| *** chlong has joined #openstack-infra | 00:04 | |
| Clint | HOST_NAME_MAX | 00:04 |
|---|---|---|
| *** harlowja_at_home has joined #openstack-infra | 00:04 | |
| *** mikeit has quit IRC | 00:08 | |
| *** _diana_ has quit IRC | 00:09 | |
| *** heyongli has quit IRC | 00:13 | |
| *** laughterwym has joined #openstack-infra | 00:14 | |
| *** heyongli has joined #openstack-infra | 00:15 | |
| *** notmyname has quit IRC | 00:27 | |
| *** morgabra has quit IRC | 00:28 | |
| *** notmyname has joined #openstack-infra | 00:28 | |
| *** jpeeler has quit IRC | 00:29 | |
| *** morgabra has joined #openstack-infra | 00:29 | |
| *** Kennan2 is now known as Kennan | 00:33 | |
| *** amotoki has joined #openstack-infra | 00:39 | |
| *** otter768 has joined #openstack-infra | 00:41 | |
| *** jpeeler has joined #openstack-infra | 00:43 | |
| *** amotoki has quit IRC | 00:43 | |
| *** sdake has quit IRC | 00:53 | |
| *** puranamr has quit IRC | 00:53 | |
| *** mikeit has joined #openstack-infra | 00:54 | |
| *** jasondotstar has quit IRC | 00:54 | |
| *** madong has quit IRC | 00:54 | |
| *** dims__ has joined #openstack-infra | 00:55 | |
| *** larainema has joined #openstack-infra | 00:56 | |
| *** dims_ has quit IRC | 00:56 | |
| *** mikeit has quit IRC | 00:56 | |
| *** esmute__ has quit IRC | 01:01 | |
| *** zz_dimtruck is now known as dimtruck | 01:03 | |
| *** dims__ has quit IRC | 01:06 | |
| *** dims_ has joined #openstack-infra | 01:07 | |
| openstackgerrit | lifeless proposed openstack/requirements: Juno specific fixups: https://review.openstack.org/198147 | 01:08 |
| openstackgerrit | lifeless proposed openstack/requirements: Update Juno requirements code https://review.openstack.org/198146 | 01:08 |
| openstackgerrit | lifeless proposed openstack/requirements: Update Kilo requirements code https://review.openstack.org/198145 | 01:09 |
| *** baoli has joined #openstack-infra | 01:10 | |
| *** gmann has joined #openstack-infra | 01:12 | |
| *** dims_ has quit IRC | 01:15 | |
| *** baoli has quit IRC | 01:15 | |
| *** _diana_ has joined #openstack-infra | 01:15 | |
| *** otter768 has quit IRC | 01:16 | |
| *** btully has joined #openstack-infra | 01:17 | |
| *** heyongli has quit IRC | 01:18 | |
| *** heyongli has joined #openstack-infra | 01:19 | |
| *** heyongli has quit IRC | 01:23 | |
| *** gmann has quit IRC | 01:26 | |
| *** esmute__ has joined #openstack-infra | 01:27 | |
| *** Alexandra_ has quit IRC | 01:29 | |
| *** gmann has joined #openstack-infra | 01:35 | |
| *** jasondotstar has joined #openstack-infra | 01:38 | |
| *** amotoki has joined #openstack-infra | 01:40 | |
| *** _diana_ has quit IRC | 01:41 | |
| *** amotoki has quit IRC | 01:44 | |
| *** mayurig has joined #openstack-infra | 01:45 | |
| *** _diana_ has joined #openstack-infra | 01:45 | |
| *** Longgeek_ has joined #openstack-infra | 01:46 | |
| *** esmute__ has quit IRC | 01:47 | |
| *** heyongli has joined #openstack-infra | 01:47 | |
| *** esmute__ has joined #openstack-infra | 01:47 | |
| *** Longgeek has quit IRC | 01:49 | |
| *** _diana_ has quit IRC | 01:49 | |
| *** baoli has joined #openstack-infra | 01:50 | |
| *** _diana_ has joined #openstack-infra | 01:50 | |
| *** jamielennox is now known as jamielennox|away | 01:51 | |
| *** jasondotstar has quit IRC | 01:55 | |
| *** mayurig has quit IRC | 01:55 | |
| *** _diana_ has quit IRC | 01:56 | |
| *** esmute__ has quit IRC | 01:57 | |
| *** baoli has quit IRC | 01:58 | |
| *** jamielennox|away is now known as jamielennox | 02:02 | |
| lifeless | jhesketh: can I has +A ? https://review.openstack.org/#/c/198186/ | 02:03 |
| openstackgerrit | James Polley proposed openstack-dev/pbr: WIP: more extensive testing of requirement parsing https://review.openstack.org/187823 | 02:04 |
| *** _diana_ has joined #openstack-infra | 02:04 | |
| *** pal has joined #openstack-infra | 02:06 | |
| jhesketh | lifeless: looking | 02:08 |
| *** _diana_ has quit IRC | 02:08 | |
| *** sdake has joined #openstack-infra | 02:11 | |
| *** ayoung has joined #openstack-infra | 02:11 | |
| *** sdake_ has joined #openstack-infra | 02:14 | |
| *** yamamoto has joined #openstack-infra | 02:15 | |
| *** ayoung has quit IRC | 02:16 | |
| *** sdake has quit IRC | 02:18 | |
| *** baoli has joined #openstack-infra | 02:20 | |
| *** marzif has quit IRC | 02:22 | |
| *** achanda has joined #openstack-infra | 02:24 | |
| openstackgerrit | James Polley proposed openstack-dev/pbr: Fix everything by requiring setuptools>=16. https://review.openstack.org/187846 | 02:26 |
| *** ayoung has joined #openstack-infra | 02:28 | |
| *** asettle has joined #openstack-infra | 02:32 | |
| *** achanda has quit IRC | 02:32 | |
| jhesketh | lifeless: reviewed | 02:36 |
| *** coolsvap|away is now known as coolsvap | 02:36 | |
| *** pal has quit IRC | 02:38 | |
| lifeless | replied | 02:39 |
| lifeless | jhesketh: ^ | 02:39 |
| *** amotoki has joined #openstack-infra | 02:41 | |
| jhesketh | lifeless: thanks, and replied | 02:44 |
| *** amotoki has quit IRC | 02:45 | |
| *** Longgeek_ has quit IRC | 02:47 | |
| *** Longgeek has joined #openstack-infra | 02:47 | |
| *** pcrews has quit IRC | 02:50 | |
| *** coolsvap is now known as coolsvap|away | 02:50 | |
| *** dimtruck is now known as zz_dimtruck | 02:51 | |
| *** armax has joined #openstack-infra | 02:53 | |
| lifeless | jhesketh: and replied :) | 02:53 |
| *** pal has joined #openstack-infra | 02:53 | |
| openstackgerrit | lifeless proposed openstack/requirements: Update Kilo requirements code https://review.openstack.org/198145 | 02:54 |
| openstackgerrit | lifeless proposed openstack/requirements: Replace coinor.pulp with pulp. https://review.openstack.org/198588 | 02:54 |
| *** puranamr has joined #openstack-infra | 02:54 | |
| *** heyongli has quit IRC | 02:56 | |
| *** baoli has quit IRC | 02:56 | |
| *** heyongli has joined #openstack-infra | 02:57 | |
| *** esmute__ has joined #openstack-infra | 02:57 | |
| *** puranamr has quit IRC | 02:59 | |
| *** pal has quit IRC | 03:00 | |
| openstackgerrit | lifeless proposed openstack/requirements: Don't permit embedded #'s in URLs. https://review.openstack.org/198590 | 03:00 |
| lifeless | jhesketh: and ^ the fix | 03:00 |
| *** AzherKhan has joined #openstack-infra | 03:01 | |
| *** linkedinyou has quit IRC | 03:01 | |
| *** BharatK has joined #openstack-infra | 03:01 | |
| *** esmute__ has quit IRC | 03:02 | |
| *** armax has quit IRC | 03:03 | |
| *** heyongli has quit IRC | 03:03 | |
| *** heyongli has joined #openstack-infra | 03:04 | |
| jhesketh | lifeless: what happens if the constraints changes the URL but keeps the same egg name? | 03:04 |
| jhesketh | I'm guessing it'll update the URL | 03:04 |
| jhesketh | but perhaps we should add a test for that? | 03:04 |
| lifeless | blink | 03:04 |
| lifeless | I can't parse what you just said | 03:04 |
| lifeless | its either what its already tested to do | 03:05 |
| lifeless | or out of context ;) | 03:05 |
| *** zz_dimtruck is now known as dimtruck | 03:06 | |
| jhesketh | lifeless: existing req is 'file://path/to/foo#egg=foo', new constraint is 'http://some/path/foobar#egg=foo' | 03:08 |
| lifeless | all lines for foo will be replaced with that line | 03:08 |
| lifeless | foo | 03:08 |
| lifeless | foo==1.2 | 03:08 |
| *** koolhead17 has joined #openstack-infra | 03:08 | |
| lifeless | foo==1.2;python_version=="2.7" | 03:08 |
| lifeless | tahts what this is tested to do | 03:08 |
| lifeless | jhesketh: the data model keys on the name; there may be multiple lines (due to markers) for a single name | 03:09 |
| lifeless | jhesketh: the change here is to parse inwards on urls correctly, existing tests + the datamodel ensure that the output is correct | 03:10 |
| jhesketh | lifeless: okay cool | 03:10 |
| jhesketh | lifeless: well lgtm, thanks for the extra explanations :-). Are you around for a bit if I +A? | 03:11 |
| lifeless | I have spotted another minor quirk to fix, but again - I'd like to get this landed | 03:11 |
| lifeless | jhesketh: yes | 03:11 |
| lifeless | 1.25 horus | 03:11 |
| lifeless | then again after C is in bed | 03:11 |
| lifeless | but its low risk - its green :) | 03:11 |
| lifeless | unlike a release which could previously go everywhere | 03:12 |
| jhesketh | lifeless: yeah not too concerned, just thought I'd check | 03:13 |
| jhesketh | thanks for your follow up patch too, that lgtm | 03:14 |
| *** amotoki has joined #openstack-infra | 03:15 | |
| *** Longgeek_ has joined #openstack-infra | 03:16 | |
| *** otter768 has joined #openstack-infra | 03:17 | |
| *** Longgeek has quit IRC | 03:18 | |
| *** otter768 has quit IRC | 03:21 | |
| *** achanda has joined #openstack-infra | 03:26 | |
| *** puranamr has joined #openstack-infra | 03:27 | |
| *** nihilifer has joined #openstack-infra | 03:28 | |
| *** yamahata has quit IRC | 03:28 | |
| *** heyongli has quit IRC | 03:29 | |
| *** yamada-h has joined #openstack-infra | 03:30 | |
| *** heyongli has joined #openstack-infra | 03:32 | |
| *** esmute__ has joined #openstack-infra | 03:33 | |
| *** oomichi has quit IRC | 03:34 | |
| *** nihilifer has quit IRC | 03:34 | |
| *** yamada-h has quit IRC | 03:34 | |
| *** achanda has quit IRC | 03:35 | |
| *** puranamr has quit IRC | 03:35 | |
| *** achanda has joined #openstack-infra | 03:35 | |
| *** harlowja_at_home has quit IRC | 03:38 | |
| *** jamielennox is now known as jamielennox|away | 03:38 | |
| mordred | greghaynes: awesome | 03:39 |
| *** achanda has quit IRC | 03:40 | |
| *** yamahata has joined #openstack-infra | 03:40 | |
| *** btully has quit IRC | 03:45 | |
| openstackgerrit | Merged openstack-infra/glean: Set hostname to the bottom level domain https://review.openstack.org/198576 | 03:48 |
| *** puranamr has joined #openstack-infra | 03:48 | |
| *** jamielennox|away is now known as jamielennox | 03:49 | |
| *** pal has joined #openstack-infra | 03:50 | |
| *** amotoki has quit IRC | 03:52 | |
| *** coolsvap|away is now known as coolsvap | 03:53 | |
| *** sdake_ is now known as sdake | 03:56 | |
| *** baoli has joined #openstack-infra | 03:57 | |
| *** ddieterly has quit IRC | 03:57 | |
| *** esmute__ has quit IRC | 03:57 | |
| greghaynes | mordred: we should cut a glean release so it goes into tomorrows images | 04:01 |
| *** baoli has quit IRC | 04:01 | |
| *** amotoki has joined #openstack-infra | 04:06 | |
| *** ildikov has quit IRC | 04:06 | |
| openstackgerrit | Merged openstack/requirements: Fix editing constraints files with urls in them. https://review.openstack.org/198186 | 04:09 |
| *** dims_ has joined #openstack-infra | 04:12 | |
| *** amotoki has quit IRC | 04:14 | |
| *** dims__ has joined #openstack-infra | 04:16 | |
| lifeless | jogo: https://bugs.launchpad.net/testrepository/+bug/902881 | 04:17 |
| openstack | Launchpad bug 902881 in Testrepository "no way to drop into a debugger safely" [Wishlist,Triaged] | 04:17 |
| lifeless | jogo: put some thoughts in there | 04:17 |
| lifeless | jogo: ping me if you want more | 04:17 |
| *** dims_ has quit IRC | 04:18 | |
| *** ricky1 has left #openstack-infra | 04:18 | |
| *** puranamr has quit IRC | 04:18 | |
| *** amotoki has joined #openstack-infra | 04:19 | |
| *** heyongli has quit IRC | 04:20 | |
| *** dims__ has quit IRC | 04:20 | |
| *** dims__ has joined #openstack-infra | 04:21 | |
| jogo | lifeless: thanks will read it a bit later | 04:21 |
| *** heyongli has joined #openstack-infra | 04:22 | |
| *** dims___ has joined #openstack-infra | 04:22 | |
| *** dims__ has quit IRC | 04:26 | |
| *** dims___ has quit IRC | 04:27 | |
| *** Sukhdev has joined #openstack-infra | 04:28 | |
| *** heyongli has quit IRC | 04:29 | |
| *** heyongli has joined #openstack-infra | 04:30 | |
| *** heyongli has quit IRC | 04:34 | |
| *** coolsvap is now known as coolsvap|brb | 04:35 | |
| *** BharatK has quit IRC | 04:37 | |
| *** heyongli has joined #openstack-infra | 04:38 | |
| *** amotoki has quit IRC | 04:40 | |
| *** dimtruck is now known as zz_dimtruck | 04:40 | |
| *** heyongli has quit IRC | 04:42 | |
| *** zz_dimtruck is now known as dimtruck | 04:45 | |
| *** heyongli has joined #openstack-infra | 04:48 | |
| *** amotoki has joined #openstack-infra | 04:51 | |
| *** heyongli has quit IRC | 04:52 | |
| *** pcrews has joined #openstack-infra | 04:53 | |
| *** yamada-h has joined #openstack-infra | 04:54 | |
| *** asettle has quit IRC | 04:55 | |
| *** heyongli has joined #openstack-infra | 04:56 | |
| *** dimtruck is now known as zz_dimtruck | 04:58 | |
| *** ddieterly has joined #openstack-infra | 04:58 | |
| *** esmute__ has joined #openstack-infra | 05:00 | |
| *** BharatK has joined #openstack-infra | 05:03 | |
| *** ddieterly has quit IRC | 05:03 | |
| *** esmute__ has quit IRC | 05:04 | |
| *** deepakcs has joined #openstack-infra | 05:05 | |
| *** heyongli has quit IRC | 05:06 | |
| *** coolsvap|brb is now known as coolsvap | 05:06 | |
| *** heyongli has joined #openstack-infra | 05:07 | |
| greghaynes | mordred: looks like I dont have tag pushing perms :( so youll have to do it | 05:09 |
| clarkb | we can likely fix that | 05:11 |
| greghaynes | clarkb: yea, in the meanwhile - you likely have perms to push a tag? | 05:14 |
| *** masayukig_ has joined #openstack-infra | 05:16 | |
| *** otter768 has joined #openstack-infra | 05:18 | |
| *** heyongli has quit IRC | 05:19 | |
| *** heyongli has joined #openstack-infra | 05:19 | |
| *** k4n0 has joined #openstack-infra | 05:19 | |
| clarkb | maybe? but no keys at the moment | 05:21 |
| greghaynes | womp womp | 05:22 |
| *** otter768 has quit IRC | 05:22 | |
| *** Sukhdev has quit IRC | 05:24 | |
| *** abregman has joined #openstack-infra | 05:26 | |
| *** _nadya_ has joined #openstack-infra | 05:28 | |
| *** btully has joined #openstack-infra | 05:29 | |
| *** coolsvap is now known as coolsvap|away | 05:29 | |
| *** s0nea has joined #openstack-infra | 05:36 | |
| *** Sukhdev has joined #openstack-infra | 05:43 | |
| *** _nadya_ has quit IRC | 05:45 | |
| *** yamamoto has quit IRC | 05:51 | |
| *** ig0r__ has joined #openstack-infra | 05:51 | |
| *** yfried_ has joined #openstack-infra | 05:52 | |
| *** heyongli has quit IRC | 05:54 | |
| *** ig0r_ has quit IRC | 05:55 | |
| *** heyongli has joined #openstack-infra | 05:55 | |
| *** shashank_hegde has joined #openstack-infra | 05:55 | |
| *** yfried__ has joined #openstack-infra | 05:55 | |
| *** yfried_ has quit IRC | 05:56 | |
| *** kdas_ is now known as kushal | 06:02 | |
| *** kushal has joined #openstack-infra | 06:03 | |
| *** heyongli has quit IRC | 06:05 | |
| *** heyongli has joined #openstack-infra | 06:06 | |
| *** mrunge has joined #openstack-infra | 06:07 | |
| *** binbincong has joined #openstack-infra | 06:07 | |
| *** matbu has joined #openstack-infra | 06:09 | |
| *** matbu has left #openstack-infra | 06:10 | |
| *** dims_ has joined #openstack-infra | 06:10 | |
| *** yamamoto has joined #openstack-infra | 06:13 | |
| *** _nadya_ has joined #openstack-infra | 06:14 | |
| *** dims_ has quit IRC | 06:15 | |
| *** baoli has joined #openstack-infra | 06:22 | |
| *** flepied has joined #openstack-infra | 06:24 | |
| *** ddieterly has joined #openstack-infra | 06:28 | |
| openstackgerrit | Marc Koderer proposed openstack-infra/project-config: Activate devstack plugin for Murano https://review.openstack.org/198325 | 06:30 |
| *** ddieterly has quit IRC | 06:32 | |
| *** scheuran has joined #openstack-infra | 06:35 | |
| *** yfried__ is now known as yfried|afk | 06:37 | |
| *** yfried|afk is now known as yfried__ | 06:38 | |
| openstackgerrit | YAMAMOTO Takashi proposed openstack-infra/project-config: Add python3 jobs for neutron-lbaas https://review.openstack.org/198611 | 06:40 |
| openstackgerrit | YAMAMOTO Takashi proposed openstack-infra/project-config: Make python34 whitespace consistent with others https://review.openstack.org/198612 | 06:42 |
| *** ociuhandu has joined #openstack-infra | 06:42 | |
| *** Sukhdev has quit IRC | 06:48 | |
| *** HeOS has quit IRC | 06:50 | |
| *** koolhead_ has joined #openstack-infra | 06:54 | |
| *** koolhead17 has quit IRC | 06:56 | |
| *** shashank_hegde has quit IRC | 06:56 | |
| *** Sukhdev has joined #openstack-infra | 06:57 | |
| *** yfried__ is now known as yfried|afk | 06:59 | |
| *** yfried|afk is now known as yfried__ | 07:00 | |
| openstackgerrit | Merged openstack-infra/puppet-accessbot: Update template to use instance variables https://review.openstack.org/198160 | 07:01 |
| openstackgerrit | Sachi King proposed openstack/requirements: WIP: Add capability to run tests in environment installed with constraints https://review.openstack.org/198620 | 07:07 |
| Nakato | lifeless: ^ | 07:09 |
| *** mrunge has quit IRC | 07:09 | |
| *** camunoz is now known as camunoz_gone | 07:10 | |
| lifeless | Nakato: interesting | 07:11 |
| lifeless | Nakato: I'm inclined to not bother with the shim | 07:11 |
| lifeless | Nakato: in that we already have other hard minimum caps on the pip within the venv | 07:11 |
| *** dims_ has joined #openstack-infra | 07:12 | |
| lifeless | Nakato: so constraints: is special tox syntax? | 07:12 |
| *** mrunge has joined #openstack-infra | 07:13 | |
| Nakato | lifeless: Yep, https://tox.readthedocs.org/en/latest/config.html#complex-factor-conditions | 07:13 |
| *** achanda has joined #openstack-infra | 07:13 | |
| Nakato | lifeless: Which venv are you meaning, is there one before tox on the gate systems, or the one installed by tox? | 07:15 |
| *** Sukhdev has quit IRC | 07:15 | |
| *** dims_ has quit IRC | 07:16 | |
| Nakato | My concern is if the environment tox is running from has virtualenv<13.1.0 pip won't be at version 1.7.0 and thus no -c. | 07:16 |
| *** ociuhandu has quit IRC | 07:16 | |
| lifeless | yes, I get that that is the concern | 07:17 |
| lifeless | right now if its < 13.0.3 then the pip in it will break anyway on env markers | 07:17 |
| lifeless | so this is no different | 07:17 |
| *** mpaolino has joined #openstack-infra | 07:18 | |
| lifeless | Nakato: full feedback in the review | 07:18 |
| openstackgerrit | Chris Hoge proposed openstack-infra/system-config: Add eavesdrop logging to #refstack channel https://review.openstack.org/198621 | 07:18 |
| Nakato | lifeless: Cool, thanks | 07:18 |
| *** otter768 has joined #openstack-infra | 07:18 | |
| *** ParsectiX has joined #openstack-infra | 07:19 | |
| *** otter768 has quit IRC | 07:23 | |
| *** _nadya_ has quit IRC | 07:24 | |
| *** achanda has quit IRC | 07:25 | |
| *** mikedillion has joined #openstack-infra | 07:26 | |
| *** mpavone has joined #openstack-infra | 07:28 | |
| *** mikedillion has left #openstack-infra | 07:28 | |
| *** ddieterly has joined #openstack-infra | 07:29 | |
| *** markus_z has joined #openstack-infra | 07:29 | |
| *** _nadya_ has joined #openstack-infra | 07:32 | |
| *** sshnaidm has joined #openstack-infra | 07:32 | |
| *** ddieterly has quit IRC | 07:33 | |
| *** devvesa has joined #openstack-infra | 07:35 | |
| *** chlong has quit IRC | 07:35 | |
| *** samuelBartel has joined #openstack-infra | 07:36 | |
| *** achanda has joined #openstack-infra | 07:36 | |
| *** Ala has joined #openstack-infra | 07:36 | |
| *** kaisers has joined #openstack-infra | 07:37 | |
| *** viktors|afk is now known as viktors | 07:37 | |
| *** kaisers has quit IRC | 07:37 | |
| *** ildikov has joined #openstack-infra | 07:42 | |
| *** yfried__ is now known as yfried|afk | 07:43 | |
| *** Krinkle_ is now known as Krinkle | 07:45 | |
| *** _nadya_ has quit IRC | 07:46 | |
| clarkb | fwiw I don't think its a major issue to say "use latest pip/virtualenv" | 07:46 |
| clarkb | the hard part is going to be communication but as a general rule its a good one | 07:46 |
| *** notnownikki has joined #openstack-infra | 07:48 | |
| *** ParsectiX has quit IRC | 07:50 | |
| *** ParsectiX has joined #openstack-infra | 07:51 | |
| openstackgerrit | Jerome Hourquebie proposed stackforge/python-jenkins: Adds Cloudbees folder plugin support https://review.openstack.org/180185 | 07:57 |
| *** laughterwym_ has joined #openstack-infra | 08:00 | |
| *** achanda has quit IRC | 08:02 | |
| *** hichihara has quit IRC | 08:02 | |
| *** achanda has joined #openstack-infra | 08:02 | |
| *** kaisers has joined #openstack-infra | 08:03 | |
| *** yfried|afk is now known as yfried__ | 08:03 | |
| *** laughterwym has quit IRC | 08:04 | |
| *** baoli has quit IRC | 08:04 | |
| *** kaisers has quit IRC | 08:04 | |
| *** vponomaryov has joined #openstack-infra | 08:08 | |
| *** yamahata has quit IRC | 08:09 | |
| *** achanda has quit IRC | 08:09 | |
| *** ParsectiX has quit IRC | 08:10 | |
| *** achanda has joined #openstack-infra | 08:10 | |
| *** HeOS has joined #openstack-infra | 08:10 | |
| *** AzherKhan has quit IRC | 08:10 | |
| *** AzherKhan has joined #openstack-infra | 08:11 | |
| openstackgerrit | Jan Klare proposed openstack-infra/project-config: add periodic chef-rake-integration job for openstack-chef-repo https://review.openstack.org/194164 | 08:15 |
| *** pal has quit IRC | 08:16 | |
| *** _nadya_ has joined #openstack-infra | 08:18 | |
| *** bauwser is now known as bauzas | 08:20 | |
| *** hashar has joined #openstack-infra | 08:20 | |
| *** btully has quit IRC | 08:23 | |
| *** soren has joined #openstack-infra | 08:27 | |
| *** _nadya_ has quit IRC | 08:28 | |
| *** achanda has quit IRC | 08:29 | |
| *** ddieterly has joined #openstack-infra | 08:30 | |
| *** rossella_s has joined #openstack-infra | 08:32 | |
| *** Ala has quit IRC | 08:33 | |
| *** masayukig_ has quit IRC | 08:33 | |
| *** masayukig_ has joined #openstack-infra | 08:33 | |
| *** ddieterly has quit IRC | 08:34 | |
| *** sshnaidm has quit IRC | 08:38 | |
| *** madong_ has joined #openstack-infra | 08:39 | |
| *** larainema has quit IRC | 08:39 | |
| *** Ala has joined #openstack-infra | 08:42 | |
| *** Murad has joined #openstack-infra | 08:42 | |
| Murad | does anybody have any idea how to install devstack with infiniband? | 08:43 |
| *** _nadya_ has joined #openstack-infra | 08:46 | |
| *** laughterwym_ has quit IRC | 08:48 | |
| openstackgerrit | Derek Higgins proposed openstack-infra/tripleo-ci: Remove unneeded nova patch https://review.openstack.org/198356 | 08:50 |
| *** pal has joined #openstack-infra | 08:54 | |
| *** sshnaidm has joined #openstack-infra | 08:57 | |
| *** dizquierdo has joined #openstack-infra | 08:58 | |
| *** ParsectiX has joined #openstack-infra | 09:04 | |
| *** baoli has joined #openstack-infra | 09:04 | |
| *** pelix has joined #openstack-infra | 09:05 | |
| *** binbincong has quit IRC | 09:08 | |
| *** baoli has quit IRC | 09:09 | |
| *** jamielennox is now known as jamielennox|away | 09:11 | |
| openstackgerrit | Merged openstack-infra/project-config: Add separate acls for 'stable/6.1' branch of fuel projects https://review.openstack.org/196130 | 09:17 |
| *** jamielennox|away is now known as jamielennox | 09:18 | |
| *** otter768 has joined #openstack-infra | 09:19 | |
| *** otter768 has quit IRC | 09:24 | |
| *** yfried__ is now known as yfried|afk | 09:25 | |
| *** yfried|afk is now known as yfried__ | 09:26 | |
| *** ddieterly has joined #openstack-infra | 09:31 | |
| *** ddieterly has quit IRC | 09:35 | |
| openstackgerrit | Antoine Musso proposed openstack-infra/nodepool: Fix Jenkins nodes methods to use POST https://review.openstack.org/193600 | 09:38 |
| *** dpose has joined #openstack-infra | 09:39 | |
| *** yfried__ is now known as yfried|afk | 09:45 | |
| *** ociuhandu has joined #openstack-infra | 09:46 | |
| *** yfried|afk is now known as yfried__ | 09:46 | |
| *** heyongli has quit IRC | 09:50 | |
| *** heyongli has joined #openstack-infra | 09:51 | |
| *** koolhead_ has quit IRC | 09:52 | |
| *** yamada-h has quit IRC | 09:52 | |
| openstackgerrit | fumihiko kakuma proposed openstack-infra/project-config: Make gate-networking-ofagent-python34 non-voting and add py34 gate https://review.openstack.org/198657 | 09:59 |
| *** ildikov has quit IRC | 09:59 | |
| *** sshnaidm has quit IRC | 09:59 | |
| ruhe | lifeless: hi! when do you plan to update upper-constraints.txt? | 10:00 |
| ruhe | for yaql there's incosistency between global-requirements.txt and upper-constraints.txt which causes murano dsvm jobs to fail | 10:01 |
| clarkb | ruhe there is an automated job to propose updates daily iirc. Does that mean your min version in global reqs isnt high enough? | 10:01 |
| ruhe | clarkb: global-requirements.txt has "yaql>=0.2.6,!=0.3.0", but upper-constraints.txt has "yaql===0.3" | 10:02 |
| clarkb | interesting, I think that means0.3 was getting selected somehow when generating the conatraints | 10:03 |
| ruhe | the part with "!=0.3.0" was merged to global-requirements.txt after upper-constraints.txt was introduced | 10:03 |
| clarkb | oh in that case you can propose the change yourself | 10:04 |
| ruhe | clarkb: will do. thank you! | 10:04 |
| *** marekd has joined #openstack-infra | 10:04 | |
| clarkb | keep in mind it may have transitive affects so you need to run the script to do it aiui | 10:04 |
| clarkb | and lifeless was going to document that in the requirements repo | 10:05 |
| marekd | Hello, To be honest I don't fully understand what fungi meant here - maybe somebody available now will be able to sched some light on that? re: https://review.openstack.org/#/c/190631/3/zuul/layout.yaml | 10:05 |
| *** Kennan has quit IRC | 10:05 | |
| *** yfried__ is now known as yfried|afk | 10:05 | |
| clarkb | marekd I dont see comments from fungi there, do you mean the cpmments from jeblair? | 10:06 |
| *** laughterwym has joined #openstack-infra | 10:06 | |
| *** park__heijlong has quit IRC | 10:07 | |
| lifeless | it'll self correct when fungi's fixes to the script job land | 10:07 |
| *** Ala has quit IRC | 10:08 | |
| marekd | clarkb: oh, sorry, my bad! i meant jeblair's comment! | 10:08 |
| clarkb | marekd jeblair is saying that there is no corresponding code removal in that diff for the old job | 10:10 |
| clarkb | marekd so a new patchset is likely needed to add the removal | 10:10 |
| *** kaisers has joined #openstack-infra | 10:11 | |
| *** kaisers has quit IRC | 10:11 | |
| *** marzif has joined #openstack-infra | 10:12 | |
| marekd | clarkb: ok i might lack some knowledge here so i'll likely need some help from you guys - what shall i remove then? When I grep for saml2 in that file I found 3 occurences and all seem relevant. | 10:12 |
| *** sshnaidm has joined #openstack-infra | 10:12 | |
| *** kaisers has joined #openstack-infra | 10:12 | |
| clarkb | marked basically when you rename you should have matching deletes and adds | 10:12 |
| clarkb | marked remove whatever corresponds to those lines that are added | 10:12 |
| *** yfried|afk is now known as yfried__ | 10:14 | |
