish, jogo: yeah, i think that's just another time bomb of the same variety, and we will need venvs (or equiv) to avoid that | 22:46 |
| jogo | dhellmann: and that is something we want t explicitly support right? | 22:46 |
| dhellmann | jogo: so I've been told | 22:46 |
| adam_g | per-service venvs would be useful in the regular devstack run for isolating tempest from the rest of the deployment | 22:46 |
| jogo | jeblair dhellmann: if you right a qa-spec about how you want the grenade job to do venvs (which I assume means add venv support to grenade) I would be happy to work on it | 22:47 |
| jeblair | adam_g: i think we already do that by running with tox | 22:47 |
| mtreinish | jeblair: well to an extent, site-packages is enabled in the tempest tox.ini | 22:47 |
| jeblair | jogo: i can't push this, sorry. | 22:48 |
| jeblair | i'm really hoping someone else can take ownership of it | 22:48 |
| mtreinish | jogo: it would have to be added to devstack first, then we could use it in grenade | 22:49 |
| jogo | jeblair: ack was hoping dhellmann could define the reqs and you review | 22:49 |
| openstackgerrit | Doug Hellmann proposed openstack/requirements: Make SQLAlchemy settings compatible with Juno https://review.openstack.org/141927 | 22:49 |
| jogo | mtreinish: right | 22:50 |
| dhellmann | jogo: I don't think I have time to do that, either. :-/ | 22:50 |
| *** sarob has joined #openstack-infra | 22:51 | |
| jogo | dhellmann: :/ mtreinish would you be able to help? | 22:51 |
| mmedvede | jeblair: the second part of elasticsearch split is now ready https://review.openstack.org/#/c/137155/ | 22:51 |
| mtreinish | jogo: maybe on the review side, but I've got too many plates spinning to write a spec for this. (I'm also not sure I understand all the reqs) | 22:52 |
| *** yfried is now known as yfried|afk | 22:52 | |
| mmedvede | jeblair: thanks | 22:52 |
| *** kumartin has joined #openstack-infra | 22:53 | |
| jeblair | jhesketh, clarkb, fungi, mordred, SergeyLukjanov: i single-core approved https://review.openstack.org/#/c/137155/ due to time constraints | 22:53 |
| jogo | mtreinish dhellmann: I don't understand all the reqs either I think | 22:54 |
| jeblair | jogo: you set up the original ncpu partial job, right? can you drive it? i think you'll have a lot of help reviewing | 22:54 |
| jogo | jeblair: I can drive it yeah, then ask you folks for reviews to make sure I am not missing anything | 22:54 |
| jeblair | jogo: thanks. yeah, none of us is going to think of everything. :) | 22:55 |
| jogo | jeblair: but first want to get some of my aiopcpu stuff landed | 22:55 |
| openstackgerrit | Wayne Warren proposed openstack-infra/jenkins-job-builder: Update 'timeout' wrapper module https://review.openstack.org/129467 | 22:56 |
| jogo | live migration is blocked on some big devstack-gate changes that hopefully dtroyer will do, but can get tempest-full without live migrate working | 22:56 |
| jeblair | jogo: *nod*. i think this is a time bomb, and if it goes off again before we fix grenade, we may have to disable some grenade jobs to dig ourselves out. | 22:57 |
| jogo | jeblair: agreed, which is why I am diving into this now | 22:57 |
| jeblair | woot | 22:57 |
| adam_g | jogo, i'd be down to help out with service venvs. i'm in jury duty today and tomorrow, though, with bad courthouse wifi | 22:58 |
| jeblair | jogo: i will pitch in by bumping the d-g changes to the top of my review queue | 22:58 |
| *** dims has quit IRC | 22:58 | |
| *** packet has quit IRC | 22:59 | |
| jogo | jeblair: thanks, rebasing my patch series to make it clear what is ready to land | 22:59 |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: DO NOT MERGE: testing aiopcpu with tempest full https://review.openstack.org/136504 | 22:59 |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: Set up ssh_known_host based on hostname https://review.openstack.org/136596 | 22:59 |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: Set custom cpu_model for live_migrate https://review.openstack.org/141530 | 22:59 |
