| *** baker_ has joined #puppet-openstack | 00:04 | |
| *** baker has quit IRC | 00:07 | |
| *** guessi has quit IRC | 00:14 | |
| EmilienM | _ody: so about https://review.openstack.org/293178 | 00:18 |
|---|---|---|
| EmilienM | when it's merged, you'll have PUPPET_VERSION that you'll be able to check | 00:18 |
| EmilienM | I imagine something here: https://github.com/openstack/puppet-openstack-integration/blob/master/run_tests.sh#L63 | 00:19 |
| EmilienM | where you check PUPPET_VERSION and install the right version | 00:19 |
| *** guessi has joined #puppet-openstack | 00:19 | |
| *** openstack has joined #puppet-openstack | 00:25 | |
| _ody | Perfect. Just what I had in mind. Running what I hacked together last week against our OpenStack install now. | 00:28 |
| EmilienM | mfisch: wdyt? https://bugs.launchpad.net/puppet-keystone/+bug/1557790 | 00:31 |
| openstack | Launchpad bug 1557790 in puppet-keystone "passwords leaking on the command line" [Undecided,New] | 00:31 |
| EmilienM | zigo: I did not close it because I'm not sure if my comment is enough to say it's not totally a security issue | 00:31 |
| mfisch | EmilienM: does keystone manage accept env vars? | 00:38 |
| EmilienM | I haven't tested but it seems like if we read zigo's report | 00:38 |
| mfisch | His example showed an openstack call not keystone-manage | 00:39 |
| mfisch | looks like yes | 00:40 |
| EmilienM | mfisch: can you review https://review.openstack.org/#/c/292658/ please? | 00:42 |
| EmilienM | so much release notes to write ! | 00:42 |
| *** angdraug has quit IRC | 00:43 | |
| mfisch | lgtm | 00:43 |
| EmilienM | mfisch: I know you love ceilometer :P | 00:43 |
| mfisch | I need to go fight CI | 00:43 |
| EmilienM | mfisch: which one? | 00:43 |
| mfisch | this is BS because it was passing | 00:43 |
| mfisch | https://review.openstack.org/#/c/283898/28 | 00:43 |
| mfisch | EmilienM: replied to the bug | 00:43 |
| EmilienM | oh right | 00:43 |
| EmilienM | your patch is not passing anymore why? | 00:44 |
| mfisch | ERROR: Publisher be.certipost.hudson.plugin.SCPRepositoryPublisher aborted due to exception | 00:44 |
| EmilienM | ah | 00:44 |
| EmilienM | 'recheck' | 00:44 |
| EmilienM | it happens | 00:44 |
| mfisch | I learned how to swear at jenkins from t he infra team | 00:44 |
| *** moravec has quit IRC | 00:45 | |
| mfisch | recheck\n\nI hate you jenkins | 00:45 |
| EmilienM | lol | 00:45 |
| mfisch | can you review it once it passes | 00:45 |
| EmilienM | of course | 00:45 |
| mfisch | tomorrow if needed | 00:45 |
| EmilienM | as soon it's passing CI, I'll review it again | 00:45 |
| openstackgerrit | Merged openstack/puppet-ceilometer: releasenotes: add notes for Mitaka cycle https://review.openstack.org/292658 | 00:47 |
| EmilienM | _ody: so when my patch on infra is merged, you'll be able to run the puppet4 jobs by letting a comment on Gerrit: check experimental | 00:48 |
| EmilienM | so first patch puppet-openstack-integration with your stuff and then let a comment "check experimental" | 00:48 |
| EmilienM | it will run the jobs, you can monitor http://status.openstack.org/zuul/ and watch experimental queue | 00:49 |
| *** brandon has quit IRC | 00:52 | |
| EmilienM | omg, so close to got tempest running with ipv6 | 00:53 |
| EmilienM | mtreinish: I found 2 tests that fail with ipv6 | 00:54 |