| clarkb | there should be a line for the old name with a branch restrixtion and that can be removed | 10:14 |
| openstackgerrit | Antoine Musso proposed stackforge/python-jenkins: Creates ability to work with views in jenkins https://review.openstack.org/194307 | 10:14 |
| *** ildikov has joined #openstack-infra | 10:14 | |
| marekd | clarkb: let me fetch the change and proper grep | 10:14 |
| marekd | clarkb: if it comes to zuul/layout.yaml file there are only two occurences of saml2 keyword. | 10:16 |
| marekd | clarkb: ./layout.yaml:1481: - name: gate-keystoneauth-saml2-requirements and ./layout.yaml:1976: - name: openstack/keystoneauth-saml2 | 10:16 |
| marekd | since i was renaming from python-keystoneclient-saml2 to keystoneauth-saml2 there would be some keystoneclient related if I had some dangling to be removed entries there, right? | 10:17 |
| *** john-davidge has joined #openstack-infra | 10:17 | |
| clarkb | ya | 10:17 |
| *** binbincong has joined #openstack-infra | 10:17 | |
| clarkb | which means that this line was likely added rather than part of the rename? perhaps that should go in a different change? also why not run that job against master? | 10:18 |
| marekd | clarkb: i will be honest with you i am following steps from some other rename-like patches :( | 10:19 |
| marekd | done by jamielennox | 10:19 |
| *** park__heijlong has joined #openstack-infra | 10:20 | |
| *** yamamoto has quit IRC | 10:20 | |
| marekd | what I need for now is to rename python-keystoneclient-saml2 into keystoneauth-saml2 and be able to depend on a keystoneauth library which is not yet fully released, I am guessing this also need relaxing of the constraints of the requirements jobs. | 10:20 |
| *** john-davidge has left #openstack-infra | 10:21 | |
| *** pc_m has joined #openstack-infra | 10:21 | |
| *** Kennan has joined #openstack-infra | 10:21 | |
| *** ddieterly has joined #openstack-infra | 10:31 | |
| *** marzif_ has joined #openstack-infra | 10:33 | |
| *** mpaolino has quit IRC | 10:35 | |
| *** ddieterly has quit IRC | 10:36 | |
| *** baoli has joined #openstack-infra | 10:36 | |
| *** park_heijlong has joined #openstack-infra | 10:37 | |
| *** mpaolino has joined #openstack-infra | 10:37 | |
| *** yfried__ has quit IRC | 10:38 | |
| *** yfried__ has joined #openstack-infra | 10:38 | |
| *** Kennan2 has joined #openstack-infra | 10:39 | |
| openstackgerrit | Evgeny Antyshev proposed openstack-infra/zuul: Make zuul-merger construct git URL, not zuul-server https://review.openstack.org/198667 | 10:41 |
| *** Kennan has quit IRC | 10:41 | |
| *** park__heijlong has quit IRC | 10:41 | |
| *** baoli has quit IRC | 10:41 | |
| *** Longgeek_ has quit IRC | 10:41 | |
| *** Longgeek has joined #openstack-infra | 10:42 | |
| *** hashar has quit IRC | 10:42 | |
| *** cdent has joined #openstack-infra | 10:44 | |
| openstackgerrit | Darragh Bailey proposed openstack-infra/jenkins-job-builder: Reformat xUnit publisher function and docstring help https://review.openstack.org/183585 | 10:47 |
| *** dims_ has joined #openstack-infra | 10:47 | |
| *** Kennan2 is now known as Kennan | 10:48 | |
| *** yfried__ has quit IRC | 10:49 | |
| *** AzherKhan has quit IRC | 10:52 | |
| *** dims_ has quit IRC | 10:52 | |
| igorbelikov | fungi SergeyLukjanov AJaeger_ hi, folks! could someone please add me to fuel-stable-core group in gerrit(https://review.openstack.org/#/admin/groups/984,members)? It was created as a result of my review(https://review.openstack.org/#/c/196130/) and I need to be able to manage it. | 10:55 |
| sdague | lifeless: if you happen to still be up and can help me figure out what I did wrong with - https://review.openstack.org/#/c/196203 (testing for the wsgi scripts pbr patch) it would be appreciated | 10:57 |
| *** ociuhandu has quit IRC | 10:58 | |
| openstackgerrit | Darragh Bailey proposed openstack-infra/jenkins-job-builder: Replace open() with io.open() and force 'utf-8' https://review.openstack.org/183488 | 11:01 |
| *** jcoufal has joined #openstack-infra | 11:01 | |
| *** chlong has joined #openstack-infra | 11:05 | |
| *** yfried__ has joined #openstack-infra | 11:13 | |
| mkoderer | hi folks, the "Depends-On" flag seems not to work for infra-config changes | 11:15 |
| mkoderer | I would like to test my change on murano ci before I merge it in infra-config | 11:16 |
| mkoderer | https://review.openstack.org/#/c/194077/ | 11:16 |
| mkoderer | but it seems the project config isn't changed at all (localrc looks completly identical) | 11:17 |
| freerunner | mkoderer: Hi! The 'Depends-On' tag ton yet implemented in Murano CI logic. We have a patch on review for this. https://review.openstack.org/#/c/197565/ | 11:17 |
| *** e0ne has joined #openstack-infra | 11:17 | |
| *** derekh has joined #openstack-infra | 11:18 | |
| *** yfried has joined #openstack-infra | 11:18 | |
| *** yfried__ has quit IRC | 11:18 | |
| *** yfried is now known as yfried|afk | 11:18 | |
| mkoderer | freerunner: ahh ok :) thanks.. I will wait until this gets merged then | 11:18 |
| mkoderer | freerunner: but that's only for murano-ci? is it the same for murano in the gate? | 11:19 |
| freerunner | mkoderer: That's only for Murano CI. In dsvm-* gates 'Depends-On' should work... | 11:20 |
| *** otter768 has joined #openstack-infra | 11:20 | |
| openstackgerrit | Darragh Bailey proposed openstack-infra/jenkins-job-builder: Wrap file objects with codecs to handle unicode https://review.openstack.org/183939 | 11:20 |
| mkoderer | freerunner: I mean the flag works but the project config change seems to get ignored though | 11:21 |
| mkoderer | freerunner: so the dependency is visible in zuul.openstack.org but the project config is still unchanged | 11:22 |
| freerunner | mkoderer: It seems.. | 11:24 |
| *** otter768 has quit IRC | 11:25 | |
| *** heyongli has quit IRC | 11:26 | |
| *** heyongli has joined #openstack-infra | 11:26 | |
| *** sshnaidm has quit IRC | 11:27 | |
| *** sshnaidm has joined #openstack-infra | 11:28 | |
| openstackgerrit | Evgeny Antyshev proposed openstack-infra/zuul: Make zuul-merger construct git URL, not zuul-server https://review.openstack.org/198667 | 11:30 |
| *** ildikov has quit IRC | 11:31 | |
| *** yfried|afk is now known as yfried | 11:31 | |
| *** hashar has joined #openstack-infra | 11:31 | |
| *** ddieterly has joined #openstack-infra | 11:32 | |
| *** yamamoto has joined #openstack-infra | 11:34 | |
| *** yamamoto_ has joined #openstack-infra | 11:34 | |
| *** baoli has joined #openstack-infra | 11:37 | |
| *** ddieterly has quit IRC | 11:37 | |
| *** ociuhandu has joined #openstack-infra | 11:37 | |
| *** amrith is now known as _amrith_ | 11:38 | |
| *** yamamoto has quit IRC | 11:38 | |
| *** baoli_ has joined #openstack-infra | 11:40 | |
| mordred | clarkb: if you're still around, https://review.openstack.org/#/c/198121/ ... I'd love to get that in before I cut a new glean release. | 11:41 |
| mordred | fungi: ^^ | 11:41 |
| mordred | morning sdague | 11:42 |
| *** baoli has quit IRC | 11:42 | |
| *** thorst has joined #openstack-infra | 11:42 | |
| *** ldnunes has joined #openstack-infra | 11:43 | |
| *** ildikov has joined #openstack-infra | 11:44 | |
| *** baoli_ has quit IRC | 11:45 | |
| *** e0ne has quit IRC | 11:45 | |
| *** baoli has joined #openstack-infra | 11:51 | |
| ruhe | turns out, to run generate-constraints you need all the system dependencies of all the python libraries listed in global-requirements.txt to be installed on the system | 11:53 |
| *** pal has quit IRC | 11:54 | |
| openstackgerrit | fumihiko kakuma proposed openstack-infra/project-config: Make gate-networking-ofagent-python34 non-voting and add py34 check https://review.openstack.org/198657 | 11:55 |
| *** yfried is now known as yfried|afk | 11:55 | |
| *** baoli_ has joined #openstack-infra | 11:55 | |
| *** _amrith_ is now known as amrith | 11:56 | |
| *** e0ne has joined #openstack-infra | 11:57 | |
| *** baoli has quit IRC | 11:58 | |
| sdague | morning mordred | 12:00 |
| SergeyLukjanov | igorbelikov, done | 12:01 |
| *** bradjones has joined #openstack-infra | 12:01 | |
| *** bradjones has quit IRC | 12:01 | |
| *** bradjones has joined #openstack-infra | 12:01 | |
| igorbelikov | SergeyLukjanov thanks! | 12:01 |
| Murad | does anybody have any idea how to install devstack with infiniband? | 12:02 |
| *** radez_g0n3 is now known as radez | 12:02 | |
| *** binbincong has quit IRC | 12:03 | |
| *** dims_ has joined #openstack-infra | 12:04 | |
| openstackgerrit | Adam Coldrick proposed openstack-infra/storyboard: Support offset/limit pagination in the API https://review.openstack.org/198297 | 12:05 |
| *** yfried|afk is now known as yfried | 12:05 | |
| openstackgerrit | garyk proposed openstack/requirements: Bump oslo.vmware to 0.15.0 https://review.openstack.org/198689 | 12:07 |
| *** zz_dimtruck is now known as dimtruck | 12:08 | |
| *** dims_ has quit IRC | 12:08 | |
| *** amrith is now known as _amrith_ | 12:12 | |
| *** leakypipes is now known as jaypipes | 12:12 | |
| *** binbincong has joined #openstack-infra | 12:15 | |
| *** nzeer has quit IRC | 12:15 | |
| *** heyongli has quit IRC | 12:15 | |
| *** nzeer has joined #openstack-infra | 12:15 | |
| *** heyongli has joined #openstack-infra | 12:16 | |
| *** ildikov has quit IRC | 12:22 | |
| openstackgerrit | Ruslan Kamaldinov proposed openstack/requirements: Adjust yaql version in upper-constraints https://review.openstack.org/198696 | 12:23 |
| *** weshay has joined #openstack-infra | 12:23 | |
| *** bdemers has joined #openstack-infra | 12:24 | |
| *** bswartz has quit IRC | 12:24 | |
| *** jcoufal has quit IRC | 12:24 | |
| *** yfried has quit IRC | 12:25 | |
| *** gordc has joined #openstack-infra | 12:27 | |
| *** jcoufal has joined #openstack-infra | 12:27 | |
| *** heyongli has quit IRC | 12:29 | |
| *** aysyd has joined #openstack-infra | 12:29 | |
| *** heyongli has joined #openstack-infra | 12:29 | |
| *** BharatK has quit IRC | 12:31 | |
| *** spzala has joined #openstack-infra | 12:33 | |
| *** ddieterly has joined #openstack-infra | 12:33 | |
| devvesa | Hi! Anyone can tell me which permissions do I have to set on gerrit acls file of my project to delete branches? | 12:35 |
| openstackgerrit | Adam Coldrick proposed openstack-infra/storyboard: Support offset/limit pagination in the API https://review.openstack.org/198297 | 12:35 |
| *** ddieterly has quit IRC | 12:38 | |
| *** jasondotstar has joined #openstack-infra | 12:38 | |
| *** rfolco has joined #openstack-infra | 12:38 | |
| *** ildikov has joined #openstack-infra | 12:39 | |
| frickler | devvesa: are you talking about your own gerrit or some repo on review.openstack.org? | 12:43 |
| frickler | devvesa: for the latter, there was a discussion recently with the result that only infra ppl are allowed to delete branches | 12:43 |
| *** jraim has quit IRC | 12:43 | |
| *** jraim has joined #openstack-infra | 12:43 | |
| frickler | because there are too many other permissions coupled to deleting branches | 12:43 |
| *** mjturek1 has quit IRC | 12:44 | |
| devvesa | A repo on review.openstack.org | 12:44 |
| devvesa | fricker: ^ | 12:44 |
| mordred | yah. we don't give out branch delete permissions | 12:45 |
| mordred | because acl related reasons | 12:45 |
| devvesa | Ok.. I understand | 12:45 |
| *** yamamoto_ has quit IRC | 12:45 | |
| devvesa | So, if I want to delete some branches, I have to ask a infrastructure owner with willingness to do it | 12:46 |
| *** yamamoto has joined #openstack-infra | 12:46 | |
| mordred | branches are a pretty heavy-weight thing in gerrit and are usually not needed - so we recommend against using them except for things that really do want to be long-term like stable/* branches | 12:46 |
| mordred | that is correct | 12:46 |
| devvesa | Well, we created a couple more for a misunderstood between us (the project developers) and we would like to rollback the creation | 12:47 |
| mordred | devvesa: which branches? (and which project?) | 12:47 |
| devvesa | https://github.com/openstack/networking-midonet | 12:47 |
| *** mjturek1 has joined #openstack-infra | 12:48 | |
| devvesa | branches `release/kilo_legacy` and `stable/kilo_legacy` | 12:48 |
| mordred | devvesa: done | 12:48 |
| devvesa | also `gerrit/refs/tags/*` which I created somehow when I tried to push tags | 12:48 |
| devvesa | (sorry for the inconveniences) | 12:48 |
| mordred | I was just going to ask about those :) | 12:48 |
| devvesa | :) Thanks | 12:48 |
| mordred | and no worried! | 12:48 |
| mordred | and no worries! | 12:49 |
| *** jcoufal has quit IRC | 12:49 | |
| mordred | oh! | 12:49 |
| mordred | I cannot delete those | 12:49 |
| mordred | fungi: ^^ when you get up - this is almost interesting enough to be interesting | 12:49 |
| devvesa | I should work as gerrit QA, then :) | 12:50 |
| devvesa | Anyway, thanks for trying and for the quick responses | 12:50 |
| *** ZZelle has quit IRC | 12:53 | |
| *** dims_ has joined #openstack-infra | 12:55 | |
| *** sshnaidm has quit IRC | 12:56 | |
| *** sshnaidm has joined #openstack-infra | 12:56 | |
| *** rlandy has joined #openstack-infra | 12:57 | |
| *** dims__ has joined #openstack-infra | 12:57 | |
| *** e0ne is now known as e0ne_ | 12:57 | |
| *** Ala has joined #openstack-infra | 12:58 | |
| *** abregman has quit IRC | 12:59 | |
| *** dims___ has joined #openstack-infra | 13:00 | |
| *** dims_ has quit IRC | 13:00 | |
| *** dims_ has joined #openstack-infra | 13:01 | |
| *** doude has quit IRC | 13:01 | |
| *** doude has joined #openstack-infra | 13:02 | |
| *** doude has quit IRC | 13:02 | |
| *** doude has joined #openstack-infra | 13:02 | |
| lennyb | Hi, are we experiencing issues with git ? | 13:03 |
| *** julim has joined #openstack-infra | 13:03 | |
| *** dims__ has quit IRC | 13:03 | |
| *** dims___ has quit IRC | 13:05 | |
| *** erlon has joined #openstack-infra | 13:06 | |
| *** dims__ has joined #openstack-infra | 13:06 | |
| *** bswartz has joined #openstack-infra | 13:07 | |
| *** ameade_vacation is now known as ameade | 13:08 | |
| *** nelsnelson has joined #openstack-infra | 13:08 | |
| *** mmedvede_ has joined #openstack-infra | 13:09 | |
| *** yamamoto has quit IRC | 13:09 | |
| *** xyang1 has joined #openstack-infra | 13:09 | |
| *** dims_ has quit IRC | 13:09 | |
| anteaya | lennyb: that is a question that just begs another to clarify | 13:10 |
| anteaya | lennyb: why don't you start off by sharing your experience | 13:10 |
| anteaya | lennyb: if you want to first see if the question has been asked previously | 13:10 |
| * lennyb preparing copy paste | 13:10 | |
| anteaya | lennyb: you can always vist http://eavesdrop.openstack.org/irclogs/%23openstack-infra/ and take a look | 13:11 |
| *** dims__ has quit IRC | 13:11 | |
| *** ari has joined #openstack-infra | 13:11 | |
| *** linkedinyou has joined #openstack-infra | 13:12 | |
| *** yamamoto has joined #openstack-infra | 13:12 | |
| *** diego__ has joined #openstack-infra | 13:14 | |
| lennyb | http://144.76.193.39/ci-artifacts/194133/6/Nova-ML2-Sriov/console.html.gz | 13:14 |
| lennyb | 15:56:57 + git clone git://git.openstack.org/openstack-dev/devstack | 13:14 |
| lennyb | 15:56:57 Cloning into 'devstack'... | 13:14 |
| lennyb | 15:58:07 fatal: read error: Connection reset by peer | 13:14 |
| *** mugsie_ is now known as mugsie | 13:15 | |
| *** maishsk_afk has joined #openstack-infra | 13:15 | |
| *** maishsk_afk is now known as maishsk | 13:15 | |
| anteaya | I had hoped you were headed for paste.openstack.org rather than pasting into channel | 13:16 |
| fungi | lennyb: it happens from time to time. this is why we don't clone over the network in our ci, we start with a warm cache of all the clones, clone from a local path on disk, and then update that from remote git servers | 13:16 |
| *** kgiusti has joined #openstack-infra | 13:16 | |
| openstackgerrit | Dawid Malinowski proposed openstack-infra/jenkins-job-builder: Adding publisher support to join-trigger plugin https://review.openstack.org/178836 | 13:17 |
| *** yamamoto has quit IRC | 13:17 | |
| lennyb | anteaya: sorry, i think it's more clear this way, but I will paste into paste.openstack.org next time. | 13:18 |
| lennyb | fungi: thanks. | 13:19 |
| *** rhefner has left #openstack-infra | 13:20 | |
| * anteaya fails to understand how text pasted into paste.openstack.org is less clear than text pasted into channel | 13:20 | |
| fungi | mordred: i was able to delete those from git cli, but not from the webui | 13:20 |
| *** peristeri has joined #openstack-infra | 13:21 | |
| *** otter768 has joined #openstack-infra | 13:21 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:24 | |
| *** e0ne_ is now known as e0ne | 13:24 | |
| *** otter768 has quit IRC | 13:26 | |
| *** pal has joined #openstack-infra | 13:26 | |
| *** amotoki has quit IRC | 13:27 | |
| *** rbrooker has joined #openstack-infra | 13:29 | |
| *** _amrith_ is now known as amrith | 13:29 | |
| *** fabrice_ has joined #openstack-infra | 13:31 | |
| *** claudiub has joined #openstack-infra | 13:34 | |
| *** btully has joined #openstack-infra | 13:34 | |
| *** dustins has joined #openstack-infra | 13:34 | |
| *** simonmcc has quit IRC | 13:35 | |
| *** jsavak has joined #openstack-infra | 13:35 | |
| *** simonmcc has joined #openstack-infra | 13:35 | |
| *** claudiub has quit IRC | 13:38 | |
| *** chlong has quit IRC | 13:39 | |
| *** mriedem has joined #openstack-infra | 13:39 | |
| *** yfried has joined #openstack-infra | 13:40 | |
| *** eharney has joined #openstack-infra | 13:40 | |
| *** chlong has joined #openstack-infra | 13:40 | |
| *** Longgeek has quit IRC | 13:43 | |
| openstackgerrit | Om Kumar proposed openstack/diskimage-builder: Removes hardcoded refrences for ethernet interface https://review.openstack.org/198720 | 13:43 |
| openstackgerrit | Anita Kuno proposed openstack-infra/irc-meetings: Add id and url for third party meetings. https://review.openstack.org/198722 | 13:47 |
| *** ctracey has quit IRC | 13:47 | |
| *** ctracey has joined #openstack-infra | 13:48 | |
| *** yfried has quit IRC | 13:48 | |
| *** yfried has joined #openstack-infra | 13:48 | |
| *** miqui has joined #openstack-infra | 13:49 | |
| *** sshnaidm has quit IRC | 13:49 | |
| *** sshnaidm has joined #openstack-infra | 13:50 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/project-config: (WIP) Test osloganalyze in devstack https://review.openstack.org/198728 | 13:51 |
| openstackgerrit | Merged openstack-infra/tripleo-ci: Update DELOREAN_URL to one from 2015-06-19. https://review.openstack.org/193874 | 13:51 |
| *** marun has joined #openstack-infra | 13:51 | |
| *** dboik has joined #openstack-infra | 13:52 | |
| *** marun has left #openstack-infra | 13:53 | |
| *** prad has joined #openstack-infra | 13:53 | |
| *** r-daneel has joined #openstack-infra | 13:56 | |
| *** yfried has quit IRC | 13:58 | |
| *** linkedinyou has quit IRC | 13:59 | |
| *** yfried has joined #openstack-infra | 13:59 | |
| *** ParsectiX has quit IRC | 13:59 | |
| *** amitgandhinz has joined #openstack-infra | 13:59 | |
| *** k4n0 has quit IRC | 14:00 | |
| pabelanger | morning | 14:02 |
| *** amotoki has joined #openstack-infra | 14:04 | |
| *** apmelton has joined #openstack-infra | 14:04 | |
| *** sshnaidm has quit IRC | 14:05 | |
| *** sshnaidm has joined #openstack-infra | 14:05 | |
| *** koolhead17 has joined #openstack-infra | 14:06 | |
| *** r-daneel has quit IRC | 14:07 | |
| *** amotoki has quit IRC | 14:08 | |
| *** deepakcs has quit IRC | 14:08 | |
| anteaya | morning pabelanger | 14:09 |
| *** ddieterly has joined #openstack-infra | 14:10 | |
| mordred | fungi: btw - https://review.openstack.org/#/c/198121/ if you have a sec - I'd like to cut a new glean before image builds | 14:10 |
| *** yfried has quit IRC | 14:10 | |
| mordred | AJaeger_: how do I generate new po files in a doc repo? | 14:11 |
| mordred | (openstack-manuals) | 14:11 |
| mordred | or maybe I don't ... | 14:11 |
| fungi | mordred: sure thing. also do you happen to know why automodule isn't working in https://review.openstack.org/#/c/198504/4/doc/source/usage.rst | 14:12 |
| mordred | fungi: nope - that looks right to me | 14:14 |
| fungi | for some reason it generates no output in that part of the page | 14:15 |
| *** wenlock has joined #openstack-infra | 14:15 | |
| mordred | fungi: oh - you know what? | 14:17 |
| mordred | fungi: bindep/__init__ doesn't import bindep/depends.py in any way - you might try doing automodule:: bindep.depends? or something. I dunno | 14:18 |
| mordred | I'm always less than thrilled with what sphinx thinks 'auto' means, tbh | 14:19 |
| fungi | mordred: aha, that's the missing bit of recursion magic i was looking for--thanks | 14:19 |
| *** yfried has joined #openstack-infra | 14:19 | |
| *** r-daneel has joined #openstack-infra | 14:20 | |
| *** woodster_ has joined #openstack-infra | 14:21 | |
| mordred | that worked? *sigh8 | 14:22 |
| *** smoser has joined #openstack-infra | 14:23 | |
| *** soren has quit IRC | 14:23 | |
| openstackgerrit | Om Kumar proposed openstack/diskimage-builder: Removes hardcoded refrences for ethernet interface https://review.openstack.org/198720 | 14:23 |
| smoser | hey, can someone take a look at https://review.openstack.org/#/c/191081/ for me ? it has the +2 from jeblair_ but need a workflow. i believe that sdague's -1 comment is addressed. | 14:23 |
| openstackgerrit | Merged openstack-infra/glean: Wait longer for interface carrier to come up https://review.openstack.org/198121 | 14:23 |
| *** tonytan4ever has joined #openstack-infra | 14:24 | |
| openstackgerrit | vikram.choudhary proposed openstack-infra/project-config: Adding networking-onos as a part of neutron project. https://review.openstack.org/198740 | 14:25 |
| openstackgerrit | yolanda.robla proposed openstack-infra/system-config: Start using jfryman/puppet-selinux https://review.openstack.org/188357 | 14:28 |
| openstackgerrit | Dawid Malinowski proposed stackforge/python-jenkins: Adds Cloudbees folder plugin support https://review.openstack.org/180185 | 14:28 |
| *** sshnaidm has quit IRC | 14:28 | |
| mordred | smoser: I agree - it does look like sdague's concern is addressed | 14:28 |
| *** sshnaidm has joined #openstack-infra | 14:29 | |
| fungi | mordred: don't know if your suggestion worked yet. in a vmt meeting at the moment | 14:29 |
| sdague | smoser: yep | 14:29 |
| mordred | fungi: bah. meetings. | 14:29 |
| *** jeblair_ is now known as jeblair | 14:30 | |
| fungi | mordred: also i started in on trying to convert bindep from mox to mock so we can get py3k testing going for it. and discovered just how much i actually don't understand about writing tests after all | 14:32 |
| *** marzif_ has quit IRC | 14:32 | |
| *** marzif_ has joined #openstack-infra | 14:33 | |
| jeblair | fungi, mordred: good morning. anything i ought to know about? | 14:33 |
| fungi | jeblair: nothing i'm aware of. did you see fbo-'s blog post about zuul last week? good stuff | 14:34 |
| jeblair | oh cool | 14:35 |
| *** dustins_ has joined #openstack-infra | 14:35 | |
| fungi | jeblair: if you missed the url, http://techs.enovance.com/7542/dive-into-zuul-gated-commit-system-2 | 14:35 |
| openstackgerrit | yolanda.robla proposed openstack-infra/system-config: Move manage_projects logic to puppet-jeepyb https://review.openstack.org/171565 | 14:35 |
| *** hashar has quit IRC | 14:36 | |
| openstackgerrit | Merged openstack-infra/project-config: Add coverage jobs for cloud-init for both supported Pythons. https://review.openstack.org/191081 | 14:36 |
| *** rossella_s has quit IRC | 14:37 | |
| *** sshnaidm has quit IRC | 14:37 | |
| *** rossella_s has joined #openstack-infra | 14:37 | |
| *** sshnaidm has joined #openstack-infra | 14:38 | |
| *** dustins has quit IRC | 14:38 | |
| *** cindyo has joined #openstack-infra | 14:38 | |
| *** mrunge has quit IRC | 14:38 | |
| yolanda | mordred, fungi, i'm no longer seeing the zuul results in review.openstack.org in a table at the right, but down into the change. Is that a known thing? | 14:40 |
| *** sabeen has quit IRC | 14:41 | |
| *** Longgeek has joined #openstack-infra | 14:43 | |
| jeblair | yolanda: looks right to me; link to change where you are seeing the problem? | 14:44 |
| *** otter768 has joined #openstack-infra | 14:45 | |
| jeblair | yolanda: maybe shift-reload? | 14:45 |
| yolanda | yep, some cache thing | 14:45 |
| eantyshev | yolanda: Hello! let's discuss https://review.openstack.org/#/c/198667 ? | 14:47 |
| yolanda | hi, eantyshev, my point there is that there is an spec for zuul v3 nearly in. And this spec covers that situation, or shall be covering that | 14:48 |
| yolanda | jeblair, what are your thoughts there? ^ | 14:48 |
| *** amrith is now known as _amrith_ | 14:49 | |
| *** Longgeek has quit IRC | 14:49 | |
| *** _amrith_ is now known as amrith | 14:49 | |
| *** tbarron_afk is now known as tbarron | 14:49 | |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Allow neutron-milestone to tag networking-foo projects https://review.openstack.org/198749 | 14:50 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add stable ACLs for networking-foo projects https://review.openstack.org/198750 | 14:50 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add a create ACL for networking-bgpvpn https://review.openstack.org/198751 | 14:50 |
| openstackgerrit | Victor Stinner proposed openstack-infra/project-config: Move gate-keystone-python34 to check https://review.openstack.org/198350 | 14:52 |
| *** pal has quit IRC | 14:53 | |
| eantyshev | yolanda: Does my change require any modifications to that spec? | 14:54 |
| *** sshnaidm has quit IRC | 14:54 | |
| *** ddieterly has quit IRC | 14:54 | |
| yolanda | in my opinion, your change should be enclosed into the effort of that spec, so i'd say that you check that spec to find where it fits, and reference that into the commit message | 14:54 |
| *** sshnaidm has joined #openstack-infra | 14:54 | |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add stable ACLs for networking-foo projects https://review.openstack.org/198750 | 14:55 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add a create ACL for networking-bgpvpn https://review.openstack.org/198751 | 14:55 |
| pabelanger | yolanda, not sure if ianw replied to you about https://bugzilla.redhat.com/show_bug.cgi?id=1237395 or not | 14:55 |
| openstack | bugzilla.redhat.com bug 1237395 in git "extremley slow https clone performance" [Unspecified,Assigned] - Assigned to pstodulk | 14:55 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Allow neutron-milestone to tag networking-foo projects https://review.openstack.org/198749 | 14:55 |