| *** ChuckC has joined #openstack-infra | 22:59 | |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: Setup /etc/hosts for aiopcpu https://review.openstack.org/136158 | 22:59 |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: Allow root user to ssh in as stack https://review.openstack.org/137176 | 22:59 |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: Enable live block migration https://review.openstack.org/135768 | 22:59 |
| jogo | as no reason to land the live migrate patches yet | 22:59 |
| dhellmann | jogo: thanks, I'll help with reviews as well | 22:59 |
| dhellmann | in the mean time, I'm going ahead with the oslo.db 1.4 plan, unless someone can point out a reason that won't work | 23:00 |
| jeblair | dhellmann: i can not think of such a reason, which means nothing. :) | 23:00 |
| dougwig | hi infra, openstack/requirements question. I'm trying to add the neutron-*aas repos to the project list, but... they reference neutron as a library in master as "-e git+https://git.openstack.org/openstack/neutron#egg=neutron", which fails in the requirements jenkins jobs. How should this case be handled? | 23:01 |
| dhellmann | jeblair: yes, I can't either, which means less and may give me the false confidence to deal with it tomorrow :-) | 23:01 |
| fungi | sorry, got pulled away for a few minutes--catching up on conversation now | 23:03 |
| *** andreaf has quit IRC | 23:03 | |
| jeblair | dougwig: we have agreed that is undesirable and you need to reference either a specific version or a branch tarball. note that in devstack jobs, if you add it to the project list in devstack-gate, it will be installed from git. then the requirements should generally only be used in unit test jobs. | 23:03 |
| *** dims has joined #openstack-infra | 23:03 | |
| *** achanda has quit IRC | 23:04 | |
| jogo | jeblair: ok everything under https://review.openstack.org/#/c/136504/ should be ready (and if all goes well that patch will pass tempest-full for aiopcpu, but maybe not neutron) | 23:04 |
| *** achanda has joined #openstack-infra | 23:04 | |
| dougwig | jeblair: so, move it to test-requirements.txt and trust the packager's to get the dependencies right? (i'm fine with that.) then we add specific requirements entries on the stable branches, maybe? | 23:05 |
| *** rkukura_ has joined #openstack-infra | 23:05 | |
| *** sandywalsh_ has joined #openstack-infra | 23:05 | |
| * jogo starts writing the qa-spec | 23:05 | |
| jogo | fungi: looks like http://status.openstack.org/elastic-recheck/data/uncategorized.html is hung again :( | 23:06 |
| mtreinish | jogo: I didn't think neutron was enabled on aiopcpu (we really should rename that eventually) | 23:06 |
| *** charz has quit IRC | 23:06 | |
| *** esker has quit IRC | 23:06 | |
| jogo | mtreinish: it is in the job with neutron and aiopcpu in the name | 23:06 |
| mtreinish | jogo: oh, I didn't think there was one like that | 23:07 |
| *** esker has joined #openstack-infra | 23:07 | |
| mtreinish | but I see it's there now, I must have just missed it before | 23:07 |
| *** rkukura has quit IRC | 23:08 | |
| *** rkukura_ is now known as rkukura | 23:08 | |
| *** sandywalsh has quit IRC | 23:08 | |
| jogo | wow https://jogo.github.io/gate/ | 23:08 |
| *** esker has quit IRC | 23:09 | |
| *** achanda has quit IRC | 23:09 | |
| *** esker has joined #openstack-infra | 23:09 | |
| *** charz has joined #openstack-infra | 23:09 | |
| jeblair | jogo: i see a blank page | 23:12 |
| *** sputnik13 has quit IRC | 23:12 | |
| mtreinish | jogo: yeah, nothing is loading for me either | 23:12 |
| *** yfried|afk is now known as yfried | 23:12 | |
| jroll | it works without ssl http://jogo.github.io/gate/ | 23:13 |
| jroll | also, wow. | 23:13 |
| jogo | oh right the ssl | 23:14 |
| *** sputnik13 has joined #openstack-infra | 23:14 | |
| *** esker has quit IRC | 23:15 | |