| EmilienM | http://logs.openstack.org/21/287521/26/check/gate-puppet-openstack-integration-scenario002-tempest-dsvm-centos7/a2cdb66/console.html#_2016-03-15_23_57_52_486 | 00:54 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack-integration: Do not merge: test RabbitMQ with IPv6 https://review.openstack.org/293186 | 00:56 |
| EmilienM | Hunner: testing my changes on puppetlabs-rabbitmq here ^ | 00:56 |
| EmilienM | Hunner: I just want to make sure catalog is not failing on curl | 00:56 |
| *** xarses has joined #puppet-openstack | 00:59 | |
| EmilienM | mtreinish: https://bugs.launchpad.net/glance/+bug/1557814 | 01:03 |
| openstack | Launchpad bug 1557814 in tempest "Tempest tests fail on Glance: "Endpoint not found" when IPv6" [Undecided,New] | 01:03 |
| *** guessi has quit IRC | 01:09 | |
| *** brandon has joined #puppet-openstack | 01:10 | |
| *** brandon has quit IRC | 01:11 | |
| *** guessi has joined #puppet-openstack | 01:13 | |
| *** johnmilton has joined #puppet-openstack | 01:17 | |
| mtreinish | EmilienM: do other services work if you use an v6 address in the catalog? | 01:17 |
| EmilienM | everything works | 01:17 |
| EmilienM | mtreinish: scenarios are working I spawn VMs etc | 01:17 |
| mtreinish | so it's just using glance that doesn't like the v6 loopback address | 01:17 |
| mtreinish | oh, I know why | 01:17 |
| mtreinish | it's the glance specific http client we use for sending and recieving images (because httplib2 doesn't like chunking) | 01:18 |
| mtreinish | it's based on httlib (not httplib2) iirc and that might not have ipv6 support | 01:18 |
| EmilienM | :-/ | 01:19 |
| EmilienM | ipv6 is future ! | 01:19 |
| bkero | EmilienM: Keep on playing that trumpet! | 01:20 |
| EmilienM | bkero: TRUMPet? | 01:21 |
| EmilienM | ok I need to sleep. | 01:21 |
| bkero | >_> | 01:21 |
| bkero | EmilienM: Maybe trumpET | 01:21 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack_spec_helper: Add matchers include_regexp and have_array_of_instances_hash. https://review.openstack.org/286092 | 01:27 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack-integration: scenario002/centos7: switch RabbitMQ and OpenStack to IPv6 https://review.openstack.org/287521 | 01:43 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack-integration: Do not merge: test RabbitMQ with IPv6 https://review.openstack.org/293186 | 01:51 |
| *** johnmilton has quit IRC | 01:54 | |
| openstackgerrit | Merged openstack/puppet-openstack_spec_helper: Add matchers include_regexp and have_array_of_instances_hash. https://review.openstack.org/286092 | 01:58 |
| openstackgerrit | Merged openstack/puppet-openstack-integration: scenario003: disable sahara on ubuntu jobs https://review.openstack.org/293109 | 01:58 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack-integration: Run Tempest plugins https://review.openstack.org/283786 | 02:00 |
| *** richm has quit IRC | 02:08 | |
| openstackgerrit | Matt Fischer proposed openstack/puppet-keystone: Keystone hooks support https://review.openstack.org/283898 | 02:11 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack-integration: Do not merge: test RabbitMQ with IPv6 https://review.openstack.org/293186 | 02:13 |
| EmilienM | Hunner, _ody: in puppetlabs-rabbitmq, node_ip_address to UNSET by default is an horrible idea. | 02:14 |