| *** pal has joined #openstack-infra | 14:55 | |
| yolanda | pabelanger, i wasn't aware of that change, it's really good | 14:56 |
| yolanda | i was able to make that tetsts work locally with older git, but yes, the performance was super slow | 14:57 |
| *** ddieterly has joined #openstack-infra | 14:57 | |
| pabelanger | yolanda, Ya, I think ianw found the issue and is working to get patchset merged into RHEL6 | 14:57 |
| eantyshev | yolanda: forgive if I'm slow, what are 'tenants' of this spec? Gerrit accounts? | 14:57 |
| *** ddieterly has quit IRC | 14:58 | |
| yolanda | so pointing to ianw repo should be temporary? | 14:58 |
| pabelanger | yolanda, yes | 14:58 |
| *** hdd has joined #openstack-infra | 14:58 | |
| yolanda | cool | 14:58 |
| *** dguitarbite has quit IRC | 14:58 | |
| pabelanger | WIP patchset until bug is triaged properly I would assume | 14:59 |
| *** sshnaidm has quit IRC | 15:00 | |
| yolanda | eantyshev, i'm re-reading latest zuul v3 changes, just a sec | 15:00 |
| *** sshnaidm has joined #openstack-infra | 15:00 | |
| *** ddieterly has joined #openstack-infra | 15:02 | |
| *** fallenpegasus has joined #openstack-infra | 15:02 | |
| *** mtanino has joined #openstack-infra | 15:02 | |
| *** ajmiller__ has joined #openstack-infra | 15:02 | |
| *** ajmiller has quit IRC | 15:03 | |
| *** ajmiller__ is now known as ajmiller | 15:03 | |
| *** jsavak has quit IRC | 15:03 | |
| *** jsavak has joined #openstack-infra | 15:04 | |
| *** mmedvede_ has quit IRC | 15:04 | |
| *** maishsk has quit IRC | 15:06 | |
| fungi | mordred: no luck. switching that line to ".. automodule:: bindep.main" still doesn't produce any content in that part of the page | 15:07 |
| *** mmedvede has left #openstack-infra | 15:07 | |
| fungi | and bindep.main does in turn import bindep.depends | 15:07 |
| yolanda | eantyshev, tenants are "users" of zuul. In new version there can be different users using same zuul, but with project isolation, so one user cannot see config and projects from the other. I think your change could be fit in the "tenant isolation" section | 15:07 |
| mordred | fungi: then I do not know | 15:07 |
| yolanda | and if you are interested on that, you should add yourself as assignee on the spec | 15:07 |
| mordred | fungi: maybe dhellmann does ... he knows things | 15:07 |
| fungi | mordred: yeah, he's got some scary sphinx-fu | 15:08 |
| *** dims_ has joined #openstack-infra | 15:09 | |
| fungi | dhellmann: not urgent at all, but trying to figure out why automodule isn't adding any content for https://review.openstack.org/#/c/198504/4/doc/source/usage.rst | 15:09 |
| *** viktors is now known as viktors|afk | 15:09 | |
| fungi | dhellmann: i also tried making that bindep.main and bindep.depends instead of just bindep, but to no avail | 15:10 |
| fungi | wait, i take that back | 15:10 |
| fungi | i tried bindep.depends and it works | 15:10 |
| fungi | bindep.main does not | 15:10 |
| *** puranamr has joined #openstack-infra | 15:11 | |
| fungi | aha! it only covers class definitions, not global function definitions. now i understand | 15:11 |
| fungi | mordred: ^ | 15:11 |
| mordred | fungi: ahhhh | 15:11 |
| *** puranamr has quit IRC | 15:12 | |
| mordred | fungi: which is why it works for shade, since shade had a class that's in its __init__.py | 15:12 |
| fungi | looks like autofunction is the secret sauce for that | 15:12 |
| *** dannywilson has joined #openstack-infra | 15:12 | |
| *** amotoki has joined #openstack-infra | 15:12 | |
| *** puranamr has joined #openstack-infra | 15:13 | |
| *** markvoelker has quit IRC | 15:14 | |
| openstackgerrit | Adam Coldrick proposed openstack-infra/storyboard-webclient: Add paging controls to the result set pager https://review.openstack.org/198768 | 15:14 |
| openstackgerrit | Adam Coldrick proposed openstack-infra/storyboard-webclient: Implement offset/limit pagination of lists https://review.openstack.org/198769 | 15:14 |
| eantyshev | yolanda: so should I just mention that in commit message and change the topic? | 15:17 |
| *** puranamr_ has joined #openstack-infra | 15:17 | |
| *** mmedvede has joined #openstack-infra | 15:17 | |
| *** dangers_away is now known as dangers | 15:18 | |
| mordred | fungi: do you happen to remember the name of the program that behaves like pip but installs things into a package-specific virtualenv and makes symlinks into your bin dir for you? | 15:18 |
| *** puranamr has quit IRC | 15:18 | |
| mordred | dstufft: ^^ ? | 15:18 |
| fungi | mordred: mkvirtualenv/virtualenvwrapper? | 15:19 |
| yolanda | yes, i think you should start by that, and also be involved into the zuul v3 spec. Once done that, i think core reviewers of zuul will have better opinion than me, on how your change can fit there, and how it can be improved | 15:19 |
| openstackgerrit | Om Kumar proposed openstack/diskimage-builder: Adds Ubuntu and Debian to ironic-python-agent Support-list https://review.openstack.org/198771 | 15:19 |
| mordred | pipsi! | 15:20 |
| mordred | fungi, https://github.com/mitsuhiko/pipsi <-- sdague | 15:20 |
| eantyshev | yolanda: okay, let's try that | 15:20 |
| sdague | mordred: awesome | 15:20 |
| *** jsavak has quit IRC | 15:21 | |
| fungi | mordred: oh, hadn't even heard of that one | 15:21 |
| *** jsavak has joined #openstack-infra | 15:21 | |
| fungi | mordred: however, it does basically look like an automation of my workflow | 15:21 |
| mordred | yah. that's why I brought it up - I believe many of us do a similar thing | 15:22 |
| jeblair | oh this makes ~/bin/gertty for me? | 15:22 |
| *** ameade has quit IRC | 15:22 | |
| sdague | jeblair: yeh | 15:22 |
| jeblair | neato | 15:22 |
| fungi | jeblair: well, ~/.local/bin/gertty according to the readme, but probably configurable | 15:22 |
| sdague | I ghetto'ed up a thing | 15:22 |
| *** ameade_vacation has joined #openstack-infra | 15:22 | |
| fungi | it's only a few commands to make a virtualenv, install something into it, and set up some ~/bin and ~/man symlinks to things i want | 15:23 |
| *** ameade_vacation is now known as ameade | 15:23 | |
| fungi | but neat that there's something to wrap that automagically | 15:23 |
| openstackgerrit | Adam Coldrick proposed openstack-infra/storyboard-webclient: Show project list rather than landing page by default https://review.openstack.org/198319 | 15:23 |
| *** flepied has quit IRC | 15:25 | |
| sdague | fungi: honestly, I sort of wish pip just had a mode to work like that out of the box | 15:26 |
| fungi | i actually even do that with virtualenv itself (i have ~/bin/virtualenv symlinked to ~/pyenvs/virtualenv/bin/virtualenv) since i only like to use the distro virtualenv binary as a first-stage bootstrapper to get latest virtualenv | 15:26 |
| fungi | sdague: me too | 15:26 |
| fungi | sdague: i mean, i never ever ever 'sudo pip ...' anything on my systems | 15:26 |
| fungi | because i don't like it fighting with the distro package mangler | 15:27 |
| fungi | admittedly, this is currently biting me in a few places. for example beets | 15:27 |
| *** tsekiyama has joined #openstack-infra | 15:27 | |
| *** hodos has joined #openstack-infra | 15:27 | |
| openstackgerrit | Daniel Izquierdo proposed openstack-infra/activity-board: OpenStack community activity report Q2, 2015 https://review.openstack.org/192126 | 15:28 |
| fungi | i want beets to do chromaprint/acoutsid classifying, but that needs gobject python extensions which can't run at all in a virtualenv and don't really work terribly well even with a virtualenv that has sitepackages enabled | 15:28 |
| fungi | mainly because gnome devs seem to not really care about virtualenv-based workflows at all | 15:29 |
| Daviey | fungi: considering i mostly use one venv, i just have an alias - alias venv="/bin/bash --rcfile ${HOME}/openstack/.venv/bin/activate" | 15:29 |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config: Create grafana.o.o under -infra https://review.openstack.org/179208 | 15:29 |
| fungi | Daviey: interesting idea. i'm just not excited about comingling dependencies between different python tools, though that may change once lifeless's dep solver makes it into a pip release | 15:30 |
| Daviey | fungi: in theory, global-requirements should have already solved the global playground problem... no? | 15:31 |
| sdague | mordred: so, I am clearly failing on the mental model of correctly building tests for pbr - https://review.openstack.org/#/c/196203/ - any chance you can help me unwind this? | 15:31 |
| mordred | sdague: I can try! | 15:31 |
| fungi | Daviey: i'm talking about random python-based tools, not necessarily anything openstack-specific | 15:31 |
| Daviey | oh | 15:31 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add stable ACLs for networking-foo projects https://review.openstack.org/198750 | 15:32 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add a create ACL for networking-bgpvpn https://review.openstack.org/198751 | 15:32 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Update ACLs for networking-foo projects https://review.openstack.org/198749 | 15:32 |
| mordred | sdague: well, in good news, it's only your test that doesn't work | 15:32 |
| openstackgerrit | Sean Dague proposed openstack-infra/project-config: remove swift-dsvm-functional from blocking changes to many projects https://review.openstack.org/196085 | 15:34 |
| openstackgerrit | Stephanie Miller proposed openstack-infra/project-config: Add tests for bifrost https://review.openstack.org/190327 | 15:34 |
| sdague | mordred: right, however, I clearly don't quite understand the setup magic I have access to here | 15:34 |
| mordred | ALL THE MAGIC | 15:35 |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config: Create grafana.o.o under -infra https://review.openstack.org/179208 | 15:36 |
| openstackgerrit | Om Kumar proposed openstack/diskimage-builder: Adds Ubuntu and Debian to ironic-python-agent Support-list https://review.openstack.org/198771 | 15:37 |
| *** annegentle has joined #openstack-infra | 15:37 | |
| *** bdemers has quit IRC | 15:38 | |
| *** dizquierdo has quit IRC | 15:38 | |
| *** ashleighfarnham has joined #openstack-infra | 15:41 | |
| *** dustins_ has quit IRC | 15:41 | |
| *** pal has quit IRC | 15:41 | |
| *** sabeen1 has joined #openstack-infra | 15:42 | |
| *** thedodd has joined #openstack-infra | 15:42 | |
| mordred | sdague: I believe it's because you're running install_scripts not install - which means that the library itself is not getting installed | 15:42 |
| mordred | sdague: patch coming | 15:42 |
| *** bdemers has joined #openstack-infra | 15:42 | |
| sdague | mordred: ok | 15:43 |
| EmilienM | nibalizer: I'm interested by bringing some log files in our puppet jobs jenkins workdir. Do you have any pointer? | 15:43 |
| *** harlowja_at_home has joined #openstack-infra | 15:44 | |
| nibalizer | look at the 'publisher' section of jjb | 15:44 |
| openstackgerrit | Monty Taylor proposed openstack-dev/pbr: WIP: tests for wsgi_scripts support https://review.openstack.org/196203 | 15:44 |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config: Bump puppet-memcached to v2.6.0 https://review.openstack.org/192741 | 15:44 |
| nibalizer | you can set paths and shellglobs of files to pull | 15:44 |
| mordred | sdague: ^^ that gets as far as hanging on the running wsgi process now | 15:44 |
| EmilienM | nibalizer: ok thanks | 15:45 |
| *** btully has quit IRC | 15:47 | |
| sdague | mordred: awesome, great, let me poke from here | 15:47 |
| *** zhiyan has quit IRC | 15:47 | |
| *** gyee has joined #openstack-infra | 15:47 | |
| *** zhiyan has joined #openstack-infra | 15:47 | |
| mordred | sdague: tl;dr is - use --prefix and setup.py install so that the library code is actually installed into a pythonpath-able location | 15:48 |
| *** Sukhdev has joined #openstack-infra | 15:48 | |
| *** yfried has quit IRC | 15:49 | |
| *** yfried has joined #openstack-infra | 15:49 | |
| *** armax has joined #openstack-infra | 15:50 | |
| *** scheuran has quit IRC | 15:51 | |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add stable ACLs for networking-foo projects https://review.openstack.org/198750 | 15:53 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add a create ACL for networking-bgpvpn https://review.openstack.org/198751 | 15:53 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Update ACLs for networking-foo projects https://review.openstack.org/198749 | 15:53 |
| *** patrickeast has joined #openstack-infra | 15:53 | |
| *** Ala has quit IRC | 15:55 | |
| greghaynes | fungi: Hey, so you said the IAD upload worked for ubuntu-trusty right? Does that mean kicking off the experimental job for it will schedule there? | 15:56 |
| *** sabeen2 has joined #openstack-infra | 15:56 | |
| *** sabeen1 has quit IRC | 15:56 | |
| zaro | pelix: re https://review.openstack.org/#/c/179455/4/jenkins_jobs/cmd.py if you feel that would failing would be better then go ahead with another patch and i'll +2. not a big deal though | 15:57 |
| zaro | morning | 15:58 |
| mordred | greghaynes: should do | 15:58 |
| *** puranamr has joined #openstack-infra | 15:58 | |
| fungi | greghaynes: i manually reran nodepool image-upload to upload them all again over the weekend too | 15:59 |
| greghaynes | oh | 15:59 |
| greghaynes | fungi: and good or bad? | 15:59 |
| zaro | pelix: any interest with helping out on this change? https://review.openstack.org/#/c/180185 | 16:00 |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config: Remove line break (\n) from modulesenv diff https://review.openstack.org/198780 | 16:00 |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config: Add modulesenv to tox envlist https://review.openstack.org/198781 | 16:00 |
| *** Longgeek has joined #openstack-infra | 16:00 | |
| *** flepied has joined #openstack-infra | 16:00 | |
| fungi | greghaynes: they all uploaded successfully. i suspect running them one at a time reduced the contention or interface saturation | 16:01 |
| greghaynes | oh, nice | 16:01 |
| *** ddieterly has quit IRC | 16:01 | |
| greghaynes | I wonder if our nodepool builders will also help this then | 16:01 |
| fungi | likely | 16:01 |
| *** puranamr_ has quit IRC | 16:02 | |
| fungi | remember that rackspace caps bandwith on instances depending on flavor. i have a feeling their "rate limiting" implementation is sort of packet-lossy | 16:02 |
| *** thorst is now known as thorst_afk | 16:02 | |
| greghaynes | yea, I think we've discovered that before | 16:02 |
| pelix | zaro: yeah, think I'll change that to exit. the exception might be a little messy, but probably need to re-org that code so that the parsing of the config options set is done earlier and any exceptions can be caught with a parser.error() used to exit | 16:02 |
| pelix | can do the clean up at a later stage? | 16:02 |
| *** ddieterly has joined #openstack-infra | 16:02 | |
| pc_m | Anyone know about db migrations? | 16:03 |
| *** rwsu has joined #openstack-infra | 16:04 | |
| *** devvesa has quit IRC | 16:05 | |
| *** Longgeek has quit IRC | 16:05 | |
| pabelanger | SergeyLukjanov, mordred Wanted to see of there was any updates from mirantis about stackalytics spec proposal? | 16:05 |
| pabelanger | was out most of last week | 16:05 |
| zaro | pelix: clean up? are you referring to folders support or jenkins connection config? | 16:06 |
| *** doug-fish has joined #openstack-infra | 16:07 | |
| fungi | pc_m: zzzeek in #openstack-dev and #openstack-oslo is the primary sqlalchemy maintainer, if your questions are of a particularly complex nature | 16:07 |
| openstackgerrit | Paul Belanger proposed openstack-infra/grafyaml: Add datasource functionality https://review.openstack.org/194265 | 16:08 |
| *** esmute__ has joined #openstack-infra | 16:09 | |
| *** ashleighfarnham has quit IRC | 16:09 | |
| *** markvoelker has joined #openstack-infra | 16:10 | |
| jeblair | i'm going to be in and out today, sorry :( | 16:11 |
| openstackgerrit | Dave McCowan proposed openstack/requirements: Fix version conflict between PBR and requirements https://review.openstack.org/198790 | 16:11 |
| *** mpavone has quit IRC | 16:11 | |
| pc_m | fungi: Thanks. Will ping over in dev | 16:12 |
| pc_m | fungi: Some are basic, and some are complex. | 16:13 |
| *** laughterwym has quit IRC | 16:13 | |
| *** pal has joined #openstack-infra | 16:13 | |
| pelix | zaro: that we essentially doing complex arg and config parsing inside the execute function where any exceptions thrown should be due to bad options or config provided by the user | 16:14 |
| openstackgerrit | Paul Belanger proposed openstack-infra/grafyaml: Add datasource functionality https://review.openstack.org/194265 | 16:14 |
| pelix | those thrown around config issues should probably be caught and just print an exit message to the user about bad input | 16:15 |
| *** bdemers_ has joined #openstack-infra | 16:16 | |
| *** koolhead_ has joined #openstack-infra | 16:16 | |
| *** ashleighfarnham has joined #openstack-infra | 16:17 | |
| *** ashleighfarnham has quit IRC | 16:17 | |
| *** ashleighfarnham has joined #openstack-infra | 16:17 | |
| *** sshnaidm has quit IRC | 16:17 | |
| *** skolekonov has joined #openstack-infra | 16:18 | |
| *** koolhead17 has quit IRC | 16:18 | |
| *** bdemers has quit IRC | 16:19 | |
| *** bdemers_ is now known as bdemers | 16:19 | |
| nibalizer | does 181965 'latest gerrit' mean 2.8 or 2.10? Is that change still needed? | 16:20 |
| *** jaypipes is now known as jay-deepthought | 16:21 | |
| openstackgerrit | Emilien Macchi proposed openstack-infra/project-config: puppet: publish logs dir for Beaker jobs https://review.openstack.org/198802 | 16:21 |
| EmilienM | nibalizer: ^ | 16:22 |
| anteaya | sdague: once you and notmyname reach an understanding let me know | 16:22 |
| sdague | anteaya: sure | 16:22 |
| openstackgerrit | Emilien Macchi proposed openstack-infra/project-config: puppet: publish logs dir for Beaker jobs https://review.openstack.org/198802 | 16:23 |
| EmilienM | nibalizer: had to repush | 16:24 |
| *** yfried has quit IRC | 16:24 | |
| *** btully has joined #openstack-infra | 16:24 | |
| pleia2 | good morning | 16:25 |
| *** otter768 has quit IRC | 16:26 | |
| *** shashank_hegde has joined #openstack-infra | 16:27 | |
| anteaya | pleia2: morning | 16:27 |
| pleia2 | anteaya: productive mid-cycle last week, I hope? :) | 16:28 |
| *** harlowja_at_home has quit IRC | 16:28 | |
| *** dustins has joined #openstack-infra | 16:28 | |
| *** dpose has quit IRC | 16:28 | |
| openstackgerrit | Kyle Mestery proposed openstack-infra/irc-meetings: Update neutron-drivers team meeting times. https://review.openstack.org/198806 | 16:29 |
| Odd_Bloke | Hello all, does anyone know why https://review.openstack.org/#/c/198308/ hasn't been queued for gate? It has two Code Review +2s and a Workflow +1, which is all I would expect to be necessary... | 16:30 |
| openstackgerrit | Paul Belanger proposed openstack-infra/infra-specs: Host trystack.o.o site https://review.openstack.org/195098 | 16:31 |
| *** ociuhandu has quit IRC | 16:31 | |
| *** pal has quit IRC | 16:31 | |
| *** whoops has joined #openstack-infra | 16:32 | |
| hogepodge | Which project is appropriate for setting up a new irc room? | 16:32 |
| *** pal has joined #openstack-infra | 16:32 | |
| openstackgerrit | Merged stackforge/python-jenkins: Creates ability to work with views in jenkins https://review.openstack.org/194307 | 16:33 |
| hogepodge | I've volunteered to set up #openstack-diversity | 16:33 |
| pleia2 | hogepodge: you start with project config, let me grab the link | 16:33 |
| pabelanger | hogepodge, http://docs.openstack.org/infra/system-config/irc.html should be a good starting point | 16:33 |
| *** kun_huang has left #openstack-infra | 16:34 | |
| *** davideagnello has joined #openstack-infra | 16:35 | |
| pleia2 | hogepodge: you edit accessbot/channels.yaml in project config, if you want gerrit to post any merge-things (doubtful for -diversity) you edit gerritbot/channels.yaml and if you want logging you go over to submit a patch in system-config and edit modules/openstack_project/manifests/eavesdrop.pp | 16:35 |
| *** abregman has joined #openstack-infra | 16:35 | |
| fungi | Odd_Bloke: looks like workflow +1 was added, zuul saw the event (because it triggers gate enqueuing on comment-added with a workflow +1), checked to see if it was otherwise ready to merge, noticed it had no +2 code-review votes, and stopped | 16:35 |
| pleia2 | hogepodge: the system-config change will have to depend on the project-config change (we need it added to accessbot before we add logging) | 16:35 |
| pabelanger | nibalizer, easy +1 workflow for you: https://review.openstack.org/#/c/190705/ | 16:35 |
| anteaya | pleia2: it is always good to be able to sit down and look into someone's eyes | 16:35 |
| pabelanger | nibalizer, at least I think you have +1 workflow | 16:35 |
| fungi | Odd_Bloke: later smoser and harlowja added code-review +2 votes, but that alone doesn't trigger zuul to try to enqueue a change into the gate pipeline | 16:35 |
| openstackgerrit | Paul Belanger proposed openstack-infra/project-config: Import trystack.org static contents https://review.openstack.org/195117 | 16:36 |
| anteaya | pleia2: plus I just can't hear gus laugh while he types :) | 16:36 |
| nibalizer | pabelanger: oh man, idunno about this change | 16:36 |
| anteaya | pleia2: tough things were discussed, much drawing on whiteboards | 16:36 |
| nibalizer | :) | 16:36 |
| pabelanger | nibalizer, MERGE ALLL THE THINGS | 16:36 |
| anteaya | pleia2: action items were actioned | 16:36 |
| anteaya | pleia2: thanks for asking | 16:36 |
| fungi | Odd_Bloke: either smoser can remove and reapply his workflow +1 to trigger gating for 198308, or harlowja or another core reviewer on stackforge/cloud-init can add an additional workflow +1 to trigger it | 16:36 |
| pleia2 | anteaya: wonderful, sounds like it went well then! | 16:37 |
| anteaya | pleia2: it was a productive gathering, yes | 16:37 |
| hogepodge | pleia2: pabelanger: thanks! | 16:37 |
| smoser | fungi, trying. thanks. | 16:37 |
| *** davideagnello has quit IRC | 16:37 | |
| anteaya | pleia2: boston next week with keystone | 16:38 |
| anteaya | common ci virtual sprint end of this week | 16:38 |
| pabelanger | yolanda, any preference for passing crontab values into puppet modules? | 16:38 |
| *** packet has joined #openstack-infra | 16:39 | |
| pleia2 | anteaya: ah, July :) | 16:40 |
| anteaya | yeah | 16:40 |
| pleia2 | anteaya: I enjoy Boston in the summer? I love their aquarium | 16:40 |
| anteaya | oh? | 16:40 |
| anteaya | I've never been | 16:41 |
| pleia2 | s/?/! | 16:41 |
| anteaya | not sure i will have time | 16:41 |
| pleia2 | it can be hot, but t he breezes off the cove are lovely | 16:41 |
| pabelanger | nibalizer, So, was thinking I'd start the task of bumping the puppet-apache module in system-config. Not sure if I'm asking for a betting or what. | 16:41 |
| anteaya | but thanks | 16:41 |
| anteaya | oh good | 16:41 |
| anteaya | i like a good breeze | 16:41 |
| pabelanger | nibalizer, maybe the effort should be driven from the 3rd party infra sprint this week | 16:41 |
| nibalizer | pabelanger: wellllllllllll so there are two schools of thought on where we could go on that | 16:42 |
| *** eharney has quit IRC | 16:42 | |
| mordred | uhoh. puppet-apache | 16:42 |
| nibalizer | one is to bump the apache module version, the other is to use our fork of the apache module pupet-httpd | 16:42 |
| * mordred gets ready to throw vegetables at nibalizer | 16:42 | |
| nibalizer | the neat thing about using puppet-httpd is that then anyone who wants to use puppetlabs-apache-new can do so while consuming our stuff | 16:43 |
| nibalizer | possibly even us | 16:43 |
| *** maishsk has joined #openstack-infra | 16:43 | |
| pabelanger | nibalizer, My thought was to actually sed puppet-apache to puppet-infra-apache, and then add puppet-apache from upstream. Then go module by module and convert | 16:43 |
| nibalizer | there is also unsubstantiated rumours that the critical features we use in 0.0.4 are available in the new puppet-apache module | 16:43 |
| nibalizer | pabelanger: ya we did that and made it openstack-infra/puppet-httpd | 16:43 |
| *** davideagnello has joined #openstack-infra | 16:44 | |
| pabelanger | nibalizer, Ya, didn't know of that module | 16:44 |
| nibalizer | the second step (module by module to convert) is what we haven't done yet | 16:44 |
| openstackgerrit | Darragh Bailey proposed openstack-infra/jenkins-job-builder: Support anonymous credentials for Jenkins plugin info query https://review.openstack.org/179455 | 16:44 |
| nibalizer | personally I'm a big fan of the fork | 16:44 |
| nibalizer | but the effort has definitely lost steam so its worth considering alternatives I think | 16:44 |
| mordred | as long as we can pass in templated apache module files with the latest puppet-apache from upstream, I think it's workable | 16:45 |
| nibalizer | personally I don't really like the newer puppetlabs-apache, and I think we'll alientate as many people by requiring a very new verson of it as we do when we require a very old version of it | 16:45 |
| mordred | the biggest problem with it is that they removed that | 16:45 |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Remove deprecated disk-image-get-kernel https://review.openstack.org/198810 | 16:45 |
| pabelanger | nibalizer, did you have a spec for puppet-httpd convert? | 16:46 |
| nibalizer | pabelanger: don't think so, I think we may have started that before specs? | 16:46 |
| *** rbrooker has quit IRC | 16:46 | |
| *** e0ne has quit IRC | 16:47 | |
| pabelanger | nibalizer, so, to make sure I'm on the same page. You were going to replace apache::vhost with httpd::vhost, which was the fork of puppet-apache 0.0.4 | 16:47 |
| openstackgerrit | Merged openstack-infra/puppet-stackalytics: Add .gitreview https://review.openstack.org/190705 | 16:47 |
| fungi | regardless, i think the temporary move to our httpd module is necessary before we could embark on a piecemeal upgrade of some of our modules to the new apache module | 16:47 |
| fungi | since we won't be able to run multiple versions of the actual apache module in parallel | 16:48 |
| *** ildikov has quit IRC | 16:48 | |
| pabelanger | fungi, right, that makes sense to me. And seem straight forward to replace apache:: with httpd:: | 16:48 |
| pabelanger | assuming httpd:: is 0.0.4 fork | 16:49 |
| *** koolhead_ has quit IRC | 16:49 | |
| nibalizer | pabelanger: yes | 16:49 |
| pleia2 | ollld fork | 16:49 |
| nibalizer | also any a2Mod resources | 16:49 |