| fungi | dougwig: any reason you need a python requirements entry to install neutron for you? that's not going to work out well outside of our ci anyway. we probably just need the relevant jobs to install neutron explicitly | 23:16 |
| mtreinish | jogo: heh, your rolling average is showing :) | 23:16 |
| jogo | mtreinish: what should the BP be for the non overlapping dependency upgrades | 23:16 |
| jogo | mtreinish: exactly, its sorta neat to see it in action | 23:16 |
| mtreinish | probably grenade although we don't have bps setup for grenade | 23:16 |
| fungi | dougwig: we shouldn't be putting things in requirements or test requirements lists which aren't released to pypi | 23:16 |
| jogo | mtreinish: well some of the change is to devstack | 23:16 |
| dougwig | fungi: just an attempt at completeness and naivete because the sub-repos import neutron. | 23:17 |
| *** Sukhdev has quit IRC | 23:17 | |
| fungi | dhellmann: i think the sqlalchemy requirement difference in oslo.db 1.3.0 was merely because the logistics around releasing a version for stable weren't forseen | 23:17 |
| fungi | dhellmann: so that plan sounds as sane as i could hope for at least | 23:17 |
| mtreinish | jogo: push it up as a devstack bp then, I think that's fine | 23:17 |
| jogo | mtreinish: cool | 23:18 |
| fungi | jogo: i don't see any lingering/hung processes nor cron e-mail to the recheck account | 23:19 |
| *** timcline_ has joined #openstack-infra | 23:19 | |
| jogo | fungi: hrmm | 23:20 |
| *** teran has joined #openstack-infra | 23:20 | |
| *** emagana has quit IRC | 23:20 | |
| fungi | jogo: do those log anywhere? the recheckwatch irc bot is logging some tracebacks but i don't see where the update queries would log other than to stdout/stderr | 23:20 |
| *** rlandy has quit IRC | 23:20 | |
| jogo | oh wait, its up to date, my bad | 23:20 |
| fungi | excellent | 23:21 |
| jogo | timestamp is not UTC its local time | 23:21 |
| fungi | local time of what/where? | 23:21 |
| jogo | Data Last Updated: Mon Dec 15 2014 15:00:01 GMT-0800 (PST) | 23:21 |
| jogo | on http://status.openstack.org/elastic-recheck/ | 23:21 |
| fungi | Data Last Updated: Mon Dec 15 2014 23:00:01 GMT+0000 (UTC) | 23:21 |
| fungi | lgtm, yeah | 23:21 |
| fungi | Last Elastic Search Index Update: Mon Dec 15 2014 22:59:32 GMT+0000 (UTC) | 23:21 |
| fungi | okay, i'm going to go grab some dinner then see if i can salvage what's left of the day | 23:22 |
| fungi | bbiaw | 23:22 |
| jogo | mtreinish: how does non-overlapping-dependency-upgrades sound? | 23:22 |
| jogo | for bp name | 23:22 |
| *** yfried is now known as yfried|afk | 23:23 | |
| *** teran_ has quit IRC | 23:23 | |
| *** timcline has quit IRC | 23:23 | |
| *** timcline_ has quit IRC | 23:24 | |
| *** unicell has quit IRC | 23:24 | |
| *** signed8bit has quit IRC | 23:24 | |
| dhellmann | fungi: that's what I thought; thank you for the confirmation | 23:25 |
| *** MarkAtwood has quit IRC | 23:25 | |
| mtreinish | jogo: I guess that's fine | 23:25 |
| *** dims has quit IRC | 23:25 | |
| *** kumartin has quit IRC | 23:26 | |
| *** amitgandhinz has quit IRC | 23:26 | |
| dhellmann | anteaya: thanks for the link to https://github.com/dougwig/split-repo/blob/master/service-split.sh (sorry for the delayed response) | 23:26 |
| dougwig | dhellmann: it's a delayed response kind of day. | 23:27 |
| *** armax has quit IRC | 23:27 | |
| *** andreykurilin_ has quit IRC | 23:27 | |
| dhellmann | dougwig: seriously, I'm still catching up :-) | 23:27 |
| anteaya | dhellmann: you ahve been busy | 23:27 |
| openstackgerrit | Merged openstack-infra/gear: Use non-blocking IO in server https://review.openstack.org/128754 | 23:27 |
| openstackgerrit | Merged openstack-infra/system-config: Split out elasticsearch module https://review.openstack.org/137155 | 23:28 |
| dhellmann | the script looks good; I have plans to refactor ours into a separate script for making a repo containing specific contents only and then the oslo-specific parts | 23:28 |