| EmilienM | https://github.com/puppetlabs/puppetlabs-rabbitmq/commit/6a2509b42e3cc37382cfa9b3012aa11b638ae137 | 02:19 |
| EmilienM | 2012 ! | 02:19 |
| *** johnmilton has joined #puppet-openstack | 02:23 | |
| *** ducttape_ has joined #puppet-openstack | 02:33 | |
| *** baker_ has quit IRC | 02:34 | |
| mwhahaha | btw for the question about env vars and keystone-manage, http://docs.openstack.org/developer/keystone/configuringservices.html | 02:34 |
| mwhahaha | also, https://github.com/openstack/keystone/blob/master/keystone/cmd/cli.py#L87-L122 | 02:35 |
| EmilienM | nice | 02:35 |
| *** johnmilton has quit IRC | 02:42 | |
| mfisch | EmilienM, denis: https://review.openstack.org/#/c/283898/ I've defeated jenkins | 02:53 |
| mfisch | and found & fixed a missing item | 02:53 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-cinder: releasenotes: add notes for Mitaka cycle https://review.openstack.org/293201 | 02:54 |
| EmilienM | mfisch: lol | 02:54 |
| EmilienM | mfisch: just waiting for tripleo & fuel CI | 02:54 |
| EmilienM | and I'll review it | 02:55 |
| *** baker has joined #puppet-openstack | 02:55 | |
| EmilienM | mfisch: I need to test your patch on tripleo HA job too, do you mind if we wait tomorrow to land it? | 02:58 |
| mfisch | sure | 02:58 |
| Hunner | EmilienM: The whole unset thing was a horrible idea in general. It's been burned in a fire in most places though | 03:00 |
| mfisch | -> eod | 03:00 |
| EmilienM | Hunner: my patch on ipv6 did not work on tripleo, it's not breaking anything but need more adjustment. I'll work on it with chem tomorrow. | 03:02 |
| EmilienM | same, EOD | 03:02 |
| Hunner | k | 03:02 |
| openstackgerrit | Merged openstack/puppet-cinder: Update the solidfire module. https://review.openstack.org/291473 | 03:12 |
| *** ducttape_ has quit IRC | 03:14 | |
| *** Blah has joined #puppet-openstack | 03:20 | |
| *** ducttape_ has joined #puppet-openstack | 03:26 | |
| *** _ducttape_ has joined #puppet-openstack | 03:29 | |
| *** ducttape_ has quit IRC | 03:29 | |
| *** morazi has quit IRC | 03:32 | |
| *** baker has quit IRC | 03:43 | |
| *** ayoung is now known as ayoung_Zzz__ | 04:01 | |
| *** fedexo has joined #puppet-openstack | 04:07 | |
| *** baker has joined #puppet-openstack | 04:07 | |
| *** ayoung_Zzz__ has quit IRC | 04:12 | |
| *** _ducttape_ has quit IRC | 04:24 | |
| *** ducttape_ has joined #puppet-openstack | 04:30 | |
| *** ducttape_ has quit IRC | 04:44 | |
| *** baker has quit IRC | 05:21 | |
| *** saneax_AFK is now known as saneax | 06:17 | |
| *** astupnikov has joined #puppet-openstack | 06:20 | |
| *** Blah has quit IRC | 06:23 | |
| *** jtomasek has joined #puppet-openstack | 06:36 | |
| *** tesseract has joined #puppet-openstack | 07:16 | |
| *** tesseract is now known as Guest64967 | 07:17 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/puppet-tripleo: Fix typo https://review.openstack.org/293281 | 07:17 |
| *** jaosorior has joined #puppet-openstack | 07:18 | |
| *** Guest64967 has quit IRC | 07:21 | |
| *** leanderthal|afk is now known as leanderthal | 07:25 | |
| *** fedexo has quit IRC | 07:32 | |
| *** tesseract- has joined #puppet-openstack | 07:35 | |
| *** foxhunt_away is now known as foxhunt72 | 07:43 | |
| *** social has quit IRC | 08:04 | |