| nibalizer | are now httpmod resouces or whatever I neded up naming them | 16:49 |
| pabelanger | okay, I'll start poking around with that today and see what work effort it requries | 16:50 |
| pabelanger | requires* | 16:50 |
| nibalizer | and I've personally used openstack-infra/puppet-httpd for a lil side project so it is a functional module | 16:50 |
| *** tonytan4ever has quit IRC | 16:50 | |
| nibalizer | we didn't unwire anything accidentally | 16:50 |
| pabelanger | then, work more on it during the sprint | 16:50 |
| nibalizer | pabelanger: sweet | 16:50 |
| *** annegentle has quit IRC | 16:50 | |
| *** jbernard has quit IRC | 16:50 | |
| *** taron has quit IRC | 16:50 | |
| *** ondergetekende_ has quit IRC | 16:50 | |
| *** ondergetekende has joined #openstack-infra | 16:50 | |
| openstackgerrit | Kyle Mestery proposed openstack-infra/irc-meetings: Update neutron-drivers team meeting times. https://review.openstack.org/198806 | 16:50 |
| *** samuelBartel has quit IRC | 16:50 | |
| *** jbernard has joined #openstack-infra | 16:51 | |
| *** taron has joined #openstack-infra | 16:51 | |
| *** Swami has joined #openstack-infra | 16:51 | |
| *** tqtran has joined #openstack-infra | 16:53 | |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Document what our stable interfaces are https://review.openstack.org/198814 | 16:53 |
| *** annegentle has joined #openstack-infra | 16:54 | |
| *** ajmiller__ has joined #openstack-infra | 16:54 | |
| *** ajmiller has quit IRC | 16:56 | |
| *** eharney has joined #openstack-infra | 16:57 | |
| *** maishsk has quit IRC | 16:58 | |
| openstackgerrit | Khai Do proposed stackforge/python-jenkins: Support scriptText api to execute groovy scripts on the server https://review.openstack.org/194468 | 16:59 |
| *** _nadya_ has quit IRC | 17:00 | |
| *** Longgeek has joined #openstack-infra | 17:01 | |
| *** derekh has quit IRC | 17:02 | |
| *** amotoki has quit IRC | 17:04 | |
| *** doude has quit IRC | 17:06 | |
| *** Longgeek has quit IRC | 17:06 | |
| *** doude has joined #openstack-infra | 17:06 | |
| *** doude has quit IRC | 17:06 | |
| *** doude has joined #openstack-infra | 17:06 | |
| *** thorst_afk is now known as thorst | 17:07 | |
| *** yfried has joined #openstack-infra | 17:10 | |
| *** markus_z has quit IRC | 17:10 | |
| *** maishsk has joined #openstack-infra | 17:10 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config: WIP: Migrate to puppet-httpd module https://review.openstack.org/198819 | 17:12 |
| ddieterly | is zuul having problems today? | 17:17 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add stable ACLs for networking-foo projects https://review.openstack.org/198750 | 17:17 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add a create ACL for networking-bgpvpn https://review.openstack.org/198751 | 17:17 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Update ACLs for networking-foo projects https://review.openstack.org/198749 | 17:17 |
| *** puranamr has quit IRC | 17:17 | |
| fungi | ddieterly: you'll have to be more specific | 17:18 |
| *** puranamr has joined #openstack-infra | 17:18 | |
| ddieterly | fungi: sure, i have 2 jobs in the 'post' queue that have been there for 3 hours | 17:18 |
| fungi | ddieterly: the post pipeline gets a lower priority than most of the others (gate, check, release, pre-release, et cetera) | 17:19 |
| ddieterly | ok, so then 3 hrs is within *normal* | 17:19 |
| *** mrunge has joined #openstack-infra | 17:19 | |
| fungi | ddieterly: when there are still changes in gate/check with jobs waiting on available workers, yes that's fairly normal | 17:20 |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-openstackci: Migrate to puppet-httpd module https://review.openstack.org/198820 | 17:21 |
| *** soren has joined #openstack-infra | 17:21 | |
| fungi | i especially like the "evidence planting" mechanisms in their rcs toolkit | 17:24 |
| *** nihilifer has joined #openstack-infra | 17:24 | |
| ddieterly | fungi: thanks | 17:24 |
| fungi | oh, heh, wrong channel on that last comment. leaked over from discussions elsewhere about the hackedteam incident | 17:25 |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-zuul: Migrate to puppet-httpd module https://review.openstack.org/198824 | 17:25 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add stable ACLs for Neutron sub-projects https://review.openstack.org/198750 | 17:26 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add a create ACL for networking-bgpvpn https://review.openstack.org/198751 | 17:26 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Update ACLs for Neutron sub-projects https://review.openstack.org/198749 | 17:26 |
| *** btully has quit IRC | 17:26 | |
| *** koolhead17 has joined #openstack-infra | 17:27 | |
| *** ajmiller__ has quit IRC | 17:28 | |
| *** nihilifer has quit IRC | 17:28 | |
| *** puranamr has quit IRC | 17:28 | |
| *** marzif_ has quit IRC | 17:28 | |
| *** ajmiller has joined #openstack-infra | 17:29 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-jenkins: Migrate to puppet-httpd module https://review.openstack.org/198825 | 17:30 |
| *** shashank_hegde has quit IRC | 17:31 | |
| *** jsavak has quit IRC | 17:32 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-openstackci: Migrate to puppet-httpd module https://review.openstack.org/198820 | 17:32 |
| anteaya | fungi: you do have some interesting wrong channel comments | 17:32 |
| *** jsavak has joined #openstack-infra | 17:32 | |
| *** tonytan4ever has joined #openstack-infra | 17:33 | |
| *** kgiusti has left #openstack-infra | 17:35 | |
| *** mpaolino has quit IRC | 17:36 | |
| *** puranamr has joined #openstack-infra | 17:36 | |
| fungi | anteaya: yeah apparently a contract firm that was ostensibly providing "cyber security" tools to oppressive governments got compromised and something like 400gb of their source code and contracts/documents have been published anonymously on the internet | 17:37 |
| pabelanger | nibalizer, shouldn't we run 0.0.4 for puppet-httpd? https://github.com/openstack-infra/puppet-httpd/blob/master/metadata.json#L3 | 17:38 |
| pabelanger | just to keep things simple? | 17:38 |
| pabelanger | or do you want to reset to 0.0.1? | 17:38 |
| openstackgerrit | Merged openstack-infra/puppet-statusbot: Add support for passing channels as list https://review.openstack.org/193802 | 17:38 |
| nibalizer | pabelanger: oh thats a whole conversation | 17:38 |
| *** ajmiller has quit IRC | 17:38 | |
| nibalizer | so the idea is the release process should write out the version from the git tag | 17:38 |
| nibalizer | so all the modules are 0.0.0 until tagged and pushed to the forge, where they will get versions | 17:38 |
| *** ajmiller has joined #openstack-infra | 17:38 | |
| *** amit213 has quit IRC | 17:39 | |
| nibalizer | the idea being that 0.0.0 is a valid version but weird enough that people would know to look at the git metadata for determining version | 17:39 |
| *** amit213 has joined #openstack-infra | 17:39 | |
| *** jay-deepthought is now known as jaypipes | 17:40 | |
| *** amit213 has quit IRC | 17:40 | |
| nibalizer | something like this https://review.openstack.org/#/c/175156/2/tools/module_versions.sh is needed to determine what version of code you're actually running if you clone from git | 17:40 |
| *** mrmartin has joined #openstack-infra | 17:40 | |
| pabelanger | nibalizer, okay, sounds good | 17:40 |
| *** amit213 has joined #openstack-infra | 17:40 | |
| *** _ody has quit IRC | 17:41 | |
| jeblair | fungi: clearly, they were framed! | 17:41 |
| nibalizer | there is code in jjb and secrets on the review slave to automate the module release process | 17:41 |
| nibalizer | but that code path hasn't really been exercised much | 17:41 |
| *** amit213 has quit IRC | 17:41 | |
| nibalizer | i suppose that I'm technically in a position to 'release' an infra module, I think the gerrit perms line that up | 17:41 |
| fungi | jeblair: indeed. someone must have planted those documents on their systems ;) | 17:41 |
| *** amit213 has joined #openstack-infra | 17:42 | |
| *** esmute__ has quit IRC | 17:42 | |
| jeblair | nibalizer: oh, i bet we should write up some docs on the puppet module release mechanics | 17:42 |
| *** amit213 has quit IRC | 17:42 | |
| anteaya | fungi: wow | 17:42 |
| *** amit213 has joined #openstack-infra | 17:42 | |
| *** mrmartin has quit IRC | 17:43 | |
| nibalizer | jeblair: and also make it work :P | 17:43 |
| jeblair | nibalizer: both of those things is a tall order, i'd settle for one. ;) | 17:43 |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/bindep: Update documentation from cookiecutter https://review.openstack.org/198504 | 17:43 |
| nibalizer | hehe | 17:43 |
| anteaya | fungi: this is going to be a fun rat's nest for the conspiracy crowd | 17:44 |
| *** mrmartin has joined #openstack-infra | 17:44 | |
| *** kgiusti has joined #openstack-infra | 17:45 | |
| *** _ody has joined #openstack-infra | 17:45 | |
| *** kgiusti has joined #openstack-infra | 17:47 | |
| SpamapS | hm | 17:47 |
| SpamapS | I wonder if, as soon-to-be-operators, we should consider sending infra people to the ops meetup. | 17:48 |
| *** dustins has quit IRC | 17:48 | |
| *** Sukhdev has quit IRC | 17:48 | |
| *** Sukhdev has joined #openstack-infra | 17:48 | |
| jeblair | hasn't been scheduled, right? | 17:49 |
| pleia2 | where is it? | 17:49 |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-askbot: Migrate to puppet-httpd module https://review.openstack.org/198827 | 17:49 |
| anteaya | SpamapS: I'm also lining up for details, as I see nothing on the sprints wikipage | 17:50 |
| pleia2 | my summer schedule is already a bit brain-explode, but I certainly would find it valuable if the timing works | 17:50 |
| SpamapS | It has not been scheduled, there's quite a large thread about it outgrowing the "just do it in our corporate HQ" model. | 17:50 |
| pleia2 | "our"? | 17:51 |
| SpamapS | as in, no one company is willing ot take on the entire cost anymore. | 17:51 |
| *** Hal has joined #openstack-infra | 17:51 | |
| pleia2 | aha | 17:51 |
| SpamapS | https://wiki.openstack.org/wiki/Operations/Meetups#Venue_Selection | 17:51 |
| *** Hal is now known as Guest98271 | 17:51 | |
| SpamapS | It's now bigger than the first, what, 6 summits? | 17:52 |
| mordred | SpamapS: the last ops midcycle was already at that point - and in fact the foundation had to step in at the last minute and help secure alternate venue | 17:52 |
| mordred | yup | 17:52 |
| jeblair | it will also not fit in my apartment | 17:52 |
| mordred | jeblair: bah | 17:52 |
| *** esmute__ has joined #openstack-infra | 17:52 | |
| mordred | jeblair: I reject that thought | 17:52 |
| jeblair | we can try for mordred's patio though | 17:52 |
| mordred | jeblair: it fits at your apartment - anything that doesn't fit is a bug :) | 17:52 |
| fungi | i heard the last operators mid-cycle was >300 attendees | 17:53 |
| * jeblair calls exterminators | 17:53 | |
| *** dustins has joined #openstack-infra | 17:53 | |
| fungi | that's well into the range of a smallish tech conference | 17:53 |
| SpamapS | jeblair: I've seen some film reels from the 1950's that suggest no matter how small your apartment is, _WE SHOULD TRY_ | 17:53 |
| *** maishsk has quit IRC | 17:53 | |
| mordred | SpamapS: ++ | 17:54 |
| SpamapS | for America | 17:54 |
| * mordred is happy to suggest hosting it on my roof | 17:54 | |
| anteaya | do they have some dates in mind? | 17:54 |
| mordred | the fireworks on saturday showed that my roof can CLEARLY support more than 300 people | 17:54 |
| anteaya | we close in on something if we expect to have representation | 17:54 |
| *** e0ne has joined #openstack-infra | 17:54 | |
| *** pal has quit IRC | 17:54 | |
| SpamapS | according to the wiki page there were 150 attendees in Philadelphia | 17:55 |
| Daviey | SpamapS: this, http://mentalfloss.com/sites/default/legacy/blogs/wp-content/uploads/2012/03/life_booth_1.jpg ? | 17:55 |
| mordred | SpamapS: there were many | 17:56 |
| SpamapS | Daviey: exactly. That looks like an extremely productive high-bandwidth working environment. | 17:56 |
| *** jasondotstar has quit IRC | 17:56 | |
| SpamapS | anyway, I suggest that we are noobs in operating OpenStack, and we should send 1 or 2 people to discuss that with operators, and also help facilitate any infra-discussion that happens as a target of opportunity. | 17:57 |
| jeblair | SpamapS: ++ | 17:57 |
| *** _diana_ has joined #openstack-infra | 17:57 | |
| SpamapS | Also there's a puppet mid-cycle, but I think we can just make crinkle go to that. | 17:57 |
| anteaya | I like montreal | 17:57 |
| *** shashank_hegde has joined #openstack-infra | 17:57 | |
| SpamapS | s/make/ask/ ;) | 17:57 |
| anteaya | and always look for an opporunity to go | 17:58 |
| SpamapS | anteaya: who doesn't? :) | 17:58 |
| greghaynes | heh | 17:58 |
| greghaynes | Yea, ops summit is nice, it has definitely exploded | 17:58 |
| anteaya | however knowing the dates for the ops mid-cycle would be helpful in voting on EmilienM's doodle poll | 17:58 |
| greghaynes | SpamapS: So - you mentioned that the bm cloud is to a point where we can maybe deploy some things? | 17:58 |
| EmilienM | anteaya: hey | 17:59 |
| anteaya | EmilienM: hey | 17:59 |
| greghaynes | SpamapS: The NOC has found some inconsistencies in our inventory data, hoping I can help resolve them if possible... | 17:59 |
| timothyb89 | fungi: looking at http://puppetboard.openstack.org/node/zuul.openstack.org it appears the a2mod cgi isn't converging. I've reproduced locally and it seems to be because a2enmod selects+enables cgid instead of cgi | 17:59 |
| anteaya | EmilienM: a chance to visit you, how could I say no? | 17:59 |
| timothyb89 | fungi: changing puppet config to use cgid allows the config to converge, but is this a safe change to make for all environments? or would it be better to conditionally accept both cgi and cgi? | 17:59 |
| EmilienM | anteaya: say yes then :) | 17:59 |
| anteaya | EmilienM: I'm trying | 17:59 |
| EmilienM | I don't see OPS midcycle in https://wiki.openstack.org/wiki/Sprints | 17:59 |
| SpamapS | greghaynes: it cannot deploy things yet, because I have to re-do the vlan+ovs setup since I used regular vlan tagging which apparently stops ovs from working. | 17:59 |
| anteaya | EmilienM: trying to find out dates for the ops mid-cycle | 17:59 |
| anteaya | EmilienM: me either | 18:00 |
| greghaynes | aye | 18:00 |
| EmilienM | so... | 18:00 |
| fungi | timothyb89: good question... asselin: do you suspect that's an upshot of the puppet-openstackci migration for our use of the zuul module? | 18:00 |
| SpamapS | greghaynes: once I do that I hope dnsmasq will see the DHCP reqs in its flow, and then we should be able to deploy. | 18:00 |
| greghaynes | SpamapS: I didnt realize we had decided to use ovs anywhere? | 18:00 |
| anteaya | EmilienM: I never planned on attending the ops mid-cycle last time then it turns out I should have been there, not saying I'm going but just trying to get a sense of things | 18:00 |
| greghaynes | SpamapS: or is that just how the puppet modules do it? | 18:00 |
| SpamapS | greghaynes: we hadn't. I am just following ironic's install guide which says this is how neutron should be configured. | 18:00 |
| fungi | timothyb89: i wonder whether that's specific to particular platforms/apache versions | 18:01 |
| greghaynes | SpamapS: ok, gotcha | 18:01 |
| SpamapS | greghaynes: I believe it's just as a bridge integration tool, not as a tenant overlay tool. | 18:01 |
| SpamapS | greghaynes: as in, it lets the API talk about physical interfaces for provider networks. | 18:01 |
| *** aysyd has quit IRC | 18:02 | |
| *** abitha has joined #openstack-infra | 18:02 | |
| SpamapS | If there's way to get neutron-dhcp-agent to just listen to --interface=eth2 instead of a tap interface though.. that would be great. | 18:02 |
| greghaynes | huh... | 18:02 |
| *** pelix has quit IRC | 18:02 | |
| greghaynes | yea, that seems like a lot of overkill for just that | 18:02 |
| SpamapS | totally agree, but I do recall we used it that way in tripleo too | 18:02 |
| SpamapS | and to be fair, ovs is actually a nice one-stop-shop for all things interface related. | 18:03 |
| SpamapS | no vconfig, no brctl, just ovs | 18:03 |
| *** Kennan2 has joined #openstack-infra | 18:03 | |
| *** aysyd has joined #openstack-infra | 18:03 | |
| greghaynes | I really dont mind either way, I am definitely in a path-of-least resistence mindset for this part of the work - and if the docs say OVS then I say we go for it | 18:03 |
| *** Kennan has quit IRC | 18:04 | |
| *** e0ne is now known as e0ne_ | 18:04 | |
| timothyb89 | fungi: that is my fear as well | 18:04 |
| *** yamahata has joined #openstack-infra | 18:05 | |
| SpamapS | greghaynes: same here | 18:06 |
| *** esmute__ has quit IRC | 18:06 | |
| *** SumitNaiksatam has quit IRC | 18:06 | |
| *** SumitNaiksatam has joined #openstack-infra | 18:06 | |
| fungi | timothyb89: ftr, we're running that on ubuntu 12.04 lts with apache 2.2 | 18:06 |
| *** gyee has quit IRC | 18:07 | |
| timothyb89 | fungi: ah, I reproduced with apache 2.4.7 on ubuntu 14.04, which could change results somewhat | 18:07 |
| *** med_` is now known as med_ | 18:07 | |
| *** med_ has joined #openstack-infra | 18:07 | |
| fungi | timothyb89: in that case, i suspect it's safe to submit a patch for openstack-infra/puppet-zuul to adjust the module name to its canonical version rather than the alias | 18:08 |
| *** jasondotstar has joined #openstack-infra | 18:08 | |
| timothyb89 | fungi: if it makes any difference, apache is selecting cgid over cgi because it detects a threaded MPM | 18:09 |
| *** liam_ has joined #openstack-infra | 18:09 | |
| *** liam_ is now known as Guest20028 | 18:10 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-openstackci: Migrate to puppet-httpd module https://review.openstack.org/198820 | 18:10 |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-zuul: Migrate to puppet-httpd module https://review.openstack.org/198824 | 18:10 |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-jenkins: Migrate to puppet-httpd module https://review.openstack.org/198825 | 18:10 |
| fungi | timothyb89: oh, got it. so maybe we still go ahead and hard-code that to cgid anyway, or perhaps some additional logic is called for if we expect to have people choosing between mpm types | 18:11 |
| *** otter768 has joined #openstack-infra | 18:11 | |
| *** maishsk has joined #openstack-infra | 18:11 | |
| fungi | timothyb89: i could honestly go either way, but i'm just one voice. we can likely get wider attention to the choice in code review | 18:11 |
| timothyb89 | fungi: fair enough, I will start on a patch then, thanks for the advice! | 18:11 |
| fungi | timothyb89: you're welcome | 18:12 |
| *** jsavak has quit IRC | 18:12 | |
| *** jsavak has joined #openstack-infra | 18:12 | |
| anteaya | I asked for some details on the ops meetup thread, let's see what happens next | 18:16 |
| pleia2 | thanks anteaya | 18:17 |
| *** esmute__ has joined #openstack-infra | 18:17 | |
| *** Guest20028 has quit IRC | 18:17 | |
| anteaya | pleia2: welcome | 18:17 |
| *** cindyo has quit IRC | 18:17 | |
| *** Guest98271 has quit IRC | 18:18 | |
| *** mrmartin has quit IRC | 18:19 | |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Document what our stable interfaces are https://review.openstack.org/198814 | 18:19 |
| *** dims_ has quit IRC | 18:20 | |
| *** jsavak has quit IRC | 18:20 | |
| *** jsavak has joined #openstack-infra | 18:23 | |
| *** e0ne_ is now known as e0ne | 18:24 | |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Document our supported distributions https://review.openstack.org/198841 | 18:27 |
| *** peristeri has quit IRC | 18:27 | |
| openstackgerrit | greghaynes proposed openstack/diskimage-builder: Document our supported distributions https://review.openstack.org/198841 | 18:27 |
| clarkb | morning | 18:28 |
| anteaya | morning clarkb | 18:28 |
| anteaya | I noticed you were up late last night | 18:28 |
| greghaynes | clarkb: are you ircing with baby in lap? | 18:28 |
| greghaynes | or has the technique not been perfected yet? | 18:28 |
| clarkb | greghaynes: no thankfully they tend to be happy during the day | 18:28 |
| clarkb | anteaya: ya nights tend to be the unfun part of the day so now I am on an australian sleep schedule | 18:28 |
| anteaya | clarkb: jhesketh will enjoy the company | 18:29 |
| *** cathy_ has joined #openstack-infra | 18:30 | |
| *** jsavak has quit IRC | 18:31 | |
| *** jsavak has joined #openstack-infra | 18:32 | |
| *** kdas_ has joined #openstack-infra | 18:32 | |
| *** btully has joined #openstack-infra | 18:32 | |
| *** Sukhdev has quit IRC | 18:34 | |
| *** mrunge has quit IRC | 18:34 | |
| *** kushal has quit IRC | 18:35 | |
| openstackgerrit | Tim Buckley proposed openstack-infra/puppet-zuul: Adjust zuul's init.pp to check for and enable cgid instead of cgi. https://review.openstack.org/198845 | 18:35 |
| cathy_ | I have a question on how OpenStack (stackalytics) count the contribution? Currently the SFC spec's owner is Contributor A, the author of the first version is contributor B, later the spec is updated many versions and the last version spec author and committer is contributor C, do they all counted as contributors to this spec and get credit as ATC in OpenStack and stackalytics when the spec is merged into an OpenStack repo, suc | 18:37 |
| greghaynes | clarkb: or whoever else knows - whats the trick to get the system logs from the jenkins url https://jenkins04.openstack.org/job/experimental-nova-dsvm-python27/4/console | 18:37 |
| greghaynes | I think ^ just ran on the ubuntu-trusty node | 18:37 |
| greghaynes | want to make sure | 18:37 |
| russellb | cathy_: it's only counted based on the final patch that is merged. if there are multiple authors, the commit message should include "Co-authored-by:" lines | 18:38 |
| clarkb | greghaynes: at hte beginning of the console log first 10 lines or so it tells you which host it ran on | 18:38 |
| fungi | cathy_: i'm not certain. stackalytics is not actually an official openstack project, but its source code is available if you want to dig into what factors they track to identify that http://git.openstack.org/cgit/stackforge/stackalytics/tree/ | 18:38 |
| pleia2 | cathy_: also, stackalytics is not what is used for counting ATCs | 18:38 |
| clarkb | greghaynes: and at the end of the log there should be an http link printed for the logs.o.o logs location | 18:38 |
| greghaynes | clarkb: ah, that works | 18:39 |
| greghaynes | Building remotely on ubuntu-trusty-rax-iad-3505927 (ubuntu-trusty) | 18:39 |
| greghaynes | \O/ | 18:39 |
| greghaynes | mordred: ^ | 18:39 |
| greghaynes | fungi: tyty | 18:39 |
| mordred | greghaynes: zomg! | 18:40 |
| fungi | cathy_: as far as what we do in an official capacity to identify contributors for generating technical election voter rolls and providing complimentary summit passes, we track the gerrit "owner" of a change (which is the committer, not author, of the first patchset of a change) | 18:40 |
| clarkb | greghaynes: so does that mean we can run devstack on a deboostrap glean based ubuntu node? | 18:40 |
| mordred | clarkb: it means we're running it - remains to be seen if it will be successful :) | 18:40 |
| *** abregman has quit IRC | 18:40 | |
| cathy_ | russellb: Thanks for your reply! | 18:41 |
| greghaynes | that test passed | 18:41 |
| *** gyee has joined #openstack-infra | 18:42 | |
| cathy_ | fungi: Thanks for your reply. So is there a way to add multiple owners to a spec since the authors are the ones that make the most contribution to the spec | 18:42 |
| nibalizer | fungi: the nonconvergence of zuul predates the openstackci stuff | 18:42 |
| anteaya | the most contribution to a spec is rather hard to define | 18:43 |
| anteaya | I would argue that reviewers add the most contribution to a spec | 18:43 |
| *** ddieterly has quit IRC | 18:44 | |
| fungi | cathy_: for technical elections and complimentary summit passes (assuming you're talking about a spec for an official openstack project) you can submit a change to the openstack/governance repository's reference/extra-atcs file to add them as officially recognized contributors | 18:44 |
| cathy_ | fungi: Thanks for all the info! | 18:44 |
| fungi | cathy_: you're welcome | 18:45 |
| *** doug-fish has quit IRC | 18:46 | |
| cathy_ | fungi: also what is meant by WIP? Work in progress? Does it need to be listed for every updated commit on that spec or just needs to be listed once? Who can lift WIP? | 18:47 |
| cathy_ | pleia2: Thanks for your reply! | 18:48 |
| fungi | cathy_: yes, a "workflow -1" on a change implies it's a work in progress. it gets undone for each new patchset and has to be readded if still relevant. the change owner and any core reviewers for the repository where the change is proposed generally have access to set a workflow -1 on it | 18:48 |
| fungi | cathy_: the custom is to set workflow -1 if you know there's more work needed on a change before it's ready for general review | 18:49 |
| *** peristeri has joined #openstack-infra | 18:49 | |
| *** Hal has joined #openstack-infra | 18:49 | |
| *** hashar has joined #openstack-infra | 18:49 | |
| *** baoli_ has quit IRC | 18:49 | |
| *** Hal is now known as Guest66421 | 18:49 | |
| *** baoli has joined #openstack-infra | 18:50 | |
| *** _diana_ has quit IRC | 18:51 | |
| *** _diana_ has joined #openstack-infra | 18:52 | |
| sdake | hey folks | 18:53 |
| mordred | greghaynes: zomg. it did??? zomgzomg | 18:53 |
| mordred | jeblair: ^^ | 18:53 |
| mordred | jeblair: devstack works on the new dib nodes | 18:53 |
| sdake | I need someone with gerrit super powers to add me to https://review.openstack.org/#/admin/groups/970,members | 18:53 |
| sdake | it is unpopulated, and we need to populate it with the correct folks | 18:53 |
| fungi | sdake: that's updated now | 18:54 |
| greghaynes | mordred: yep, probably would be good to double check that ubuntu-trusty is what I think it is... | 18:54 |
| sdake | thansk fungi | 18:54 |
| fungi | sdake: you're welcome | 18:54 |
| sdake | badjones what is your email plz | 18:55 |
| mordred | greghaynes: it is | 18:55 |
| sdake | bradjones | 18:55 |
| *** Rockyg has joined #openstack-infra | 18:55 | |
| mordred | greghaynes: those are definitely the new dib minimal glean images | 18:55 |
| * mordred does a little dance | 18:55 | |
| greghaynes | w00t | 18:55 |
| bradjones | sdake: jones.bradley@me.com | 18:55 |
| bradjones | sdake: might make my new nick BADjones! | 18:56 |
| *** aysyd has quit IRC | 18:56 | |