| *** annegent_ has joined #openstack-infra | 23:29 | |
| openstackgerrit | Merged openstack-infra/elastic-recheck: Expand failure pattern for bug 1402747 https://review.openstack.org/141912 | 23:29 |
| uvirtbot | Launchpad bug 1402747 in glance "setuptools 8 breaks multi-range version checks" [High,Confirmed] https://launchpad.net/bugs/1402747 | 23:29 |
| *** fandi has joined #openstack-infra | 23:31 | |
| openstackgerrit | Merged openstack-infra/zuul: Set gearman timeout to 300 https://review.openstack.org/141460 | 23:31 |
| anteaya | dhellmann: the instructions in the script about getting branches and tags into the seed repo were important points during dougwig's process | 23:31 |
| *** yamamoto has quit IRC | 23:33 | |
| *** teran has quit IRC | 23:33 | |
| *** emagana has joined #openstack-infra | 23:35 | |
| mordred | hey all - sorry for the absence - my day exploded | 23:37 |
| *** e0ne has quit IRC | 23:38 | |
| openstackgerrit | Anita Kuno proposed openstack-infra/infra-manual: Adds a link for new developers to use sandbox https://review.openstack.org/104951 | 23:38 |
| mordred | it's gonna take a while to digest scrollback, if there are any things that anyone needs/wants me to attention sooner rather than later - poke me | 23:38 |
| *** spzala has quit IRC | 23:38 | |
| *** fandi has quit IRC | 23:39 | |
| anteaya | exploding days, sounds familiar | 23:39 |
| sweston | anteaya: awesome, thank you :-). Do you know what the process is to acquire push rights to the repo? | 23:39 |
| anteaya | sweston: well it is a stackforge repo | 23:39 |
| dougwig | sweston: put your repo somewhere public and ask infra really, _really_, nicely. | 23:40 |
| clarkb | mordred see lifeless' comment on 141677 | 23:40 |
| anteaya | so clone it, create a patch use git review, it will end up on gerrit | 23:40 |
| clarkb | mordred then we should decide if we want to drop the shas from versions | 23:40 |
| sweston | dougwig: :-) | 23:40 |
| anteaya | sweston: you mean core reviewer rights on it? ask mikal to add you | 23:40 |
| jhesketh | anteaya: when you get a chance I've responded to your comment on https://review.openstack.org/#/c/141668/ | 23:41 |
| sweston | anteaya: correct, otherwise it won't be very useful ;-) | 23:41 |
| *** doug-fish has left #openstack-infra | 23:41 | |
| *** dmsimard is now known as dmsimard_away | 23:41 | |
| *** bswartz has joined #openstack-infra | 23:41 | |
| *** dizquierdo has quit IRC | 23:41 | |
| *** gondoi is now known as zz_gondoi | 23:42 | |
| sweston | anteaya: thank you | 23:42 |
| anteaya | sweston: welcome, go forth and create lovely things | 23:42 |
| anteaya | jhesketh: I'm all for either hyphens or underscores | 23:42 |
| *** BadCub02 has joined #openstack-infra | 23:43 | |
| sweston | ping mikal ^^ | 23:43 |
| * dstufft comes back and tries to figure out if people are more or less angry at him now | 23:43 | |
| anteaya | jhesketh: how many jobs would need to be edited if we go all hyphens in the swift jobs? | 23:43 |
| mordred | clarkb: okie | 23:43 |
| anteaya | jhesketh: I would even write the patch if you are okay with it | 23:43 |
| * mordred hands dstufft an apple pie | 23:43 | |
| anteaya | or review it in a positive way, your call | 23:43 |
| jhesketh | anteaya: I can neaten them up, it'd only be a small handful | 23:44 |
| sweston | anteaya: that is always my objective | 23:44 |
| * anteaya lines up for pie | 23:44 | |
| jhesketh | give me s a second | 23:44 |
| anteaya | sweston: awesome | 23:44 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/project-config: Add extra detail to swift indexes https://review.openstack.org/141286 | 23:44 |
| anteaya | jhesketh: thank you | 23:44 |
| dstufft | mordred: :D | 23:44 |
| *** dimtruck is now known as zz_dimtruck | 23:44 | |
| *** jaypipes has quit IRC | 23:44 | |
| anteaya | jhesketh: I will grep for <ha> in your scripts henceforth | 23:45 |
| mordred | clarkb, dstufft: how about we break 141667 into two patches - one that populates the git hash json file and provides the tool to inspect it | 23:45 |
| mordred | that's generally useful, as dstufft points out | 23:45 |
| dstufft | lifeless said he was +1 on that extra functionality too | 23:45 |
| mordred | then the question of "should we continue to have git sha directly in version number" can be considered | 23:46 |
| dstufft | he was just -1 on removing from the version number | 23:46 |
| clarkb | sure. I think the contwntious thing is removinf it from the versions | 23:46 |
| clarkb | which honestly I am no longer expecting to avoid | 23:46 |
| *** zz_zz_sabari is now known as zz_zz_zz_sabari | 23:46 | |
| dstufft | I'm poking at if setuptools/PEP 440 can provide an answer to that | 23:46 |
| *** mpaolino has joined #openstack-infra | 23:47 | |
| *** cdent has quit IRC | 23:47 | |
| jogo | dhellmann: trying to understand what went wrong | 23:47 |
| jogo | dhellmann: and why we now want to support non overlapping dependencies (for the spec) | 23:47 |
| jhesketh | anteaya: lol :P | 23:50 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/project-config: Push dg-tempest-dsvm-full logs to swift https://review.openstack.org/141668 | 23:50 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/project-config: Rename zuul-swift-upload jobs to use hypens https://review.openstack.org/141942 | 23:50 |
| *** annegent_ has quit IRC | 23:50 | |
| jhesketh | anteaya: ^ there you go | 23:50 |
| *** emagana has quit IRC | 23:50 | |
| *** mpaolino has quit IRC | 23:51 | |
| *** aysyd has quit IRC | 23:51 | |
| openstackgerrit | Dean Troyer proposed openstack-infra/devstack-gate: Switch to use local.conf for DevStack runs https://review.openstack.org/141943 | 23:52 |
| mordred | clarkb: I think I'm aligned with you on that | 23:52 |
| jogo | jeblair: ^ what is the problem statement for why we need non overlapping dependency upgrades? | 23:53 |
| clarkb | I must return to not being here today. I am +2ish to the change as it exists byt was unable to test properly so didnt actually +2 | 23:53 |
| clarkb | jogo doing upgrades one service at a time | 23:54 |
| jogo | clarkb: we can do that today if we assume overlapping dependency sets | 23:54 |
| anteaya | jhesketh: thank you, I shall eat toast and then review | 23:54 |
| jhesketh | sounds good, I'm going to go cafehack :-) | 23:55 |
| clarkb | jogo yes so I thibk saying non overlapping means we dont do that | 23:55 |
| clarkb | jogo maybe I dont understand your question | 23:55 |
| *** dizquierdo has joined #openstack-infra | 23:55 | |
| notmyname | remember when I was asking about getting shortlinks for gerrit dashboards? I made an upstream feature request and have been told "you just need to be made a project owner". since I don't know much about the details of our gerrit deploy, I don't really know how to repond https://code.google.com/p/gerrit/issues/detail?id=3057 | 23:55 |
| jogo | clarkb: the question is why do we want to allow juno to have require foo<1.0 and kilo require foo>1.1 | 23:56 |
| *** annegent_ has joined #openstack-infra | 23:56 | |
| *** asettle has quit IRC | 23:57 | |
| jeblair | jogo: juno require foo <1.0 and kilo require foo >= 1.0 | 23:57 |
| *** alex7376 has joined #openstack-infra | 23:57 | |
| jeblair | to reduce it to the simple case... | 23:57 |
| *** annegent_ has quit IRC | 23:57 | |
| *** tnovacik has quit IRC | 23:57 | |
| *** alex7376 is now known as asettle | 23:57 | |
| clarkb | we want to do that? | 23:58 |
| jeblair | jogo: for two reasons: 1) because, particularly if we are not the authors of foo, there is no guarantee than foo 1.0 is in any way compatible with 0.9. we should be able to handle that. | 23:59 |
| *** avozza is now known as zz_avozza | 23:59 | |
| jeblair | (in our current situation, it would be easier to switch from foo to bar in that case than to upgrade foo; that does not make sense to me) | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!