| fungi | bradjones: i like it | 18:58 |
| *** DericHorn-HP has joined #openstack-infra | 18:58 | |
| fungi | bradjones: and less typing than basketballjones | 18:59 |
| *** _diana_ has quit IRC | 18:59 | |
| SpamapS | Another thought has just ocurred to me... | 18:59 |
| SpamapS | that we should probably talk about infra-cloud at the tokyo summit in a presentation format. | 18:59 |
| clarkb | hrm I just upgraded gertty to latest master and AttributeError: 'App' object has no attribute 'git_root' | 19:00 |
| fungi | mordred: greghaynes: unfortunately now i have to wait a while for another ubuntu-trusty node to build, since we're starved for quota at the moment | 19:00 |
| * SpamapS prepares an abstract | 19:00 | |
| *** aysyd has joined #openstack-infra | 19:00 | |
| clarkb | looks like yaml format changed /me updates config | 19:00 |
| mordred | SpamapS: ++ | 19:00 |
| mordred | SpamapS: I'd be in for some of that action | 19:00 |
| *** amitgandhinz has quit IRC | 19:01 | |
| fungi | looks like we have unhappy jenkins masters again. i'm going to gracefully cycle them a couple at a time | 19:01 |
| *** _diana_ has joined #openstack-infra | 19:01 | |
| openstackgerrit | Kyle Mestery proposed openstack-infra/irc-meetings: Update neutron-drivers team meeting times. https://review.openstack.org/198806 | 19:01 |
| clarkb | fungi: let me know if I can help | 19:01 |
| *** amitgandhinz has joined #openstack-infra | 19:01 | |
| *** signed8bit has joined #openstack-infra | 19:02 | |
| *** signed8b_ has joined #openstack-infra | 19:02 | |
| fungi | clarkb: eh, shouldn't take me but a second | 19:02 |
| *** Longgeek has joined #openstack-infra | 19:02 | |
| fungi | i believe that's why we have so many nodes in a delete state in nodepool | 19:02 |
| fungi | it's taking too long to hear back from some of the jenkins servers | 19:03 |
| clarkb | ya those deletes pile up i fjenkins is unable to process them | 19:03 |
| fungi | okay, jenkins01 is in prepare for shutdown | 19:03 |
| fungi | jenkins02 looks relatively healthy at the moment | 19:03 |
| fungi | jenkins03 seems fine as well | 19:04 |
| fungi | jenkins04 looks like it needs help | 19:04 |
| fungi | but it's not terrible yet | 19:04 |
| fungi | jenkins05 is in a bad way. putting it into prepare for shutdown now | 19:05 |
| fungi | jenkins06 is looking good | 19:05 |
| openstackgerrit | Paul Belanger proposed openstack-infra/system-config: WIP: Migrate to puppet-httpd module https://review.openstack.org/198819 | 19:06 |
| fungi | i think jenkins07 is really bad off. it's taking forever to even paint the initial browser window | 19:06 |
| *** e0ne has quit IRC | 19:06 | |
| *** HeOS has quit IRC | 19:07 | |
| * anteaya pictures a lethargic butler barely able to lift a brush from his reclined posture on a fainting couch | 19:07 | |
| *** maishsk_ has joined #openstack-infra | 19:07 | |
| *** maishsk has quit IRC | 19:08 | |
| *** maishsk_ is now known as maishsk | 19:08 | |
| *** cathy_ has quit IRC | 19:09 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/puppet-askbot: Migrate to puppet-httpd module https://review.openstack.org/198827 | 19:10 |
| anteaya | AJaeger_: did you miss my comment where I posted that jeblair likes to be aware of changes to devstack-gate? https://review.openstack.org/#/c/196085/ | 19:15 |
| *** flepied has quit IRC | 19:17 | |
| *** cindyo has joined #openstack-infra | 19:17 | |
| *** _diana_ has quit IRC | 19:18 | |
| *** reed has quit IRC | 19:18 | |
| *** reed has joined #openstack-infra | 19:18 | |
| *** hdd has quit IRC | 19:20 | |
| openstackgerrit | Merged openstack-infra/project-config: remove swift-dsvm-functional from blocking changes to many projects https://review.openstack.org/196085 | 19:21 |
| openstackgerrit | Andreas Jaeger proposed openstack-dev/specs-cookiecutter: Remove .coveragerc https://review.openstack.org/198857 | 19:22 |
| clarkb | fungi: did you see the constraints proposal job is now failing on something zuul-cloner related? I didn't get a chance to dig in just read a console log late last night | 19:22 |
| openstackgerrit | Andreas Jaeger proposed openstack-dev/specs-cookiecutter: Update requirements https://review.openstack.org/198859 | 19:23 |
| *** _diana_ has joined #openstack-infra | 19:23 | |
| openstackgerrit | Andreas Jaeger proposed openstack-dev/specs-cookiecutter: Cleanup .gitignore https://review.openstack.org/198860 | 19:24 |
| openstackgerrit | Andreas Jaeger proposed openstack-dev/specs-cookiecutter: Remove .mailmap https://review.openstack.org/198862 | 19:25 |
| fungi | clarkb: er, it shouldn't be... i switched it to use gerrit-branch-prep instead of zuul-git-prep | 19:25 |
| clarkb | fungi: hrm the log I saw had it ocmplaining about a missing zuul_ref | 19:26 |
| clarkb | maybe the job idnd't update yet? | 19:26 |
| *** e0ne has joined #openstack-infra | 19:26 | |
| fungi | clarkb: oh! the script itself also uses zuul-cloner, not just the builder macro | 19:26 |
| clarkb | https://jenkins.openstack.org/job/propose-requirements-constraints-master/11/console | 19:26 |
| fungi | clarkb: i think i just assumed everywhere that we have zuul-cloner installed on all our workers, when in actuality that's only true for nodepool workers | 19:27 |
| clarkb | oh though there is a -z test direclty that is failing? maybe its the script and not clonre? | 19:27 |
| fungi | no, i think the job will need to obtain and install zuul-cloner before using it, or do something else | 19:27 |
| *** Longgeek has quit IRC | 19:27 | |
| fungi | or we work out puppetry for getting zuul-cloner continuously deployed to proposal.slave.openstack.org | 19:28 |
| openstackgerrit | Andreas Jaeger proposed openstack-dev/specs-cookiecutter: Remove .testr.conf https://review.openstack.org/198866 | 19:28 |
| mordred | fungi: I vote on the puppetry | 19:28 |
| mordred | fungi: or something | 19:29 |
| clarkb | are we using tox to run the job? | 19:29 |
| *** linkedinyou has joined #openstack-infra | 19:29 | |
| clarkb | we could install zuul via that if so | 19:29 |
| fungi | clarkb: no, the script is run directly from the builder, like with the requirements update job | 19:29 |
| *** cdent has quit IRC | 19:29 | |
| *** heyongli has quit IRC | 19:30 | |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add stable ACLs for Neutron sub-projects https://review.openstack.org/198750 | 19:31 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/project-config: Add a create ACL for networking-bgpvpn https://review.openstack.org/198751 | 19:31 |
| *** heyongli has joined #openstack-infra | 19:32 | |
| *** dims_ has joined #openstack-infra | 19:36 | |
| mordred | I think on the proposal slave, we could just install the zuul software - there should be no pollution issues there | 19:36 |
| clarkb | ya | 19:37 |
| clarkb | I can write that change real quick | 19:37 |
| clarkb | or is it up and needs review? | 19:37 |
| jeblair | yes, i agree +puppet zuul-cloner for all slaves | 19:39 |
| jeblair | mordred, greghaynes: yaay! | 19:39 |
| *** Sukhdev has joined #openstack-infra | 19:40 | |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/system-config: Add a refstack.org server https://review.openstack.org/198869 | 19:40 |
| openstackgerrit | Merged openstack-dev/specs-cookiecutter: Improve CONTRIBUTING.rst for clarity https://review.openstack.org/183794 | 19:40 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/shade: Raise exception for nova egress secgroup rule https://review.openstack.org/198871 | 19:40 |
| jeblair | clarkb: er, i thought i fixed that, can you confirm your gertty sha and paste the traceback? | 19:41 |
| openstackgerrit | Merged openstack-dev/specs-cookiecutter: Replace ci.o.o links with docs.o.o/infra https://review.openstack.org/183231 | 19:41 |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/system-config: Add a refstack.org server https://review.openstack.org/198869 | 19:41 |
| fungi | hogepodge: https://review.openstack.org/198869 is a starting point for gluing the refstack puppet module to a server | 19:44 |
| sigmavirus24 | Is https://docs.openstack.org/developer/python-keystoneclient/authentication-plugins.html down for anyone else? (I'm getting a connection refused in firefox and chrome) | 19:44 |
| *** puranamr has quit IRC | 19:45 | |
| fungi | sigmavirus24: yep. that's hosted on rackspace cloudfiles | 19:45 |
| fungi | er, cloudsites | 19:45 |
| sigmavirus24 | new name is "cloudcities" | 19:46 |
| sigmavirus24 | so say we all =P | 19:46 |
| fungi | heh | 19:46 |
| *** soren has quit IRC | 19:46 | |
| jeblair | so say we all | 19:46 |
| *** _diana_ has quit IRC | 19:46 | |
| fungi | but yeah, i wonder if the cloudfiles incident going in on dfw/ord/iad is affecting cloudsites too | 19:46 |
| fungi | hrm, though the description of that incident looks like it shouldn't be related | 19:47 |
| *** e0ne has quit IRC | 19:47 | |
| fungi | [Cloud Sites] Scheduled Maintenance for DFW Environment Migration | 19:48 |
| mordred | heh | 19:48 |
| mordred | I guess that might be related | 19:48 |
| fungi | that's a ticket opened by phrenetic support for our openstack tenant | 19:48 |
| * sigmavirus24 sighs | 19:48 | |
| fungi | hrm, though that's scheduled for 60 days out from last tuesday when the ticket was opened | 19:49 |
| fungi | so also likely not related | 19:49 |
| *** _diana_ has joined #openstack-infra | 19:49 | |
| *** cindyo has quit IRC | 19:50 | |
| yolanda | pabelanger, sorry, i was out. I think that for cronjob, just a string with the cron times will be ok | 19:50 |
| clarkb | jeblair: sorry I moved db aside and started over | 19:51 |
| clarkb | but let me see if I hav ethe traceback | 19:51 |
| clarkb | jeblair: http://paste.openstack.org/show/349491/ | 19:53 |
| fungi | sigmavirus24: oh!!! http not https | 19:54 |
| sigmavirus24 | fungi: oh | 19:54 |
| *** esmute__ has quit IRC | 19:54 | |
| * sigmavirus24 sighs | 19:54 | |
| sigmavirus24 | HTTPS Everywhere, why do you have rules for those subdomains? | 19:54 |
| fungi | sigmavirus24: i only realized it when i started browsing through from the root of docs.openstack.org, saw it work to that same html file, and compared the urls | 19:54 |
| sigmavirus24 | Guess I should go fix those rules tonight or something | 19:55 |
| *** _diana_ has quit IRC | 19:55 | |
| sigmavirus24 | I'm welcome or something, right? | 19:55 |
| *** bswartz has quit IRC | 19:55 | |
| *** heyongli has quit IRC | 19:57 | |
| *** heyongli has joined #openstack-infra | 19:58 | |
| *** Guest20028 has joined #openstack-infra | 19:58 | |
| openstackgerrit | Merged openstack-infra/puppet-puppet: Add Gemfile and puppet 4 checks https://review.openstack.org/175231 | 19:59 |
| *** _diana_ has joined #openstack-infra | 20:00 | |
| *** whayutin_ has joined #openstack-infra | 20:00 | |
| *** jasondotstar has quit IRC | 20:00 | |
| fungi | sigmavirus24: you're always welcom | 20:00 |
| fungi | e | 20:00 |
| clarkb | mriedem: re 193835 why does d-g need to flip any bits? devstack should just disable that feature if using ceph | 20:01 |
| clarkb | mriedem: we shouldn't encode how to make things work into d-g that is devstack job | 20:01 |
| *** puranamr has joined #openstack-infra | 20:01 | |
| *** DericHorn-HP has quit IRC | 20:02 | |
| openstackgerrit | James E. Blair proposed stackforge/gertty: Fix get_repo call even more https://review.openstack.org/198875 | 20:03 |
| *** maishsk has quit IRC | 20:03 | |
| jeblair | clarkb: if you still have the db and can test that ^, that would be great | 20:03 |
| *** DericHorn-HP has joined #openstack-infra | 20:03 | |
| *** amrith is now known as _amrith_ | 20:04 | |
| *** maishsk has joined #openstack-infra | 20:04 | |
| jeblair | it's probably about time for migration tests | 20:05 |
| nibalizer | fix it more with the fixes! | 20:06 |
| jeblair | if it's worth fixing once it's worth fixing again | 20:06 |
| *** _diana_ has quit IRC | 20:06 | |
| clarkb | jeblair: ya I moved the db aside will put that in place and give it a go once current review is done | 20:07 |
| *** heyongli has quit IRC | 20:07 | |
| *** _diana_ has joined #openstack-infra | 20:07 | |
| *** yamahata has quit IRC | 20:08 | |
| *** heyongli has joined #openstack-infra | 20:08 | |
| *** abitha has quit IRC | 20:08 | |
| *** yamahata has joined #openstack-infra | 20:08 | |
| *** _diana_ has quit IRC | 20:08 | |
| *** esmute__ has joined #openstack-infra | 20:09 | |
| *** abitha has joined #openstack-infra | 20:10 | |
| *** abitha has quit IRC | 20:11 | |
| *** flepied has joined #openstack-infra | 20:11 | |
| clarkb | for changes like 190601, every project we add there extends the time it taks to set up the workspace and increases the chances that a git op will fail | 20:11 |
| clarkb | do we maybe want to reduce PROJECTS to a base list then have features.yaml update it as necessary? or maybe have features.yaml define it completely? | 20:12 |
| clarkb | thinking that using a minimal list will make everything happier | 20:12 |
| clarkb | greghaynes: any progress on the multinode job? | 20:12 |
| clarkb | greghaynes: we still need a glean release? | 20:12 |
| greghaynes | clarkb: yes, release still neeeded. The patch mordred wanted to land has landed. | 20:12 |
| *** abitha has joined #openstack-infra | 20:13 | |
| mordred | oh. piddle. on it | 20:13 |
| greghaynes | and then image building time | 20:13 |
| mordred | greghaynes: I think 1.1.1 - all of these are bugfixes, not interface additions, yeah? | 20:14 |
| greghaynes | the set hostname differently is kind of a interface change | 20:14 |
| greghaynes | our output is different | 20:15 |
| openstackgerrit | Stephanie Miller proposed openstack-infra/project-config: Add tests for bifrost https://review.openstack.org/190327 | 20:15 |
| fungi | mordred: greghaynes: heh, great timing. i just replied to armax's message on the multi-node grenade thread saying we needed a new glean release ;) | 20:15 |
| clarkb | uhm multinode grenade? | 20:16 |
| clarkb | can we get working multinode neutron first? | 20:16 |
| clarkb | grumble grumble | 20:16 |
| *** fallenpe_ has joined #openstack-infra | 20:16 | |
| *** soren has joined #openstack-infra | 20:16 | |
| mordred | greghaynes: 1.2.0 cut | 20:16 |
| *** maishsk has quit IRC | 20:16 | |
| greghaynes | mordred: awesome | 20:16 |
| *** dims_ has quit IRC | 20:17 | |
| mriedem | clarkb: i suppose i can throw it in here http://git.openstack.org/cgit/openstack-dev/devstack/tree/extras.d/60-ceph.sh | 20:17 |
| *** whayutin_ has quit IRC | 20:17 | |
| mriedem | clarkb: i was following what was already in d-g, but i suppose those are wrong too | 20:17 |
| clarkb | mriedem: I left a -1 on the chang ebut typically I would expect logic to make ceph + cinder + nova work to be in devstack | 20:17 |
| *** fallenpegasus has quit IRC | 20:18 | |
| mordred | me too | 20:18 |
| mordred | devstack-gate should only really have logic about how to request the right thing from devstack | 20:18 |
| SpamapS | mordred: want me to list you as co-presenter? | 20:19 |
| mordred | SpamapS: sure! | 20:19 |
| *** heyongli has quit IRC | 20:19 | |
| SpamapS | mordred: cool, that will help as I'm still pretty infra-stupid. :) | 20:19 |
| *** Guest20028 has quit IRC | 20:20 | |
| fungi | mordred: greghaynes: though in other good news, my experimental nova unit test job is working on ubuntu-trusty nodes (this one ran in rax-dfw) http://logs.openstack.org/44/164444/1/experimental/experimental-nova-dsvm-python27/fb1ecf2/console.html | 20:20 |
| armax | fungi: thanks for the headsup | 20:20 |
| greghaynes | fungi: nice | 20:20 |
| clarkb | jeblair: why is 189314 WIP? anything I can do to help with moving that along? removing the d-g reexec would be awesome | 20:21 |
| *** ddieterly has joined #openstack-infra | 20:21 | |
| fungi | armax: yeah, mordred just tagged glean 1.2.0 so when our images update tomorrow the new multinode images should hopefully be back to working again | 20:21 |
| *** HeOS has joined #openstack-infra | 20:21 | |
| clarkb | armax: have we made any progress getting neutron working on multinode? | 20:22 |
| clarkb | armax: the hostname issues don't help but curious where that work got too | 20:22 |
| *** koolhead17 has quit IRC | 20:22 | |
| armax | clarkb: I left it in the hands of Swami, when I checked it last the job was passing fairly regularly, but then a breaking change snuck in Tempest, and that caused the failure rate to spike at 100% | 20:24 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/infra-specs: JavaScript in Infra https://review.openstack.org/198884 | 20:24 |
| jeblair | clarkb: i think there was something happening with d-g at the time; i'll check back on that soon and unwip it | 20:24 |
| *** Longgeek has joined #openstack-infra | 20:24 | |
| jeblair | gotta run, biab. | 20:25 |
| Swami | armax: clarkb: hi | 20:25 |
| krotscheck | finally | 20:25 |
| clarkb | hello | 20:25 |
| *** heyongli has joined #openstack-infra | 20:25 | |
| Swami | armax: yes I am working on a fix, I found out the problem. there is a first patch that addresses one of the issue. | 20:25 |
| *** packet has quit IRC | 20:25 | |
| Swami | armax: I will be pushing in another patch that would completely remove the IPv6 related failures. | 20:26 |
| * mordred is a fan of IPv6 | 20:26 | |
| Swami | armax: the fix that I will be proposing is for stabilizing the single node, then I will get back to the multinode. | 20:26 |
| anteaya | Swami: have you help on dvr yet? | 20:27 |
| *** annegentle has quit IRC | 20:27 | |
| *** peristeri has quit IRC | 20:27 | |
| Swami | anteaya: hi | 20:27 |
| anteaya | Swami: when we discussed the dvr situation at the mid-cycle it appeared you were the only one working on it | 20:27 |
| Swami | anteaya: Yes I am one that is working on it right now. | 20:27 |
| anteaya | which seems a precarious situation given the importance neutron is giving dvr | 20:28 |
| anteaya | Swami: well if dvr is going to be adequately maintained more attention needs to be given to it | 20:28 |
| anteaya | Swami: you need more help | 20:28 |
| anteaya | and that includes getting testing working | 20:28 |
| Swami | anteaya:Yes, we wanted to make it voting at the first job. | 20:28 |
| anteaya | who is we? | 20:28 |
| Swami | anteaya: that's what I am working on. | 20:28 |
| Swami | anteaya: we is the DVR team in Neutron | 20:29 |
| anteaya | are you aware of any efforts to gather more help for the work | 20:29 |
| anteaya | it sounds like the DVR team in neutron consists of yourself | 20:29 |
| anteaya | unless I mis-understood your prior comment | 20:29 |
| Swami | anteaya: we do have a small team, but they are just beginners. | 20:29 |
| anteaya | are they in the neutron channel? | 20:30 |
| Swami | anteaya:yes they are listening in the neutron channel. | 20:30 |
| clarkb | fungi: found a fun situation with git review/gerrit | 20:30 |
| anteaya | let's see if we can get them partcipating in the neutron channel perhaps | 20:30 |
| *** Guest66421 has quit IRC | 20:30 | |
| Swami | anteaya: right now the only one that is blocking the DVR to become voting job is the bug that we hit with the IPv6 multiple subnet. | 20:30 |
| fungi | clarkb: oh? | 20:31 |
| *** jsavak has quit IRC | 20:31 | |
| clarkb | fungi: https://review.openstack.org/#/c/168146 can't merge because it is based on an old patchset of its parent chnage. If I rebase it onto master (the parent has merged) git review wont let me push it due to a lck of changes | 20:31 |
| anteaya | Swami: I'm saying you don't have enough help | 20:31 |
| Swami | anteaya: I do have a handle on it right now, and will be pushing a patch today. | 20:31 |
| *** jsavak has joined #openstack-infra | 20:31 | |
| fungi | clarkb: tried the rebase button in gerrit?> | 20:31 |
| Swami | anteaya: that's right. | 20:31 |
| anteaya | Swami: you need more help and you need kyle to be aware of that | 20:31 |
| clarkb | fungi: have not, but I did get a local conflict. Will see how rebaes button does | 20:31 |
| fungi | oh, if there's a conflict then why is git review thinking nothing's changed once you resolve it? | 20:32 |
| lifeless | sdague: I'll be happy to give you some more pointers on pbr tests | 20:32 |
| clarkb | ya failed due to a conflict | 20:32 |
| Swami | anteaya: Yes I will escalate it to Kyle if I need more help. | 20:32 |
| lifeless | sdague: once I get C off to kindy | 20:32 |
| Swami | anteaya: thanks for asking this. | 20:32 |
| clarkb | fungi: maybe I did the conflict resultion the wrong way and set it the old way and not the new way | 20:32 |
| fungi | clarkb: trying myself now. you've got me curious | 20:32 |
| anteaya | Swami: don't escalate if, make a point of keeping him aware of your progress | 20:32 |
| clarkb | time to go read diffs | 20:32 |
| Swami | anteaya: Last couple of weeks I was trying to triage this critical bug, which took most of my time. | 20:32 |
| anteaya | communication only when you are in trouble is not helpful | 20:32 |
| *** heyongli has quit IRC | 20:33 | |
| *** yamamoto has joined #openstack-infra | 20:33 | |
| lifeless | sdague: in the meantime https://review.openstack.org/#/c/198580/ and https://review.openstack.org/#/c/198579/ are backports on the steps to unbreaking stable requirements check jobs | 20:33 |
| sdague | lifeless: so, mordred unstuck me on my first issue, now I'm just trying to figure out how to communicate with a subprocess without it blocking on my stdout read. But it's end of day here, so I need to run away. | 20:33 |
| Swami | anteaya: Yes I did update him through the launchpad that I am working on fixing this bug. | 20:33 |
| clarkb | fungi: nope pretty sure I resolved it correctly | 20:33 |
| anteaya | Swami: keep kyle informed of your situation | 20:33 |
| Swami | anteaya: hope this help.s | 20:33 |
| sdague | lifeless: ok, I'll look quick | 20:33 |
| Swami | anteaya:sure | 20:33 |
| Swami | anteaya: will do | 20:33 |
| anteaya | honestly for my confience, no, not yet | 20:33 |
| lifeless | sdague: thanks. They'd depended on by the backports in openstack/requirements for juno and kilo | 20:33 |
| mordred | lifeless, sdague: I helped! | 20:33 |
| openstackgerrit | Monty Taylor proposed openstack-infra/project-config: Retire devstack-trusty in favor of ubuntu-trusty https://review.openstack.org/198886 | 20:34 |
| *** heyongli has joined #openstack-infra | 20:34 | |
| sdague | lifeless: are they backports of known changes? | 20:34 |
| sdague | ah, yeh | 20:34 |
| mordred | greghaynes, clarkb, fungi, jeblair, pleia2: ^^ so - sake of argument - I think that's the change needed to jobs to move to ubuntu-trusty - but I'm not sure how I think we should bleed the quota in nodepool - thoughts? | 20:34 |
| sdague | lifeless: in future can you do the cherry-pick -x thing so that it is a little more clear that it's a straight backport | 20:34 |
| lifeless | sdague: I did do that | 20:35 |
| clarkb | mordred: you will need to set devstack-trusty min ready to 0 then after a day or so remove the config for it | 20:35 |
| anteaya | clarkb: does the multinode testing setup rely on dvr to function? | 20:35 |
| greghaynes | mordred: for starters - I think uploads need to be working reliably before we start | 20:35 |
| lifeless | sdague: in devstack, though I had to fold three patches together, so git may have decided to throw away that metadata | 20:35 |
| clarkb | anteaya: for neutron it does, I pushed using dvr in multinode since we can test non multinode neutron failry effectively on single node | 20:35 |
| sdague | lifeless: hmmm normally it gives the original git hash | 20:36 |
| mordred | clarkb: right - but we're going to have a moment where we're going to have a supply/demand mismatch | 20:36 |
| clarkb | anteaya: er we can test non dvr effectively | 20:36 |
| *** austin81 has joined #openstack-infra | 20:36 | |
| mordred | clarkb: which I suppose is probably ok | 20:36 |
| sdague | lifeless: ok, that's fair. | 20:36 |
| lifeless | sdague: and in requirements its a fold together of hundreds of patches, so really not feasible | 20:36 |
| clarkb | anteaya: same reason we use nova nets multihost feature on multinode tests | 20:36 |
| *** ZZelle has joined #openstack-infra | 20:36 | |
| lifeless | sdague: oh! I missed the -x on my cherrypick command. Will add that in next time. | 20:36 |
| pleia2 | mordred: in theory I'm in support of our new dib overloads | 20:36 |
| anteaya | clarkb: okay, just thinking that in the setup for multinode for grenade purposes, it might be worthwhile to consider not using dvr for that | 20:36 |
| sdague | lifeless: yeh, just that flag, thanks | 20:37 |
| clarkb | mordred: if you are really worreid about that you can do devstack-trusty || ubuntu-trusty and make the functions selct either node | 20:37 |
| clarkb | anteaya: no I completely disagree | 20:37 |
| anteaya | okay | 20:37 |
| mordred | clarkb: oh - that's a good point - do that, then change the quota, then remove devstack-trusty | 20:37 |
| sdague | +2 on both of those patches. If no one else reviews them by morning, I'll kick them in. But I need to go help with dinner now | 20:37 |
| *** simondodsley has joined #openstack-infra | 20:37 | |
| clarkb | anteaya: because we can effectively test the other config without multinode and because dvr is the way we are telling people to deploy | 20:37 |
| lifeless | sdague: ciao, thanks. | 20:37 |
| anteaya | clarkb: okay well for the grenade tests sdague wants multinode | 20:38 |
| anteaya | which relys on dvr | 20:38 |
| *** ZZelle has quit IRC | 20:38 | |
| clarkb | anteaya: yes and I am asking grenade to wait | 20:38 |
| anteaya | which is broken | 20:38 |
| clarkb | I need to write an email but I think we are getting way ahead of ourselves if these jobs can be broken for am onth without anyone noticing | 20:38 |
| openstackgerrit | Merged openstack-infra/project-config: Do not call sys.exit inside ZanataUtils https://review.openstack.org/198218 | 20:38 |
| clarkb | and then trying to add a completely new set of testing on top of it | 20:38 |
| anteaya | well this is creating a heck of a backlog for the neutron partial upgrade grenade scenario | 20:38 |
| lifeless | clarkb: ++ | 20:38 |
| anteaya | clarkb: oh I agree | 20:38 |
| simondodsley | having trouble with gerrit accepting my SSH key - any advice gladly received. New user and this is the first time I've tried to connect. | 20:39 |
| anteaya | clarkb: and I think the reason, at least to my eyes, is that a key function is maintained by a team of one currently | 20:39 |
| clarkb | simondodsley: can you paste the full command and error to http://paste.openstack.org? | 20:39 |
| clarkb | anteaya: ya so maybe neutron needs to reevaluate their suggestion to use dvr | 20:39 |
| anteaya | clarkb: agreed | 20:39 |
| *** cindyo has joined #openstack-infra | 20:40 | |
| simondodsley | trying to debug just the SSH connection. Paste #349564 | 20:40 |
| *** heyongli has quit IRC | 20:40 | |
| clarkb | simondodsley: you are running ssh as root, is the public key you gave gerrit belonging to that root user? | 20:41 |
| simondodsley | yes | 20:41 |
| clarkb | oh I know | 20:42 |
| clarkb | you are telling gerrit you want to ssh as the root user | 20:42 |
| clarkb | simondodsley: set your gerrit username when sshing eg `ssh -p 29418 user@review.openstack.org` | 20:42 |
| *** jsavak has quit IRC | 20:43 | |
| *** jsavak has joined #openstack-infra | 20:43 | |
| simondodsley | clarkb: OK _ that works - I see that now, however git review -s is still asking for a password which I assume means its still not accepting the key | 20:43 |
| clarkb | simondodsley: can you past ethe git review -s full command and output too? maybe add a -v | 20:44 |
| *** heyongli has joined #openstack-infra | 20:44 | |
| *** notnownikki has quit IRC | 20:44 | |
| *** bradjones is now known as bradjones|away | 20:44 | |
| simondodsley | clarkb: #349601 | 20:45 |
| *** ZZelle has joined #openstack-infra | 20:45 | |
| clarkb | simondodsley: fwiw its easier if you provide the full http url | 20:45 |
| *** Rockyg has quit IRC | 20:45 | |
| simondodsley | sorry - http://paste.openstack.org/show/349601/ | 20:45 |
| *** ildikov has joined #openstack-infra | 20:45 | |
| *** RichardRaseley has joined #openstack-infra | 20:46 | |
| clarkb | simondodsley: it is asking for your password because you have instructed it to use https not ssh | 20:46 |
| simondodsley | clarkb: ok - so how do i tell it not to use 443 as the port | 20:46 |
| clarkb | simondodsley: remove whatever config you have used telling it to use https. From that output I am gessing /etc/git-review/git-review.conf | 20:47 |
| *** peristeri has joined #openstack-infra | 20:48 | |
| *** ari has quit IRC | 20:49 | |
| simondodsley | clarkb: nothing in that file and when I delete it I still get the same error | 20:49 |
| *** gordc has quit IRC | 20:49 | |
| *** zns has joined #openstack-infra | 20:49 | |
| *** vnod has joined #openstack-infra | 20:50 | |
| sc68cal | clarkb: just catching up | 20:50 |
| *** Longgeek has quit IRC | 20:50 | |
| sc68cal | clarkb: when you say neutron says to use DVR, do you mean docs or neutron developers | 20:50 |
| clarkb | simondodsley: can you check the .gitreview file? cinder should definitely default to ssh, but also check ~/.gitconfig and maybe it already has a gerrit remote `git remote -v` if there is already a gerrit remote try delting it and starting over | 20:51 |
| anteaya | sc68cal: http://eavesdrop.openstack.org/irclogs/%23openstack-infra/%23openstack-infra.2015-07-06.log.html#t2015-07-06T20:37:46 | 20:51 |
| clarkb | sc68cal: neutron devs have said "Use dvr to get this feature from nova net that was otherwise missing" | 20:51 |
| *** thedodd has quit IRC | 20:51 | |
| clarkb | sc68cal: nova net ha long had "multihost" routing which made each compute node a router resuling int redundancy | 20:52 |
| clarkb | sc68cal: my understanding of neutron is that the prescribed way to get that functionality is via dvr | 20:52 |
| clarkb | so with multinode testing I am far more intrested in testing that scenario than the single router setup which we already test with every other job | 20:53 |
| *** hichihara has joined #openstack-infra | 20:53 | |
| sc68cal | agree | 20:53 |
| clarkb | simondodsley: but defaults for our repos and git-review is to use ssh so something has to be overriding it to use https which would be a local thing on your end | 20:53 |
| openstackgerrit | Merged openstack-infra/shade: Make sure we are returning floating IPs in current domain https://review.openstack.org/196899 | 20:53 |
| clarkb | sc68cal: and I think the failures that we have run into for both nova and neutron with >1 router show that there is a tremendous amount of value in testing this. Particulary on the nova side at least since I am told no one runs nova net without multihost | 20:54 |
| *** fitoduarte has joined #openstack-infra | 20:56 | |
| sc68cal | clarkb: agreed - I think Sam-I-Am in -docs has a lot of bones to pick with DVR | 20:58 |
| openstackgerrit | Russell Bryant proposed openstack-infra/project-config: Add check-grenade-dsvm-partial-ncpu-neutron https://review.openstack.org/189426 | 20:58 |
| openstackgerrit | Russell Bryant proposed openstack-infra/project-config: Add check-grenade-dsvm-partial-ncpu-neutron-dvr https://review.openstack.org/189727 | 20:58 |
| sc68cal | related to how well it works, reliability, etc. | 20:58 |
| anteaya | my concern is that for something so important only one person is working on it | 20:58 |
| *** adduarte has joined #openstack-infra | 20:58 | |
| *** amotoki has joined #openstack-infra | 20:59 | |
| *** soren has quit IRC | 20:59 | |
| mordred | clarkb, sc68cal: of course, I'll just keep saying "The equiv functionality in neutron is not dvr, it's provider-network + linuxbridge" | 21:01 |
| *** eharney has quit IRC | 21:01 | |
| sc68cal | AMEN! | 21:01 |
| *** john-davidge has joined #openstack-infra | 21:01 | |
| *** fitoduarte has quit IRC | 21:01 | |
| *** ociuhandu has joined #openstack-infra | 21:01 | |
| * sc68cal does a preach it dance | 21:01 | |
| clarkb | mordred: sc68cal does that actually give you multiple routers? | 21:01 |
| anteaya | so can we have multinode use that? | 21:01 |
| clarkb | I think what that does is offload the problem which isn't the same iether | 21:02 |
| clarkb | it says "you run X routers because neutron can't figure out how to do it for you" | 21:02 |
| anteaya | because taht would get multinode possibly moving again? maybe? | 21:02 |
| mordred | clarkb: no - it direct routes | 21:02 |
| sc68cal | clarkb: sadly no, it just pushes L3 up into physical equipment | 21:02 |
| clarkb | mordred: ^ | 21:02 |
| mordred | yah | 21:02 |
| clarkb | so ya | 21:02 |
| clarkb | thats still broken | 21:02 |
| mordred | it's not broken | 21:02 |
| mordred | it's EXACTLY what you want | 21:02 |
| clarkb | it is as a replacement | 21:02 |
| clarkb | mordred: no its not | 21:02 |
| mordred | ok. you and I want different things then | 21:02 |
| mordred | it's exactly what I want | 21:02 |
| clarkb | mordred: I want what nova net does | 21:02 |
| mordred | I don't | 21:03 |
| * anteaya wants a grenade story for neutron | 21:03 | |
| clarkb | ok thats fine, the nice thing about what nova net does is your spof is what is already the spof a single hypervisor | 21:03 |
| *** john-davidge has left #openstack-infra | 21:03 | |
| mordred | I don't want any software routers if I don't need them | 21:03 |
| sc68cal | What are we trying to test with the grenade job? | 21:03 |
| mordred | I want my packets to go on to the copper unless I have requested something fancy | 21:03 |
| sc68cal | just an upgrade, with multiple nodes? | 21:03 |
| anteaya | sc68cal: a partial upgrade | 21:03 |
| clarkb | mordred: you can also do that with nova net | 21:03 |
| anteaya | sc68cal: hence the need for multinode, one node upgrades one does not | 21:04 |
| clarkb | mordred: the nice thing about what nova net has done is provide a working out of the box config that is redundant and simple to use | 21:04 |
| mordred | clarkb: I do not care about nova net | 21:04 |
| clarkb | mordred: no but you should care about use cases | 21:04 |
| clarkb | "redundant easy to use neutron setup" | 21:04 |
| clarkb | DNE | 21:04 |
| sc68cal | clarkb: do you need to have DVR in the first iteration of the grenade job | 21:04 |
| clarkb | sc68cal: yes because that is what we are running | 21:04 |
| clarkb | IMO there is 0 value to multinode neutron + grenade without dvr | 21:05 |
| *** DericHorn-HP has quit IRC | 21:05 | |
| clarkb | we already test that | 21:05 |
| *** rfolco has quit IRC | 21:05 | |
| sc68cal | clarkb: who is "we" | 21:05 |
| clarkb | sc68cal: openstack | 21:05 |
| *** annegentle has joined #openstack-infra | 21:05 | |
| clarkb | sc68cal: there is a 2 node job that sets up neutron with dvr and runs tempest against it | 21:05 |
| simondodsley | clarkb: thank you - I have it working now | 21:05 |
| sc68cal | clarkb: My thinking is, if we can get provider networking as the network layout, to shake out the bugs on the grenade side, and then transition to using DVR when that job stabilizes, at least we don't have a blocker | 21:06 |
| anteaya | sc68cal: context: http://lists.openstack.org/pipermail/openstack-dev/2015-June/067156.html | 21:06 |
| clarkb | sc68cal: but that odens't add anything | 21:06 |
| clarkb | sc68cal: we alreay test neutron with grenade in a single router setup | 21:06 |
| *** simondodsley has quit IRC | 21:06 | |
| clarkb | we don't need a new job on a different test env to test that | 21:06 |
| mordred | clarkb: it DOES add something | 21:07 |
| mordred | clarkb: it tests provider network | 21:07 |
| mordred | which is the simple version | 21:07 |
| clarkb | mordred: well thats just a new variable thorwn out there right now and wasn't part of the previous discussion | 21:07 |
| mordred | and SHOULD be what the docs point you at for the "simple neutron install" | 21:07 |
| clarkb | if you want provider networking testing start on a single node | 21:07 |
| *** thorst has quit IRC | 21:07 | |
| clarkb | Don't start on multinode | 21:07 |
| mordred | it's also what we're going to run in infra-cloud, so it would be neat if it's tested | 21:07 |
| *** esmute__ has quit IRC | 21:07 | |
| clarkb | on multinode its just noise until tested single node | 21:08 |
| mordred | sure | 21:08 |
| mordred | but I believe we could make progress towardsa multi-node provider network config without blocking on dvr | 21:08 |
| mordred | in that we can make progress on single-node provider | 21:08 |
| mordred | then multi-node provider | 21:08 |
| clarkb | mordred: so its not necessarily the simple mode because it assumes you have control of your router | 21:08 |
| mordred | and none of that is dependent on dvr | 21:08 |
| anteaya | "without blocking on dvr" are my golden words | 21:09 |
| *** zns has quit IRC | 21:09 | |
| mordred | clarkb: to accomplish what purpose? | 21:09 |
| sc68cal | ++ - let's not block on dvr - let's at least solve as much of the multi node upgrade story that doesn't depend on dvr | 21:09 |
| clarkb | mordred: to test it to start | 21:09 |
| clarkb | mordred: We will have to ste up software routers to make that work on multinode | 21:09 |
| sc68cal | hopefully re-using as much as possible when we drop DVR int | 21:09 |
| *** dims_ has joined #openstack-infra | 21:09 | |
| clarkb | mordred: because we do not control our routers | 21:09 |
| mordred | clarkb: no - I mean what do you need to do to the theoretical router other than have a routable network? | 21:09 |
| mordred | "here's a network. awesome" | 21:10 |
| clarkb | mordred: set up routes for networks | 21:10 |
| mordred | what networks? | 21:10 |
| clarkb | mordred: the networks that VMs live on | 21:10 |
| mordred | you're skipping ahead to something | 21:10 |
| mordred | you're describing multiple networks | 21:10 |
| clarkb | mordred: I am skipping ahead to "Use the VM" | 21:10 |
| mordred | I am not | 21:10 |
| mordred | I am too | 21:10 |
| clarkb | mordred: for example in the multinode setup | 21:10 |
| mordred | I'm describing an ACTUAL simple install | 21:10 |
| clarkb | mordred: tempest has to ssh into the VMs | 21:10 |
| mordred | yes | 21:10 |
| mordred | on the one network | 21:10 |
| clarkb | you can't do that without control of the router | 21:10 |
| mordred | that preexists | 21:10 |
| clarkb | mordred: NO not on the same network | 21:10 |
| mordred | sure you can | 21:10 |
| mordred | yes | 21:10 |
| sc68cal | clarkb: you can if it's a flat network | 21:10 |
| clarkb | completely not on the same network | 21:10 |
| mordred | on the same network | 21:10 |
| mordred | flat network | 21:10 |
| mordred | flat network | 21:10 |
| clarkb | NO because rax | 21:11 |
| mordred | screw rax | 21:11 |
| clarkb | rax gives you a rax IP on each test node | 21:11 |
| clarkb | VMs live on 172.24.5.0/24 | 21:11 |
| clarkb | you cannot ssh between those addrs without a router you control | 21:11 |
| mordred | yes... wait - you're talking about a COMPLETELY different thing | 21:11 |
| clarkb | this is also true on hpcloud unless we go router in hpcloud per test env | 21:11 |
| mordred | yes. WE need a software router in INFRA to provide something that approximates an actual real world environment | 21:11 |
| mordred | but that is not a neutron software router | 21:11 |
| mordred | that we're testing | 21:12 |
| clarkb | mordred: and anyone else in a real world env that is like ours | 21:12 |
| mordred | that is a fake real router | 21:12 |
| mordred | nobody is deploying a cloud in an env where they ahve one ip | 21:12 |
| clarkb | mordred: right my point is that calling this the simple neutron setup is particularly troublesome beause of ^ | 21:12 |
| clarkb | if we can't even use it in a simple way then its not simple | 21:12 |
| mordred | our env is pathological | 21:12 |
| mordred | you can't use nova-net that way eitehr | 21:12 |
| clarkb | mordred: anyone with a laptop is | 21:12 |
| clarkb | mordred: yes you can | 21:12 |
| mordred | you ALSO have to set up a software router for nova-net | 21:12 |
| clarkb | mordred: because nova net sets the router up for you so you don't have to think about it | 21:12 |
| clarkb | mordred: no nova net sets it up | 21:13 |
| clarkb | I don't do anything | 21:13 |
| mordred | no - you had to set up your own GRE tunnel between tehe nodes to create a network in which nova-net could do its thing | 21:13 |
| mordred | no? | 21:13 |
| *** zns has joined #openstack-infra | 21:13 | |
| clarkb | mordred: yes thats all l2 and has nothing to do with routing | 21:13 |
| mordred | I mean, I seem to remember you doing a bunch of crazy networking for like 6 months | 21:13 |
| clarkb | yes but none of it is routing... | 21:13 |
| *** peristeri has quit IRC | 21:13 | |
| mordred | I believe we are talking about massively different things in a non-productive way | 21:14 |
| *** dims_ has quit IRC | 21:14 | |
| clarkb | I am saying that your "simple" is completely not simple | 21:14 |
| clarkb | so lets no pretend | 21:14 |
| clarkb | I do agree that in the infra cloud env it is the sane way to deploy | 21:14 |
| mordred | I'm saying it's every bit as simple in a REAL WORLD case as nova-net | 21:14 |
| mordred | it's not simple in teh fake environment of the gate | 21:14 |
| mordred | becasue it's a pathological environment | 21:15 |
| mordred | and yes, that is hard | 21:15 |
| clarkb | mordred: and thats teh diff with nova net | 21:15 |
| mordred | and we should not expect it to be simple | 21:15 |
| clarkb | nova net is relatively simple even in the crazy test env | 21:15 |
| mordred | but who cares? | 21:15 |
| mordred | ti's not even close to a real env | 21:15 |
| mordred | like, not even close to being close | 21:15 |
| mordred | a cloud ona laptop is a toy that devs use | 21:15 |
| clarkb | yes it is i mena doesn't cern still use it? | 21:15 |
| clarkb | and for precisely this reason | 21:15 |
| mordred | that is apples and oranges | 21:16 |
| clarkb | wat its a real env | 21:16 |
| SpamapS | cern still uses it because the upgrade path is torturous IIRC | 21:16 |
| clarkb | in the real world users use nova netowrk for this functionality | 21:16 |
| mordred | neutron provider es | 21:16 |
| ctracey | can anyone point me at the official pypi mirrors? | 21:16 |
| *** dims has joined #openstack-infra | 21:16 | |
| mordred | yes. what SpamapS said | 21:16 |
| mordred | ctracey: there are none | 21:16 |
| SpamapS | but there's a plan. | 21:16 |
| clarkb | mordred: SpamapS yes beacuse this functionality does not exist | 21:16 |
| clarkb | so you can't just upgrade | 21:16 |
| *** esmute__ has joined #openstack-infra | 21:16 | |
| mordred | clarkb: no - it DOES exist | 21:16 |
| ctracey | there used to be though, correct? | 21:16 |
| clarkb | mordred: provider networks is not where it exists | 21:16 |
| mordred | ctracey: yes. we do not run mirrors for not-the-gate anymore | 21:16 |
| clarkb | mordred: provider networks is pawning the respponsibility off on the underlying infra | 21:16 |
| clarkb | which is fine | 21:16 |
| clarkb | but its completely different | 21:17 |
| mordred | clarkb: I disagree. I believe the problem is taht we've been trying to provide an upgrade path from nova-net to full-sdn-neutron | 21:17 |
| ctracey | mordred: thanks. that is what I thought, but was not sure. | 21:17 |
| mordred | which is insane | 21:17 |
| clarkb | mordred: and nova net to full sdn is also different and also part of the problem | 21:17 |
| *** aysyd has quit IRC | 21:17 | |
| clarkb | there is no multihost equivalent that works | 21:17 |
| clarkb | there are alternatives that provide similar functonality | 21:17 |
| clarkb | at variying levels of complicatedness and working | 21:18 |
| mordred | yes. that is correct. there is nothing in neutron that is architected exactly like the thing in nova, because nova and neutron are different | 21:18 |
| mordred | there is, however, a thing that provides the use case | 21:18 |
| clarkb | no the use case is different | 21:18 |
| mordred | which is "I want to just route my vm traffic to my existing network kthxbai" | 21:18 |
| clarkb | mordred: the use case is run the router for me | 21:18 |
| clarkb | mordred: because that makes it simple | 21:18 |
| clarkb | and that is where dvr come sin | 21:19 |
| mordred | clarkb: wow. I just ferverently disagree with yo | 21:19 |
| mordred | simple is no software routers | 21:19 |
| clarkb | DVR i sthe closest thing aiui to that ues case | 21:19 |
| mordred | software routers == more complex | 21:19 |
| anteaya | here is my problem | 21:19 |
| anteaya | dvr is currently broken and has one person working on it | 21:19 |
| *** ldnunes has quit IRC | 21:19 | |
| mordred | routers that sit on my physical network and do their job work and work now and exist just about everywhere | 21:19 |
| anteaya | if we require dvr for multinode for grenade | 21:19 |
| mordred | I don't need a software router to get packets from my VM to the internet | 21:19 |
| *** hashar has quit IRC | 21:19 | |
| anteaya | I can't see us making any progress in this direction this release | 21:19 |
| mordred | that is the simplest case, IMO | 21:19 |
| clarkb | mordred: its typically a huge ask to get things configured on routers for you | 21:19 |
| clarkb | mordred: which hsa also been a problem for infra cloud right? | 21:20 |
| mordred | clarkb: nope | 21:20 |
| clarkb | mordred: oh I thought one of the two envs needed routes set up for ipv6 | 21:20 |
| mordred | nope | 21:20 |
| mordred | we did get a new IP block | 21:20 |
| mordred | so they're configuring the gear to advertise that on our VLAN | 21:20 |
| mordred | and they're deleting a couple of vlans we don't need | 21:20 |
| clarkb | anyways my point still stands I don't think dvr vs provider network has anything to do with the multinode discussion until provider network has a single node test | 21:21 |
| clarkb | mordred: right thats taken how long? | 21:21 |
| mordred | clarkb: I agree- provider network needs a single node test | 21:21 |
| mordred | clarkb: how would a software router help me? | 21:21 |
| sc68cal | I think we actually do test provider networking in temepst, but I don't remember if it's just an API test | 21:21 |
| clarkb | mordred: you could use an arbirtary range and route through the existing routes for your hypervisors | 21:21 |
| clarkb | mordred: because your hypervisor is now a router | 21:22 |
| mordred | clarkb: right. but that's not the thing they're changing | 21:22 |
| clarkb | sc68cal: there is no grenade test iirc | 21:22 |
| mordred | clarkb: they're changing my ENTIRE NETWORK | 21:22 |
| mordred | clarkb: like, from a 16.x.x.x/y something to a 15.x.x.x/z | 21:22 |
| mordred | completely different inbound networking range | 21:22 |
| *** esmute__ has quit IRC | 21:23 | |
| SpamapS | It sounds to me like there's a need for test harnesses for Neutron setups to do software routing, because Neutron is trying to be an abstraction of just the first layer between vms and networking, and not just "all the networking things". | 21:23 |
| clarkb | mordred: with the soft router none of your VM stuff changes | 21:23 |
| mordred | not much I can do about that inside of my rack if the upstream gear doesn't grok it | 21:23 |
| clarkb | mordred: you only care at the router | 21:23 |
| sc68cal | clarkb: ok, let me do some more reading to catch up to you - my thought is that having DVR as a must for the grenade job is going to have us blocked for a while | 21:23 |
| mordred | clarkb: what VM stuff? | 21:23 |
| clarkb | mordred: their IPs | 21:23 |
| clarkb | mordred: and routes | 21:23 |
| mordred | clarkb: I want each of these VMs to have a direct IP from the pool we're getting | 21:23 |
| sc68cal | clarkb: we did just discuss in -meeting and consensus is to just wire up as much as possible and enable dvr later | 21:23 |
| clarkb | mordred: so them changing your IP doesn't affect your cloud | 21:23 |
| clarkb | mordred: I understand | 21:23 |
| clarkb | mordred: I am saying in the general case | 21:23 |
| mordred | clarkb: not my IP - my 1000s of IPs | 21:23 |
| clarkb | mordred: if I have a crappy network situation and it takes months to get anything done being able to control this for the VMs can be useful | 21:24 |
| clarkb | because now I have a cloud | 21:24 |
| clarkb | and it works just fine | 21:24 |
| clarkb | and the only thing the network ops folks need to know about are my racked machines | 21:24 |
| mordred | that's right | 21:24 |
| mordred | I totally agree | 21:24 |
| clarkb | which is all they dealt with prior | 21:24 |
| *** ociuhandu has quit IRC | 21:25 | |
| mordred | but I do have to have some amount of network connectivity in the first palce | 21:25 |
| *** jasondotstar has joined #openstack-infra | 21:25 | |
| mordred | and if I don't have that, software router can't change that | 21:25 |
| clarkb | sc68cal: for multinode? | 21:25 |
| sc68cal | clarkb: yes | 21:25 |
| clarkb | mordred: yes but usually get connectivity for pizza box is easy | 21:26 |
| clarkb | mordred: ask network ops to hand over contorl of a /23 and router is not | 21:26 |
| SpamapS | well thanks, now I want pizza. | 21:26 |
| clarkb | sc68cal: ok then we need to stop testing dvr with multinode I guess... which is ewrid to me since we already test that case single node | 21:26 |
| sc68cal | clarkb: no, I think we need to keep a DVR job | 21:27 |
| *** esmute__ has joined #openstack-infra | 21:27 | |
| sc68cal | clarkb: it may have just been that we were testing two objectives with one job | 21:27 |
| clarkb | sc68cal: I really don't want two jobs that dont work | 21:27 |
| clarkb | sc68cal: what is your/neutrons objective? | 21:27 |
| clarkb | my objective was run neutron on two nodes in manner prescirbed by neutron so that we tes tthat | 21:27 |
| SpamapS | Perhaps I missed the actual problem. You have vms with a single real IP, bridged onto a physical network, and you're testing with nested vms, which need to be able to talk to other nested vms on an entirely different machine. Yes? | 21:27 |
| *** bknudson has joined #openstack-infra | 21:27 | |
| clarkb | it was also the same objective with nova net and multihost | 21:27 |
| sc68cal | clarkb: we have a number of ways we perscribe running neutron.... see the networking guide | 21:28 |
| sc68cal | clarkb: obviously, DVR should be one of those, but with the stability of DVR, that is not a viable option I think | 21:28 |
| clarkb | SpamapS: yes thats completely solved at this point, but doing so requires routers and nova net and dvr both provide that for us rather than rolling our own | 21:28 |
| *** SumitNaiksatam has quit IRC | 21:28 | |
| sc68cal | clarkb: basically, can we do a multi node deployment, with no DVR, to accomplish the grenade story? | 21:28 |
| clarkb | SpamapS: with provider netowrks we would have t oroll our own | 21:28 |
| sc68cal | clarkb: and later enable DVR on it, once DVR is passing reliably? | 21:29 |
| mordred | clarkb: right. but that's what people using provider networks in a pathologicailly insane environment will have to do | 21:29 |
| clarkb | SpamapS: in the non dvr none multihost case we backhaul via gre to single router on controller | 21:29 |
| mordred | and it's ALWAYS what they'll have to do | 21:29 |
| SpamapS | clarkb: so without dvr, you can still have old schoole single-node l3 agent software routing as long as you create an l2 overlay, can't you? | 21:29 |
| clarkb | sc68cal: I don't know what the grenade story is on multinode we already test the none dvr case there | 21:29 |
| clarkb | sc68cal: so what vlaue is there in multi node without dvr? | 21:29 |
| mordred | and testing dvr does not test that provider networks work in the context of a preexisting network | 21:29 |
| clarkb | is there any? | 21:29 |
| anteaya | a partial upgrade | 21:29 |
| sc68cal | ^ this | 21:30 |
| mordred | partial upgrade is important | 21:30 |
| clarkb | sc68cal: of what? | 21:30 |
| clarkb | without dvr everything runs on the controller | 21:30 |
| anteaya | sdague: wants multinode for a partial upgrade | 21:30 |
| sc68cal | clarkb: not true | 21:30 |
| anteaya | http://lists.openstack.org/pipermail/openstack-dev/2015-June/067156.html | 21:30 |
| clarkb | sc68cal: I believe that is true in th enone dvr case, we just backhaul over l2 | 21:30 |
| sc68cal | clarkb: most deployments of neutron without DVR, that use the L3 agent have at least one dedicated network node for the l3 agent | 21:30 |
| clarkb | sc68cal: yes the controller node | 21:30 |
| clarkb | sc68cal: we can test that on a single node | 21:30 |
| sc68cal | clarkb: no, not the controller node | 21:30 |
| SpamapS | if you've a flat network of real IPs, the l3 agent isn't actually doing anything or even needed | 21:31 |
| clarkb | sc68cal: its functionally the same... | 21:31 |
| *** fawadkhaliq has joined #openstack-infra | 21:31 | |
| clarkb | sc68cal: we need to stop thinking in terms of nodes and controller/network and in terms of daemon + host pairs | 21:31 |
| clarkb | whether you use a controller + network node or just a controller node we can test that use case today on the single node setup | 21:31 |
| openstackgerrit | Emma Barber proposed openstack-infra/puppet-hound: Add module boilerplate https://review.openstack.org/198899 | 21:31 |
| *** dustins has quit IRC | 21:31 | |
| sc68cal | clarkb: so then *what* are we looking to test in multinode | 21:32 |
| clarkb | sc68cal: services running on multiple nodes | 21:32 |
| clarkb | sc68cal: so with dvr you are running services on controller and hypervisor and there is value in running on >1 node | 21:32 |
| sc68cal | clarkb: and how is that different than having a controller, then at least one network node | 21:32 |
| clarkb | sc68cal: because controller + network node is no different than running all those services on a single node if you are testing an upgrade story | 21:32 |
| *** zns has quit IRC | 21:32 | |
| clarkb | juts don't upgrade the l3 sercice on the controller... | 21:33 |
| clarkb | you dont need two nodes to test that | 21:33 |
| clarkb | SpamapS: correct | 21:33 |
| SpamapS | clarkb: just pointing out that there's value in testing that particular case. | 21:33 |
| clarkb | SpamapS: yes thats orthogonal, and I hav esaid we shoul test that in a single node | 21:34 |
| clarkb | then move to multi node if we want | 21:34 |
| *** julim has quit IRC | 21:34 | |
| SpamapS | well in multi-node it has other value. The biggest one being that metadata agent works. | 21:34 |
| *** jsavak has quit IRC | 21:34 | |
| *** jsavak has joined #openstack-infra | 21:34 | |
| clarkb | SpamapS: yes but we don't go straight to multinode | 21:34 |
| *** esmute__ has quit IRC | 21:34 | |
| clarkb | because mutlinode tsops working for a month and no one cares | 21:35 |
| sc68cal | clarkb: so, multinode means having DVR, and any other deployment scenario that uses multiple nodes for different functionality does not qualify? | 21:35 |
| sc68cal | even if the L3 agent is using vrrp or another HA tech? | 21:35 |
| SpamapS | clarkb: perhaps people do care, but it takes a month to fix? | 21:35 |
| clarkb | sc68cal: not necessarily, for nova testing nova runs real services on multiple nodes | 21:35 |
| clarkb | SpamapS: no it was just ignored | 21:35 |
| clarkb | sc68cal: so there is vlaue in multinode tetsing nova regardless of network setup | 21:35 |
| sc68cal | clarkb: nova runs real services? as opposed to neutron's.... fake services? | 21:35 |
| SpamapS | hm, yes, a test which remains unstable for an extended period of time is a test that should go away. | 21:36 |
| *** soren has joined #openstack-infra | 21:36 | |
| clarkb | sc68cal: without dvr neutron doesn't run anythong on the second node | 21:36 |
| clarkb | sc68cal: everything runs on the controller and is backhauled over l2 | 21:36 |
| sc68cal | clarkb: that's simply not true | 21:36 |
| clarkb | it is true | 21:36 |
| *** linkedinyou has quit IRC | 21:36 | |
| SpamapS | and any guarantees made based on that test having existed should be reevaluated. | 21:36 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/project-config: Added eslint-config-openstack https://review.openstack.org/194938 | 21:36 |
| clarkb | what runs on the hypervisor with neutron without dvr? | 21:36 |
| lifeless | SpamapS: dvr and l3 agent aren't needed for metadata-agent | 21:36 |
| lifeless | clarkb: l3-agent | 21:36 |
| sc68cal | clarkb: the OVS or LB agent | 21:36 |
| *** amitgandhinz has quit IRC | 21:36 | |
| SpamapS | lifeless: I agree, we were talking about what the value is running multi-node with a flat network. | 21:36 |
| SpamapS | s/running/testing/ | 21:37 |
| lifeless | SpamapS: aack. Well we know everything breaks all the time. | 21:37 |
| lifeless | so more coverage and less breakage | 21:37 |
| clarkb | sc68cal: thats also not backhauled over l2? | 21:37 |
| lifeless | :) | 21:37 |
| SpamapS | lifeless: and clarkb was pointing out that it did break and nobody responded. | 21:37 |
| sc68cal | clarkb: no, because the control plane and RPC commands have to go to the agent so they actually plug VIFs | 21:37 |
| lifeless | SpamapS: yup | 21:37 |
| clarkb | sc68cal: ok so there is value in running that then | 21:37 |
| lifeless | SpamapS: I totally agree that we need to fix that :) | 21:37 |
| sc68cal | clarkb: agreed | 21:38 |
| clarkb | sc68cal: now to go back to grenade, in your patial upgrade case is that agent one that wold not be upgraded? | 21:38 |
| mordred | woot! we've accomplished new understands! | 21:38 |
| sc68cal | clarkb: yes, it's most likely to not be upgraded at the same chronological time that the control plane is | 21:38 |
| clarkb | sc68cal: and if not, couldn't we just test that with a single node test first? | 21:38 |
| sc68cal | clarkb: it might be upgraded before, or after control plane | 21:38 |
| clarkb | so my suggestion would be to test that on a single node first | 21:39 |
| clarkb | before we ever worry about multinode grenade | 21:39 |
| clarkb | Then when that works and when multinode owrks (in whatever config) we do multinode grenade | 21:39 |
| * sc68cal thinks that when we say multinode grenade, we really should be saying DVR | 21:39 | |
| *** pc_m has quit IRC | 21:39 | |
| sc68cal | dvr+grenade | 21:39 |
| anteaya | sdague wants multinode for the grenade story | 21:39 |
| sc68cal | because apparently L3 agent running on a different node does not meet your definition | 21:39 |
| *** Sukhdev has quit IRC | 21:40 | |
| clarkb | sc68cal: l3 agent on different node is the def of dvr isn't it? | 21:40 |
| clarkb | eg do routing over there | 21:40 |
| sc68cal | clarkb: no | 21:40 |
| clarkb | sc68cal: oh do you mean netowkr node? | 21:40 |
| sc68cal | clarkb: yes, the network node | 21:40 |
| clarkb | ya I don't find do a controller + network node test to be compelling right now | 21:40 |
| clarkb | because its a single set of dameons which we can test on a single node | 21:40 |
| clarkb | adding a second node is just noise and more complications | 21:41 |
| sc68cal | clarkb: ok, then I guess we just see things differently. | 21:41 |
| clarkb | sc68cal: what is the benefit of testing controller + network node with either tempest or grenade? | 21:41 |
| clarkb | sc68cal: sorry when compared to single node | 21:41 |
| clarkb | that isn't covered by the existing single node and mutlinode test setups | 21:41 |
| sc68cal | clarkb: what is the benefit of testing multiple nodes vs single node with DVR? | 21:42 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/infra-specs: JavaScript in Infra https://review.openstack.org/198884 | 21:42 |
| *** Sukhdev has joined #openstack-infra | 21:42 | |
| clarkb | sc68cal: with multiple nodes + dvr we are running two sets of services that must work toegether | 21:42 |
| clarkb | sc68cal: instead of a single set, this has found much flakyness | 21:42 |
| sc68cal | clarkb: compared to non-dvr? where there is the same situation? two sets of services (at least) that must work together? | 21:43 |
| clarkb | sc68cal: with non dvr my understanding is/was that where is a single set of sercices for neutron on the controller | 21:43 |
| clarkb | (or network node, wherever but there is one set) | 21:43 |
| clarkb | but there is a single agent (the ovs/lb agent?) that runs on multinode non dvr | 21:44 |
| clarkb | aiui now | 21:44 |
| sc68cal | clarkb: no, I don't think that is the case | 21:44 |
| lifeless | ianw: around ? | 21:44 |
| sc68cal | clarkb: each node would have the lb/ovs agent running | 21:44 |
| clarkb | sc68cal: right | 21:44 |
| sc68cal | you could have 3 network nodes (like we do in fuel) for the L3 agent | 21:44 |
| sc68cal | *l3 agents | 21:44 |
| clarkb | sc68cal: but l3 agent != dvr? | 21:45 |
| sc68cal | then the control plane (which could also be 3 nodes, with HA technique) | 21:45 |
| sc68cal | clarkb: no, the L3 agent *can* do DVR | 21:45 |
| clarkb | ok | 21:45 |
| clarkb | sc68cal: so there is value in doing tempest + neutron without dvr | 21:45 |
| sc68cal | clarkb: on multiple nodes, yes. | 21:45 |
| *** esmute__ has joined #openstack-infra | 21:45 | |
| clarkb | sc68cal: now if we look at grenade, what are we testing there that we cna't test on single node | 21:45 |
| clarkb | I was told partial upgrades, we can test that on single node | 21:45 |
| clarkb | so now my question is can we do that first and make multinode work at all before we conflate gernade and multinode? | 21:46 |
| clarkb | because right now multinode doesnt work period | 21:46 |
| *** dboik has quit IRC | 21:46 | |
| clarkb | dvr or no dvr | 21:46 |
| openstackgerrit | Emma Barber proposed openstack-infra/puppet-hound: Add module boilerplate https://review.openstack.org/198899 | 21:46 |
| sc68cal | clarkb: then that's a problem. Let's try and fix multinode before we introduce DVR | 21:47 |
| sc68cal | in a way that will be useful for grenade | 21:47 |
| *** Longgeek has joined #openstack-infra | 21:47 | |
| anteaya | clarkb: sdague wants partial upgrades to use multinode | 21:47 |
| clarkb | anteaya: and thats fine, BUT we must first make partial upgrades without multinode work and make multinode work independently | 21:47 |
| anteaya | clarkb: agreed | 21:47 |
| *** esmute__ has quit IRC | 21:47 | |
| anteaya | if we have agreement that those are the stepping stones to getting sdague what he wants I'm fine with that | 21:48 |
| anteaya | clarkb: thanks for your passion and persistence here | 21:48 |
| *** esmute__ has joined #openstack-infra | 21:48 | |
| sc68cal | clarkb: that sounds like a goo plan to me | 21:48 |
| anteaya | any objections from the crowd? | 21:49 |
| sc68cal | to be clear, when you say multinode clarkb do you mean with DVR? | 21:49 |
| clarkb | sc68cal: I mean either way | 21:49 |
| *** yamahata has quit IRC | 21:49 | |
| clarkb | multinode does not work at all currently. There have been multiple glean fixes going in to get the jobs to run at all | 21:49 |
| sc68cal | clarkb: ok, I guess as long as we don't block on DVR getting stabilized, I'm OK | 21:49 |
| clarkb | then once thats done we hae t osee what else has broken in the interim | 21:49 |
| anteaya | working multinode (however), partial upgrade without multinode, partial upgrade with multinode | 21:50 |
| anteaya | thank you | 21:50 |
| sc68cal | sounds like we have a way forward | 21:50 |
| clarkb | sc68cal: you can run check experimental on d-g chnge sto run the multinode jobs which by default are dvr, so you can push a change to flip it to not dvr | 21:50 |
| anteaya | yay | 21:50 |
| anteaya | a direction | 21:50 |
| lifeless | ianw: if you are - https://review.openstack.org/#/c/198579/ and https://review.openstack.org/#/c/198580/ need +2A | 21:50 |
| sc68cal | clarkb: ok - I'll also write up a reply to armax's thread with what we've discussed | 21:51 |
| sc68cal | http://lists.openstack.org/pipermail/openstack-dev/2015-July/068693.html | 21:51 |
| clarkb | also | 21:52 |
| anteaya | sc68cal: keep in mind sdague hasn't weighed in yet | 21:52 |
| clarkb | I think provider netowrks are just noise at this point | 21:52 |
| anteaya | sc68cal: so leave room for him to offer his thoughts | 21:52 |
| sc68cal | clarkb: possibly | 21:52 |
| *** fawadkhaliq has quit IRC | 21:52 | |
| clarkb | lets not worry about thta at all until later as that will require some major ish changes to how we set up multi node stuff | 21:52 |
| sc68cal | anteaya: will do | 21:52 |
| anteaya | sc68cal: thank you | 21:52 |
| sc68cal | clarkb: good point, agree | 21:52 |
| clarkb | as right now multinode testing relies on routing set up by either neutron or nova net | 21:52 |
| clarkb | (whether dvr or not or mutlihost or not | 21:53 |
| anteaya | sc68cal: if you feel so inclinded you can put a draft in an etherpad and we can look at it before it goes up | 21:53 |
| anteaya | there are some emotions here, might be nice to review the wording | 21:53 |
| anteaya | if you are open to that | 21:53 |
| *** jsavak has quit IRC | 21:53 | |
| sc68cal | sounds good - do we have one currently for tracking this work? | 21:53 |
| anteaya | not as of yet, no | 21:53 |
| clarkb | no my understanding was the work was mostly done then I afk for a while and its all broken again :/ | 21:54 |
| sc68cal | clarkb: :( | 21:54 |
| anteaya | clarkb: :/ | 21:54 |
| pleia2 | that'll teach you for leaving | 21:54 |
| pleia2 | :) | 21:54 |
| anteaya | ha ha ha | 21:54 |
| clarkb | when I left my understanding was that DVR was being worked on an dwas still the way forward | 21:54 |
| sc68cal | clarkb: would an etherpad named "neutron-partial-upgrades" be sufficient? or does it need multinode in there somewhere | 21:54 |
| clarkb | but we seem to have changed opinions on that so adding or changing back to non dvr is fine | 21:55 |
| clarkb | sc68cal: thats fine | 21:55 |
| *** DericHorn-HP has joined #openstack-infra | 21:55 | |
| anteaya | well strictly speaking dvr is still being worked on, but if we all line up waiting for Swami we will be waiting a while | 21:55 |
| sc68cal | https://etherpad.openstack.org/p/neutron-partial-upgrades | 21:56 |
| anteaya | hey I got a pretty blue in the etherpad roulette | 21:56 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/project-config: Added non-voting gate-merlin-npm-run-lint https://review.openstack.org/198903 | 21:57 |
| nibalizer | fungi: pleia2 i don't have +2 on puppet-hound, can you fix that? | 21:57 |
| *** mayurig has joined #openstack-infra | 21:57 | |
| clarkb | sc68cal: anteaya I am going to push a change to add a dvr specific multihost job and switch the default job to non dvr | 21:57 |
| anteaya | sc68cal: I'm going to grab a bite, will review after you have had a pass | 21:57 |
| openstackgerrit | Monty Taylor proposed openstack-infra/project-config: Retire devstack-trusty in favor of ubuntu-trusty https://review.openstack.org/198886 | 21:57 |
| clarkb | sc68cal: anteaya so that we can iterate on them independently | 21:57 |
| sc68cal | clarkb: that sounds reasonably | 21:57 |
| sc68cal | *reasonable | 21:57 |
| anteaya | clarkb: oh that sounds good, thank you | 21:57 |
| pleia2 | nibalizer: looking | 21:57 |
| mordred | clarkb: updated that ^^ to do the || ubuntu-trusty thing you suggested | 21:57 |
| sc68cal | clarkb: can you help me a bit on the problem statement on the etherpad? | 21:58 |
| anteaya | nibalizer: thought that said puppy-hound for a minute | 21:58 |
| mordred | anteaya, nibalizer that would be awesome | 21:58 |
| nibalizer | anteaya: i would contribute to that project | 21:58 |
| anteaya | nibalizer: ha ha ha | 21:58 |
| anteaya | and a doe on the lawn | 21:58 |
| anteaya | mordred: it would | 21:58 |
| mordred | nibalizer: done | 21:59 |
| nibalizer | mordred: ty ty | 21:59 |
| pleia2 | mordred is quick | 22:00 |
| sc68cal | clarkb: can we deploy 3 nodes for the multinode story, that way we have one as the "network node" ? | 22:00 |
| mordred | pleia2: :) | 22:00 |
| pleia2 | reminds me to add it to my watched projects though! | 22:00 |
| sc68cal | so control node, network node, then hypervisor? | 22:00 |
| *** sharikapongubala has joined #openstack-infra | 22:01 | |
| *** sharikapongubala has quit IRC | 22:01 | |
| *** Sukhdev has quit IRC | 22:01 | |
| *** Sukhdev_ has joined #openstack-infra | 22:02 | |
| *** heyongli has quit IRC | 22:02 | |
| clarkb | sc68cal: currently we only have 2 node setups | 22:02 |
| *** hodos has quit IRC | 22:02 | |
| clarkb | sc68cal: but we hvae the ability to go to 3 node with nodepool if we need to | 22:02 |
| *** Kennan has joined #openstack-infra | 22:02 | |
| clarkb | sc68cal: that would be a future thing | 22:02 |
| *** heyongli has joined #openstack-infra | 22:02 | |
| sc68cal | clarkb: ok, I guess I just want to make sure we deploy the services on the nodes in a way that achieves the partial upgrade objectie | 22:03 |
| *** soren has quit IRC | 22:03 | |
| sc68cal | clarkb: my thought is, control node that has the neutron API service, then the compute node that has the L3 agent, n-cpu, and also the neutron agent | 22:03 |
| *** Kennan2 has quit IRC | 22:03 | |
| *** _diana_ has joined #openstack-infra | 22:03 | |
| sc68cal | clarkb: and upgrade the control node services, while not touching the compute node - does that fit the objective? | 22:03 |
| *** tonytan4ever has quit IRC | 22:04 | |
| sc68cal | clarkb: also, sorry for dumb questions - just getting my head around things | 22:04 |
| clarkb | sc68cal: can't we do that upgrade/or non upgrade independent of the number of nodes? | 22:04 |
| clarkb | sc68cal: this is the thing I was trying to point out before, its less about number of nodes and more about sets of services mapped onto nodes | 22:05 |
| clarkb | and we can do upgrades or not of services independent of number of nodes | 22:05 |
| clarkb | ok looks like a multinode neutron + tempest smoke + no dvr is running against neutron now | 22:06 |
| *** Longgeek has quit IRC | 22:06 | |
| clarkb | going to add a full tempest job with no dvr as that is what is missing | 22:06 |
| *** ildikov has quit IRC | 22:06 | |
| anteaya | doe on my lawn: http://imgur.com/a/mL77A | 22:06 |
| sc68cal | clarkb: ok, so then what services should not be upgraded, for a partial upgrade test | 22:06 |
| *** Sukhdev_ has quit IRC | 22:06 | |
| pleia2 | anteaya: aww :) yay canada | 22:06 |
| anteaya | pleia2: :) | 22:06 |
| clarkb | sc68cal: q-agt | 22:07 |
| clarkb | sc68cal: or the full set that would be a network node | 22:07 |
| clarkb | so everything but api | 22:07 |
| sc68cal | clarkb: ok - so for a network node that isn't also a compute host, that means just q-l3 | 22:07 |
| clarkb | does anyone know why we added the smoke multinode tests? | 22:07 |
| openstackgerrit | Thai Tran proposed openstack-infra/project-config: Create a repository for stackforge/xstatic-angular-gettext https://review.openstack.org/198905 | 22:07 |
| clarkb | at least for nova the full set were working before things went sideways | 22:07 |
| * anteaya does not know why we added the smoke multinode tests | 22:08 | |
| *** rlandy has quit IRC | 22:08 | |
| *** ddieterly has quit IRC | 22:08 | |
| clarkb | ok git log says its first step to grenade multinode | 22:09 |
| clarkb | mtreinish: is smoke not a proper subset of full? | 22:09 |
| clarkb | I would expect we get that coverage from full | 22:09 |
| *** ddieterly has joined #openstack-infra | 22:09 | |
| *** yamamoto has quit IRC | 22:10 | |
| *** Sukhdev has joined #openstack-infra | 22:10 | |
| mtreinish | clarkb: smoke is a subset of full | 22:10 |
| mtreinish | but it runs much less | 22:10 |
| clarkb | mtreinish: so we shouldn't need special smoke jobs if running full? | 22:10 |
| mtreinish | the idea of the smoke multinode job was to see if the pass rate was good enough just running smoke | 22:10 |
| mtreinish | so we could just flip the switch on multinode grenade | 22:11 |
| *** cindyo has quit IRC | 22:11 | |
| mtreinish | because full is still having some issue | 22:11 |
| mtreinish | s | 22:11 |
| openstackgerrit | Clark Boylan proposed openstack-infra/project-config: Run neutron multinode without dvr https://review.openstack.org/198906 | 22:11 |
| clarkb | anteaya: sc68cal ^ | 22:11 |
| clarkb | mtreinish: well all mutlinode is having issues | 22:11 |
| clarkb | mtreinish: so I think we proxied multinode issues as being full issues | 22:12 |
| mtreinish | clarkb: sure, but aside from that there were a couple of bugs in nova that have been identified by the full multinode job | 22:12 |
| *** cindyo has joined #openstack-infra | 22:12 | |
| mtreinish | and the idea was that maybe just running smoke (which is all grenade does) wouldn't hit those | 22:12 |
| *** yamahata has joined #openstack-infra | 22:12 | |
| mtreinish | the multinode job issues are an independent problem | 22:12 |
| *** ddieterly has quit IRC | 22:12 | |
| mtreinish | which is kinda blocking anything on this | 22:13 |
| *** ddieterl_ has joined #openstack-infra | 22:13 | |
| * mtreinish needs to respond to armax on the ML about this | 22:13 | |
| sc68cal | mtreinish: true, but I imagine the instability of DVR is also a contributing factor, yes? | 22:13 |
| mtreinish | sc68cal: I'd say there really isn't a good way to tell yet because multinode jobs have had too many other issues | 22:14 |
| mtreinish | once the infra for running them is more stable, then I'd probably be safe saying that :) | 22:14 |
| *** prad has quit IRC | 22:15 | |
| dims | fungi: i ran into semver trouble this morning, can we get rid of some versions from pypi? | 22:15 |
| Swami | anteaya: I hope I will not make you all waiting for a while. As I mentioned to kyle I will post a patch today that would at least stabilize the single node DVr job. | 22:15 |
| *** Sam-I-Am has joined #openstack-infra | 22:16 | |
| *** ddieterly has joined #openstack-infra | 22:16 | |
| clarkb | Swami: ya we are adding extra jobs so both sets of work can run independently | 22:16 |
| Swami | clarkb: thanks got it. | 22:16 |
| Sam-I-Am | howdy. | 22:16 |
| anteaya | Swami: the fact you have no help deeply concerns me | 22:16 |
| harlowja | fungi ya, so seems like taskflow 1.14 got released (which should have been 0.14) and tooz 1.18 (which should have been 0.18) | 22:16 |
| Sam-I-Am | sc68cal: here now | 22:17 |
| *** ddieterl_ has joined #openstack-infra | 22:17 | |
| * harlowja is not discounting 1.x as being bad (just unexpected) | 22:17 | |
| anteaya | Swami: I am not calling into question your dedication to task, jsut that you don't have enough hours in the day | 22:17 |
| *** jasondotstar has quit IRC | 22:17 | |
| mtreinish | harlowja: heh, I think you're stuck with 1.X now :) | 22:17 |
| mtreinish | I don't think there is a way to remove that from the wild | 22:17 |
| harlowja | :-/ | 22:17 |
| anteaya | clarkb: got a new dryer today, was putting the 5th load in to dry | 22:17 |
| mtreinish | harlowja: but someone will correct me if I'm wrong | 22:17 |
| Sam-I-Am | anteaya: thats a lot of laundry | 22:17 |
| Swami | anteaya: understood | 22:17 |
| harlowja | afaik things can get pulled from pypi | 22:17 |
| anteaya | Swami: thanks | 22:17 |
| anteaya | Sam-I-Am: I have a lot of towels | 22:18 |
| harlowja | but i want to make sure its done right mtreinish | 22:18 |
| Sam-I-Am | anteaya: i hear there's some discussion about dvr jobs | 22:18 |
| anteaya | I dislike running out of towels | 22:18 |
| fungi | dims: harlowja: there's generally not a way to go back in time and unrelease things, unless you just want to ignore anyone who has installed it and will now never upgrade | 22:18 |
| anteaya | Sam-I-Am: the backscroll be full of it | 22:18 |
| Sam-I-Am | anteaya: i just joined :/ | 22:18 |
| anteaya | sc68cal is working on a recap: https://etherpad.openstack.org/p/neutron-partial-upgrades | 22:18 |
| Sam-I-Am | sc68cal alerted me :) | 22:18 |
| anteaya | :) | 22:18 |
| harlowja | fungi dims ya, that was my worry :-/ | 22:18 |
| anteaya | you cna't be everywhere | 22:18 |
| *** ddieter__ has joined #openstack-infra | 22:18 | |
| sc68cal | Sam-I-Am: I have a link to eavesdrop where we started chatting, in that etherpad | 22:19 |
| dims | fungi: ack, we haven't broken anything so far, will wait for dhellmann and ttx to figure out next steps | 22:19 |
| anteaya | sc68cal: ah good, I was just about to link in channel | 22:19 |
| lifeless | the only way is to do epochs | 22:19 |
| * anteaya reviews clarkb's patch | 22:19 | |
| lifeless | but a) we don't support epochs in pbr yet | 22:19 |
| *** ddieterly has quit IRC | 22:19 | |
| lifeless | and b) they don't really make sense | 22:19 |
| lifeless | harlowja: whats the harm from v1 being set? | 22:20 |
| Sam-I-Am | sc68cal: reading | 22:20 |
| harlowja | lifeless harm, define harm :-P | 22:20 |
| lifeless | harlowja: I dunno, you do it ? | 22:20 |
| *** ddieter__ has quit IRC | 22:20 | |
| *** dimtruck is now known as zz_dimtruck | 22:20 | |
| harlowja | lol | 22:20 |
| *** esmute__ has quit IRC | 22:20 | |
| mordred | harlowja, dims: yah. what fungi said - once it's released, it exists - we tend to not delete things | 22:20 |
| *** doug-fish has joined #openstack-infra | 22:21 | |
| *** ddieterly has joined #openstack-infra | 22:21 | |
| dims | ack mordred | 22:21 |
| harlowja | just unexpected, was going to clean things up, remove deprecated stuff in 0.x, a few functionalities that we wanted to add | 22:21 |
| dims | we'll roll with it | 22:21 |
| mordred | harlowja: you can always do that and cut a 2.x | 22:21 |
| *** ddieterl_ has quit IRC | 22:21 | |
| mordred | and sorry | 22:21 |
| harlowja | ya, if need-be | 22:21 |
| lifeless | fungi: http://logs.openstack.org/periodic/propose-requirements-constraints-master/0ead78f/console.html#_2015-07-06_06_07_11_781 | 22:22 |
| clarkb | also keep in mind removing from pypi doesn't mean removed from pypi caches | 22:22 |
| clarkb | which pip now handily build severywhere for you | 22:22 |
| lifeless | removing from the index removes from caches | 22:22 |
| mordred | harlowja: fwiw - we're looking to hire someone to work on adding proposing-tags to gerrit so that tags can get code reviewed | 22:22 |
| lifeless | it doesn't remove from wheelhouses | 22:22 |
| clarkb | right that | 22:22 |
| fungi | lifeless: yes, clarkb pointed it out to me. i either need to do some work to add things to the server that runs on, or reqrrange that script not to call things that we aren't installing | 22:22 |
| mordred | harlowja: hasn't been the easiest thing to find | 22:22 |
| harlowja | mordred cool | 22:22 |
| lifeless | and pip doesnt' build its own wheelhouses | 22:22 |
| *** mriedem is now known as mriedem_away | 22:23 | |
| lifeless | its only workaround-infrastructure like we have that will cause wheelhouse issues | 22:23 |
| lifeless | ... | 22:23 |
| lifeless | *which we have* | 22:23 |
| lifeless | e.g. we can't sanely remove stuff. | 22:23 |
| clarkb | and many other orgs have | 22:23 |
| lifeless | yup | 22:23 |
| fungi | okay, jenkins01, 05 and 07 have been cleaned up and restarted now. 04 is in prepare for shutdown and i'll do it as soon as it's quiesced | 22:23 |
| dims | clarkb: ack thanks | 22:23 |
| *** signed8__ has joined #openstack-infra | 22:24 | |
| mordred | yah. that we have it is a good reminder that other people do too | 22:24 |
| mordred | fungi: wow. they were really unhappy weren't they? | 22:24 |
| fungi | yeah | 22:24 |
| *** signed8b_ has quit IRC | 22:25 | |
| *** signed8bit has quit IRC | 22:25 | |
| *** doug-fish1 has joined #openstack-infra | 22:25 | |
| lifeless | ok today is backport mock day | 22:25 |
| lifeless | I think I'll learn es's new repo migrate thing and see if its worth standardising on | 22:25 |
| anteaya | mtreinish and jeblair, https://review.openstack.org/#/c/198906/1 adds a new multinode job to the experimental queue for tempest, devstack and devstack-gate | 22:25 |
| mordred | lifeless: I can't tell if you're going to mock backporting, backport mocking or something else | 22:26 |
| anteaya | thanks clarkb | 22:26 |
| mordred | lifeless: also ... "es's new repo migrate thing"? | 22:26 |
| lifeless | mordred: backport mock changes since 2.7 in the std lib to the mock package on pypi | 22:26 |
| clarkb | anteaya: I did add it to neutron too right? | 22:26 |
| *** doug-fish has quit IRC | 22:27 | |
| lifeless | and then I can tell everyone to stfu and just use 'mock' | 22:27 |
| *** bswartz has joined #openstack-infra | 22:27 | |
| lifeless | since AFAICT everything else of my direct contribution work is pending reviews | 22:27 |
| anteaya | clarkb: and nova yes, just didn't feel the need to alert them to an experiemental change in their queue | 22:28 |
| anteaya | clarkb: jeblair likes to know about devstack-gate changes | 22:28 |
| clarkb | anteaya: kk just double checking I did what I thought I did | 22:28 |
| anteaya | and mtreinish is getting weary with how big the tempest and devstack experiemental queue is growing | 22:29 |
| anteaya | clarkb: yup | 22:29 |
| anteaya | so courtesy alert | 22:29 |
| clarkb | sc68cal: left a note on etherpad about what simplest next step for parital upgrade may be | 22:29 |
| clarkb | sc68cal: I think that existing job uses nova net though so it may be more copy paste then update in place | 22:29 |
| *** hogepodge has quit IRC | 22:30 | |
| *** whoops has quit IRC | 22:30 | |
| fungi | lifeless: speaking of that, i started trying to convert bindep's tests from mox to mock over the weekend (so that we can start testing it against python 3.4) but couldn't figure out where to start. i found http://mock.readthedocs.org/en/latest/compare.html but didn't know if there were more applicable examples | 22:30 |
| fungi | i'm guessing i need to spend some time familiarizing myself with both mox and mock first, because the conversion seems far from mechanical | 22:31 |
| openstackgerrit | Paul Van Eck proposed openstack-infra/puppet-refstack: Added .gitreview file https://review.openstack.org/198910 | 22:31 |
| fungi | i pulled up some of the patches where people have been going in and converting sets of tests in nova and cinder to mock, but only managed to end up more confused than when i started, i think | 22:32 |
| sc68cal | clarkb: thanks :) | 22:33 |
| mtreinish | anteaya: I'm weary of adding more random one off configs to any tempest or devstack queue :) | 22:33 |
| lifeless | fungi: yeah, they're quite different in feel | 22:33 |
| lifeless | fungi: and all of its awful | 22:33 |
| mtreinish | anteaya: but the pressure was relieved a little bit because we removed the icehouse jobs | 22:34 |
| clarkb | mtreinish: yes that is why I was trying to juts do dvr and multihost | 22:34 |
| clarkb | mtreinish: to keep the number of jobs down :/ | 22:34 |
| clarkb | and I thought I was instructed to do it that way. But things chnage | 22:35 |
| *** _diana_ has quit IRC | 22:35 | |
| mtreinish | clarkb: sure | 22:35 |
| anteaya | mtreinish: yay relief | 22:36 |
| *** rossella_s has quit IRC | 22:37 | |
| mtreinish | I'm fine with adding jobs for other configs if there are good reasons to add them. But, if the reason is just that potentially a tempest or devstack change could break the job I'm less inclined | 22:37 |
| *** jasondotstar has joined #openstack-infra | 22:37 | |
| mtreinish | especially if it's just a backend config, because things shouldn't break if the backend config changes slightly | 22:37 |
| *** rossella_s has joined #openstack-infra | 22:37 | |
| *** annegentle has quit IRC | 22:40 | |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/project-config: Switched merlin to using npm-run-test https://review.openstack.org/198911 | 22:41 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/project-config: Switched merlin to use javascript-jobs https://review.openstack.org/198912 | 22:41 |
| *** reed_ has joined #openstack-infra | 22:41 | |
| *** signed8__ is now known as signed8bit_ZZZzz | 22:41 | |
| *** cindyo has quit IRC | 22:42 | |
| anteaya | mtreinish: feel like commenting on that patch? | 22:43 |
| *** annegentle has joined #openstack-infra | 22:44 | |
| mtreinish | anteaya: I am writing it right now | 22:44 |
| anteaya | ah thank you | 22:44 |
| anteaya | week of August 17th (days to be determined) for the ops meetup: http://lists.openstack.org/pipermail/openstack-operators/2015-July/007590.html | 22:44 |
| anteaya | EmilienM: ^ | 22:44 |
| Sam-I-Am | anteaya: i wasn't here for the meat of the convo, but it sounds like we need to work out the neutron test cases | 22:44 |
| anteaya | Sam-I-Am: you pick up things fast you do | 22:45 |
| openstackgerrit | Kyle Mestery proposed openstack-infra/irc-meetings: Update neutron-drivers team meeting times. https://review.openstack.org/198806 | 22:45 |
| anteaya | Sam-I-Am: we have a few scenarios that need to be tested, clarkb and sc68cal have agreement, now we are going to post to the ml and ensure we have broader support | 22:45 |
| mtreinish | anteaya: that overlaps with linuxcon na :( | 22:45 |
| anteaya | mtreinish: :( | 22:45 |
| mtreinish | guess I'm not going to the ops meetup this time | 22:46 |
| anteaya | they need to get faster at picking their dates, and have a full schedule in front of them | 22:46 |
| anteaya | I likely will | 22:46 |
| *** hogepodge has joined #openstack-infra | 22:46 | |
| anteaya | if you have objectives to accomplish there, tell me and I will do my best | 22:46 |
| anteaya | sc68cal: I have to bail, I made an appointment for 7pm my time fully confident I wouldn't have conflicts | 22:47 |
| Sam-I-Am | anteaya: seems like a lot of it was grenade, but i'm just concerned about the general test scenarios | 22:48 |
| anteaya | sc68cal: no problem I said, I'll be done work by that time | 22:48 |
| Sam-I-Am | anteaya: i built/tested all of the networking guide scenarios in the pubcloud, so it works. | 22:48 |
| sc68cal | anteaya: haha - enjoy, we'll all chat again tomorrow I think :) | 22:48 |
| anteaya | Sam-I-Am: yes this conversation was grenade focused, as sdague had specifically said multinode for grenade | 22:48 |
| sc68cal | keep pushing the boulder up the hill :) | 22:48 |
| anteaya | and we were completely blocked so this was the conversation to unblock that | 22:48 |
| clarkb | Sam-I-Am: the general test cases exist and are run they just need some TLC | 22:48 |
| anteaya | Sam-I-Am: yay \o/ you are so awesome | 22:49 |
| anteaya | sc68cal: thank you, and yay to everyone for progress | 22:49 |
| openstackgerrit | Thai Tran proposed openstack-infra/project-config: Create a repository for stackforge/xstatic-angular-gettext https://review.openstack.org/198905 | 22:49 |
| *** chlong_ has joined #openstack-infra | 22:50 | |
| *** chlong_ has quit IRC | 22:50 | |
| *** camunoz_gone is now known as camunoz | 22:50 | |
| Sam-I-Am | clarkb: long story short, anything that tests redundancy is going to need more than one node. dvr and l3ha come to mind. the other neutron configs work fine on one node. | 22:50 |
| Sam-I-Am | dvr | 22:50 |
| Sam-I-Am | er | 22:50 |
| mtreinish | anteaya: ok, commented | 22:50 |
| clarkb | and for those that may not be aware the current multinode test setup is 2 nodes in the same cloud region (rax or hpcloud) with OVS GRE tunnel(s) between them to provide shared l2 | 22:50 |
| clarkb | Sam-I-Am: yup we have that | 22:50 |
| *** chlong has quit IRC | 22:51 | |
| anteaya | mtreinish: thank you | 22:51 |
| *** chlong has joined #openstack-infra | 22:51 | |
| clarkb | so we set up an overlay that puts at least one interface on each node on the same logical l2 network | 22:51 |
| Sam-I-Am | dvr is an interesting case because it was only tested on 1 node for juno. turns out, it broke horribly with more than 1 node (the actual use case) | 22:51 |
| Sam-I-Am | i think i might have been the first person who actually tried to deploy it | 22:51 |
| anteaya | Sam-I-Am: and my current concern is that dvr only has one person working on it, so we can't wait for it | 22:51 |
| Sam-I-Am | (mostly because i had to reverse-engineer it for the docs) | 22:52 |
| Sam-I-Am | anteaya: only one person? wow | 22:52 |
| *** ddieterly has quit IRC | 22:52 | |
| clarkb | Sam-I-Am: ya I had to do a little of that to figure out what srevices need to run where | 22:52 |
| anteaya | Sam-I-Am: yeah | 22:53 |
| anteaya | and my appointment cancelled | 22:53 |
| * anteaya moves to the etherpad | 22:53 | |
| Sam-I-Am | dvr ideally needs 3 nodes, mostly because of the central fixed-ip snat and legacy network compatibility | 22:53 |
| * clarkb maintains that we don't need 3 :) | 22:54 | |
| clarkb | if we can't test this stuff on 2 we hav failed | 22:54 |
| *** dims has quit IRC | 22:54 | |
| *** whoops has joined #openstack-infra | 22:55 | |
| *** esmute__ has joined #openstack-infra | 22:55 | |
| Sam-I-Am | i think you can test it on two, but it won't be a real world scenario imo | 22:55 |
| clarkb | Sam-I-Am: thats fine | 22:55 |
| clarkb | because this will never be anywhere near real world (see GRE tunnel details above) | 22:55 |
| greghaynes | clarkb: we hope | 22:56 |
| sc68cal | I'd like to see us move to 3 for multihost :) | 22:56 |
| mordred | I dunno | 22:56 |
| sc68cal | eventually | 22:56 |
| mordred | I'm guaranteeing someone is going to copy the gate GRE setup and run a production cloud on it | 22:56 |
| sc68cal | quantity has a quality all of its own, after all | 22:56 |
| mordred | sc68cal: yeah it does | 22:56 |
| clarkb | sc68cal: as an eventual thing thats fine, I just don't want to conflate that with making stuff work today | 22:56 |
| Sam-I-Am | mordred: haha | 22:56 |
| sc68cal | clarkb: totally agree | 22:56 |
| Sam-I-Am | mordred: how do you think i build/test the scenarios? (except i use vxlan) | 22:56 |
| clarkb | Sam-I-Am: so we don't use vxlan because we need to support nova net too | 22:57 |
| clarkb | and vxlan requires silly things like replacing arp | 22:57 |
| openstackgerrit | K Jonathan Harker proposed openstack-infra/puppet-kibana: Add kibana3 support https://review.openstack.org/188907 | 22:57 |
| sc68cal | clarkb: what do you mean? | 22:57 |
| clarkb | in a theoretical future where nova net does not exist we can probably give vxlan a go | 22:57 |
| clarkb | sc68cal: vxlan uses multicast for broadcasts, we cannot multicast on our clouds | 22:57 |
| *** DericHorn-HP has quit IRC | 22:58 | |
| clarkb | sc68cal: this is fine with neutron because it runs l2pop and similar | 22:58 |
| clarkb | sc68cal: but is not good for nova net | 22:58 |
| Sam-I-Am | clarkb: i've run multinode in pubcloud without l2pop | 22:58 |
| Sam-I-Am | and vxlans | 22:58 |
| clarkb | Sam-I-Am: which cloud? | 22:58 |
| sc68cal | clarkb: I thought vxlan can use a multicast, but will fall back when that's not present | 22:58 |
| Sam-I-Am | rax | 22:58 |
| mordred | clarkb, fungi, SpamapS: could one of you do me a quick favor and +A https://review.openstack.org/#/c/198871/ ? | 22:58 |
| clarkb | sc68cal: if thats the case I don't think I got it working for nova net when this was all being built | 22:58 |
| clarkb | and GRE just worked so it swas a meh mmoment | 22:59 |
| mordred | would like to cut a new shade for an ansible module to land | 22:59 |
| clarkb | wans't worth spending more effort on | 22:59 |
| Sam-I-Am | for a while vxlan was sort of iffy on some kernels | 22:59 |
| Sam-I-Am | still is, like on rhel 7 | 22:59 |
| clarkb | Sam-I-Am: good to know, in that case GRE is likely best for us | 22:59 |
| clarkb | as in theory we would want to run this on centos7 at some point | 22:59 |
| *** amotoki has quit IRC | 22:59 | |
| clarkb | mordred: looking | 23:00 |
| Sam-I-Am | iirc, kernel 3.13 or newer seems to do vxlan well | 23:00 |
| Sam-I-Am | rhel claims to backport stuff, but sometimes i wonder | 23:00 |
| Sam-I-Am | vxlan or gre doesn't matter, as long as the tests can be run | 23:00 |
| clarkb | mordred: Shrews would a better message be "NOva does not support egress rules"? | 23:01 |
| clarkb | mordred: Shrews as is it isn't clear what doesn't support it | 23:01 |
| Shrews | clarkb: we generally try to hide what the provider is using | 23:01 |
| mordred | clarkb: well, we are currently hiding the nova/neutron-ness of the security group rules in general - what do you think Shrews ? | 23:01 |
| clarkb | Shrews: sure but as a user of shade if I get that message then read docs I see that egress is supported and get all confused | 23:01 |
| jhesketh | Morning | 23:01 |
| *** asettle has joined #openstack-infra | 23:02 | |
| mordred | clarkb: it's an interesting point | 23:02 |
| mordred | Shrews: maybe "your cloud seems to be using nova for security groups, which do not support egress rules" | 23:02 |
| anteaya | morning jhesketh | 23:02 |
| *** annegentle has quit IRC | 23:02 | |
| mordred | Shrews: or "you're drunk, go home" :) | 23:03 |
| Shrews | clarkb: well now you'd want us to rethink the entire shade philosophy. which i'm not against, but let's be consistent right now | 23:03 |
| clarkb | I don't think its against the philosophy | 23:03 |
| clarkb | shade says don't make me worry about this stuff unless I have to | 23:03 |
| greghaynes | cowsay Egress rules are not supported in your cloud | 23:03 |
| *** heyongli has quit IRC | 23:03 | |
| clarkb | trying to perform an operation that may be supported in some cases is a situation where I have to worry about it | 23:03 |
| Shrews | clarkb: maybe not philosophy, but our axiom of "let's totally hide these things" | 23:03 |
| Shrews | which, perhaps we shouldn't be doing | 23:04 |
| mordred | but we are doing it fairly consistently | 23:04 |
| *** heyongli has joined #openstack-infra | 23:04 | |
| clarkb | because by writing an egress rule I have asserted that is what I want | 23:04 |
| clarkb | now that I don't have it I need to know why | 23:04 |
| Shrews | mordred: yeah. but i do like the "you're drunk, go home" way | 23:04 |
| mordred | Shrews: one of the public clouds just says "let's be nice" | 23:05 |
| *** austin81 has left #openstack-infra | 23:05 | |
| lifeless | mordred: http://www.catb.org/~esr/reposurgeon | 23:05 |
| mordred | I would like to say that I do not care for it as an error message | 23:05 |
| clarkb | I would expect a similar message i ftrying to floating ip with rax | 23:06 |
| clarkb | cloud doesn't support floating IPs, not "unsupported" because again if I read docs I see floating IPs are totally supported | 23:06 |
| anteaya | clarkb: would you like to respond to matt? https://review.openstack.org/#/c/198906/ | 23:06 |
| Shrews | clarkb: depends on which docs you read :) | 23:06 |
| clarkb | Shrews: the shade docs | 23:07 |
| lifeless | mordred: its a think I've heard references to but haven't looked into | 23:07 |
| mordred | clarkb, Shrews: so - turns out we have a good pattern of doing both | 23:07 |
| openstackgerrit | Merged openstack-infra/puppet-hound: Add module boilerplate https://review.openstack.org/198899 | 23:07 |
| Shrews | mordred: do we? i know we log.debug() the things | 23:07 |
| mordred | clarkb, Shrews: we return exception messages that do not mention service names, and then we log debug messages taht do | 23:07 |
| Shrews | right | 23:07 |
| mordred | so - perhaps the answer here is a log.debug with nova info | 23:07 |
| mordred | and a more genertic "your cloud doesn't like you" initial error message | 23:08 |
| mordred | so clarkb has the data, but so that we're consistent with the rest of the error messagse - at least until we decided to change more of them | 23:08 |
| mordred | clarkb: work for you? | 23:08 |
| Shrews | mordred: ++ | 23:08 |
| clarkb | mordred: sounds good | 23:08 |
| mordred | woot | 23:08 |
| Shrews | although fungi went and +2'd it | 23:09 |
| anteaya | sc68cal: the existing nova cpu partial upgrade job is the job that sdague wants re-written using multinode | 23:09 |
| Shrews | as the cow says, it's mooooot | 23:09 |
| fungi | yeah, i approved it while i was reading scrollback | 23:09 |
| Shrews | clarkb: i'll put in another review to add the debug line | 23:09 |
| fungi | missed that there was a -1 raised not-in-gerrit | 23:09 |
| Shrews | fungi: no worries | 23:09 |
| fungi | there's still time! | 23:09 |
| fungi | ;) | 23:10 |
| anteaya | sc68cal: the existing nova partial upgrade job existance is why we are in the mess we are in with regards to russellb's direction | 23:10 |
| clarkb | anteaya: mtreinish responded | 23:10 |
| *** yamamoto has joined #openstack-infra | 23:10 | |
| anteaya | sc68cal: and sdague's contribution to the conversation is sharing the perspective that he isn't interested in expanding that story | 23:10 |
| fungi | oh, seems clarkb gave in to obtuse error messages based on the promise of future perfection and rainbow unicorns | 23:10 |
| anteaya | as that increases our technical debt | 23:11 |
| anteaya | clarkb: thank you | 23:11 |
| clarkb | fungi: I wnat my unicorn! | 23:11 |
| *** jamielennox is now known as jamielennox|away | 23:11 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/shade: Raise exception for nova egress secgroup rule https://review.openstack.org/198871 | 23:11 |
| Shrews | mordred: clarkb: fungi: ^^^ | 23:11 |
| fungi | clarkb: slow-smoked unicorn steak a la Shrews coming right up, it seems | 23:12 |
| *** weshay has quit IRC | 23:12 | |
| fungi | lgtm. i'll let someone else approve | 23:13 |
| Shrews | fungi: are you cooking shark steaks these days? | 23:13 |
| fungi | Shrews: nah, they're good for driving away the tourists | 23:13 |
| anteaya | sc68cal: http://lists.openstack.org/pipermail/openstack-dev/2015-June/067915.html | 23:13 |
| clarkb | Shrews: rereviewing now | 23:14 |
| *** achanda has joined #openstack-infra | 23:14 | |
| anteaya | fungi: I found something else that will quieten the tourists though not for as long as I had hoped | 23:14 |
| anteaya | two air ambulances, one right behind the other | 23:14 |
| clarkb | approved | 23:14 |
| anteaya | the fireworks stopped and everything | 23:15 |
| fungi | anteaya: oh, i had hoped you started importing lake sharks | 23:15 |
| anteaya | I'm thinking about it | 23:15 |
| anteaya | we had a caymen a few years ago | 23:15 |
| anteaya | didn't work | 23:15 |
| fungi | those are so cute. they're like gators that never grow up | 23:15 |
| anteaya | if you hear of a supplier of lake sharks let me know | 23:15 |
| *** reed has quit IRC | 23:15 | |
| *** reed_ is now known as reed | 23:15 | |
| anteaya | that's what they said | 23:15 |
| fungi | chibi-gator | 23:15 |
| anteaya | and took pictures of them | 23:15 |
| *** reed has quit IRC | 23:16 | |
| *** reed has joined #openstack-infra | 23:16 | |
| *** dmellado_ has joined #openstack-infra | 23:16 | |
| *** reed_ has joined #openstack-infra | 23:16 | |
| *** yamamoto has quit IRC | 23:16 | |
| *** dmellado has quit IRC | 23:17 | |
| *** dmellado_ is now known as dmellado | 23:17 | |
| fungi | okay, jenkins04 has been cleaned up and restarted now too | 23:18 |
| *** annegentle has joined #openstack-infra | 23:18 | |
| anteaya | oh look at that happy graph | 23:18 |
| *** jaypipes has quit IRC | 23:18 | |
| fungi | seems the mass of excess delete nodes and ready nodes has been solved as a side effect | 23:18 |
| anteaya | yay | 23:19 |
| anteaya | good side effects | 23:19 |
| *** sabeen2 has quit IRC | 23:19 | |
| mordred | yay! | 23:20 |
| *** Krinkle is now known as Krinkle_ | 23:20 | |
| *** annegentle has quit IRC | 23:20 | |
| *** btully has quit IRC | 23:21 | |
| *** jamielennox|away is now known as jamielennox | 23:22 | |
| *** thorst has joined #openstack-infra | 23:24 | |
| *** ZZelle has quit IRC | 23:26 | |
| clarkb | anteaya: you have fireworks? | 23:26 |
| fungi | lifeless: do you have any specific input on https://review.openstack.org/198790 (bumping pbr minimum version in master global requirements to a version which strips environment markers for tests_require)? | 23:29 |
| *** heyongli has quit IRC | 23:29 | |
| *** esmute__ has quit IRC | 23:29 | |
| clarkb | its basically a noop chnge so should be fine | 23:30 |
| *** heyongli has joined #openstack-infra | 23:30 | |
| clarkb | since setuptools doesn't upgrade right? | 23:30 |
| clarkb | so if you have old pbr it odens't matter what version is listed you lose | 23:30 |
| sc68cal | anteaya: I see | 23:30 |
| *** thorst has quit IRC | 23:30 | |
| *** achanda has quit IRC | 23:30 | |
| lifeless | fungi: other than it being pointless? | 23:31 |
| clarkb | right that | 23:31 |
| mordred | clarkb: yay losing! | 23:31 |
| sc68cal | anteaya: I still think we should discuss the proposed change - since having a single person fixing DVR bugs, having pressure on him is not a good thing (tm) | 23:32 |
| *** yamamoto has joined #openstack-infra | 23:32 | |
| *** Longgeek has joined #openstack-infra | 23:32 | |
| fungi | lifeless: yeah, i think the between-the-lines expectation there is that projects adding versioned requirements will drag in a newer pbr, that is if you install anything with a versioned requirement for it. but for most projects it's unversioned in setup_requires instead | 23:33 |
| fungi | so won't help at all | 23:33 |
| clarkb | reflecting reality of the min is likely a good thing for humns though | 23:33 |
| clarkb | then they can go "but I have 0.7 installed wtf" | 23:33 |
| mordred | clarkb: wait. humans are going to use something? | 23:34 |
| fungi | not when the machines finally overthrow their oppressive meat-based masters | 23:35 |
| clarkb | then we will hav eto overthrow them and mine the spice | 23:35 |
| clarkb | ok now back to where I started today | 23:35 |
| lifeless | clarkb: so ish | 23:36 |
| lifeless | clarkb: but we already pip install -U pbr at the top of devstack | 23:36 |
| clarkb | fungi: should I go ahead and write a chnge to install zuul-cloner everywhere or is that already done? | 23:36 |
| lifeless | fungi: I've answered on the review | 23:36 |
| lifeless | tl;dr - get constraints in use everywhere, problem actually soled | 23:36 |
| lifeless | solved | 23:36 |
| fungi | clarkb: i haven't started it yet. just now emerging from evening chores | 23:36 |
| *** asettle has quit IRC | 23:36 | |
| clarkb | kk will take a look | 23:36 |
| anteaya | clarkb: not right now there are no fireworks | 23:36 |
| clarkb | we will have to do that in puppet right since we aren't using dib/nodepool there | 23:36 |
| fungi | clarkb: the trick will be that we build the zuul-cloner env in nodepool prep scripts and dib elements right now | 23:37 |
| *** heyongli has quit IRC | 23:37 | |
| anteaya | sc68cal: we should discuss a proposed change, agreed | 23:37 |
| clarkb | fungi: right, I wonder if I can pull it into slave-common and use the virtualenv puppet module | 23:37 |
| mordred | clarkb: yah - we just need it on the long-lived-slaves | 23:37 |
| lifeless | sigmavirus24: so do you have a 3.4 fix for mock? | 23:37 |
| clarkb | fungi: and stop doing it as part of dib/nodepool | 23:37 |
| mordred | clarkb: I would support that | 23:38 |
| lifeless | sigmavirus24: (if you don't thats fine, I'm just doing my backports today, so I can write one on the spot if needed) | 23:38 |
| fungi | clarkb: so will likely need to rip it out of them if we want to make puppet do it | 23:38 |
| *** heyongli has joined #openstack-infra | 23:38 | |
| anteaya | sc68cal: that the path of the proposed change including expanding the partial upgrade story in grenade following the nova example will be contrary to sdague's express wishes needs to be acknowledged | 23:38 |
| fungi | clarkb: alternatively, i can make that job bootstrap a zuul-cloner virtualenv in the workspace | 23:38 |
| clarkb | let me see how nasty it is to use the virtualenv module | 23:39 |
| anteaya | we have agreement we need to fix the multinode job, yes | 23:39 |
| anteaya | let's focus on that | 23:39 |
| lifeless | clarkb: shell out to it | 23:39 |
| anteaya | and let's leave space for sdague to express himself when he reads the email | 23:39 |
| lifeless | clarkb: oh sorry, puppet virtualenv, I dunno that | 23:39 |
| *** abitha has quit IRC | 23:39 | |
| lifeless | perhaps zuul-cloner should be its own repo and published to pypi ? | 23:39 |
| anteaya | proposing the very scenario he said no to previously gets us off to a bad start | 23:39 |
| *** whitenoise has joined #openstack-infra | 23:40 | |
| clarkb | this doesn't look bad, you can feed it a requirements file so I can just put zuul in there | 23:41 |
| greghaynes | Shrews: mordred hrm, pep8 failed for some wierd reason on https://review.openstack.org/#/c/198871/ | 23:45 |
| *** nibalizer has quit IRC | 23:45 | |
| *** achanda has joined #openstack-infra | 23:45 | |
| *** whoops has quit IRC | 23:45 | |
| *** yamamoto has quit IRC | 23:46 | |
| *** ashleighfarnham has quit IRC | 23:46 | |
| openstackgerrit | Anita Kuno proposed openstack-infra/irc-meetings: Add id and url for third party meetings. https://review.openstack.org/198722 | 23:46 |
| Shrews | greghaynes: weird reasons are the best reasons | 23:46 |
| mordred | greghaynes: that is ... what? | 23:47 |
| *** Longgeek has quit IRC | 23:50 | |
| greghaynes | yea... the test failed because it failed | 23:50 |
| *** achanda has quit IRC | 23:51 | |
| sigmavirus24 | lifeless: I don't | 23:51 |
| *** nibalizer has joined #openstack-infra | 23:52 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 23:52 | |
| jeblair | fungi: docs.o.o is still down, but there is still not a clear indication that's expected on status.r.c | 23:53 |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config: Install zuul-env venv with puppet https://review.openstack.org/198926 | 23:53 |
| clarkb | fungi: mordred ^ thats a terrible untested first pass at what this may look like | 23:53 |
| jeblair | fungi: should we start looking into that some more? | 23:53 |
| fungi | jeblair: no, it was a typo | 23:53 |
| fungi | jeblair: there is no https only http | 23:53 |
| jeblair | oh neat | 23:54 |
| *** RichardRaseley has quit IRC | 23:54 | |
| clarkb | thats a limitation of cloud sites? | 23:57 |
| fungi | dunno, i mean we haven't purchased a docs.o.o ssl cert either | 23:57 |
| *** esmute__ has joined #openstack-infra | 23:57 | |
| fungi | so i wouldn't exactly expect that to be a 100% transparent add-on | 23:57 |
| openstackgerrit | fumihiko kakuma proposed openstack-infra/project-config: Make gate-networking-ofagent-python34 non-voting https://review.openstack.org/198657 | 23:58 |
| fungi | the start of this story is that the https-anywhere plugin for firefox decided to send someone to the nonexistent https url for docs.o.o instead of http, and i didn't notice the s at first when helping troubleshoot | 23:58 |
| *** esmute__ has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