| *** starmer has joined #puppet-openstack | 08:13 | |
| *** paramite has joined #puppet-openstack | 08:20 | |
| *** kbyrne has joined #puppet-openstack | 08:22 | |
| *** starmer has quit IRC | 08:29 | |
| *** jpena|off is now known as jpena | 08:50 | |
| *** starmer has joined #puppet-openstack | 08:55 | |
| *** social has joined #puppet-openstack | 09:05 | |
| openstackgerrit | Babu Shanmugam proposed openstack/puppet-vswitch: Added custom type vs_config and OVS provider for it https://review.openstack.org/285996 | 09:21 |
| *** jistr has joined #puppet-openstack | 09:24 | |
| *** derekh has joined #puppet-openstack | 09:34 | |
| *** v1k0d3n has quit IRC | 09:35 | |
| *** v1k0d3n has joined #puppet-openstack | 09:36 | |
| *** trown has joined #puppet-openstack | 09:38 | |
| *** imcsk8_ has joined #puppet-openstack | 09:53 | |
| *** imcsk8 has quit IRC | 09:53 | |
| *** openstackgerrit has quit IRC | 09:53 | |
| *** openstackgerrit_ is now known as openstackgerrit | 09:53 | |
| *** openstackgerrit has quit IRC | 09:53 | |
| *** openstackgerrit_ has joined #puppet-openstack | 09:54 | |
| *** openstackgerrit_ is now known as openstackgerrit | 09:54 | |
| *** openstackgerrit_ has joined #puppet-openstack | 09:55 | |
| *** skolekonov has joined #puppet-openstack | 09:55 | |
| *** openstackgerrit has quit IRC | 09:55 | |
| *** openstackgerrit has joined #puppet-openstack | 09:56 | |
| *** paramite has quit IRC | 09:56 | |
| *** karimb has joined #puppet-openstack | 10:03 | |
| *** magicboiz has joined #puppet-openstack | 10:07 | |
| *** gfidente has joined #puppet-openstack | 10:13 | |
| *** gfidente has quit IRC | 10:13 | |
| *** gfidente has joined #puppet-openstack | 10:13 | |
| *** starmer has quit IRC | 10:16 | |
| *** starmer has joined #puppet-openstack | 10:17 | |
| *** cdelatte has quit IRC | 10:19 | |
| *** fwdit has joined #puppet-openstack | 10:21 | |
| *** starmer has quit IRC | 10:22 | |
| openstackgerrit | Babu Shanmugam proposed openstack/puppet-vswitch: Added custom type vs_config and OVS provider for it https://review.openstack.org/285996 | 10:30 |
| *** leanderthal has quit IRC | 10:39 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/puppet-ironic: [inspector] create a separate class for configuring logging https://review.openstack.org/293366 | 10:51 |
| *** paramite has joined #puppet-openstack | 10:53 | |
| openstackgerrit | Ivan Berezovskiy proposed openstack/puppet-oslo: Rework oslo-messaging-rabbit class https://review.openstack.org/292378 | 10:54 |
| *** johnmilton has joined #puppet-openstack | 10:55 | |
| *** trown is now known as trown|relocate | 11:12 | |
| *** openstackgerrit has quit IRC | 11:18 | |
| *** openstackgerrit has joined #puppet-openstack | 11:18 | |
| *** karimb has quit IRC | 11:30 | |
| *** magicboiz has left #puppet-openstack | 11:35 | |
| *** cdelatte has joined #puppet-openstack | 11:44 | |
| *** cdelatte has quit IRC | 11:46 | |
| *** cdelatte has joined #puppet-openstack | 11:48 | |
| EmilienM | hello | 11:49 |
| *** foxhunt72 is now known as foxhunt_away | 11:52 | |
| skolekonov | hello | 11:58 |
| _degorenko|afk | hey :) | 12:05 |
| *** _degorenko|afk is now known as degorenko | 12:05 | |
| *** jaosorior has quit IRC | 12:08 | |
| *** foxhunt_away is now known as foxhunt72 | 12:09 | |
| *** jaosorior has joined #puppet-openstack | 12:11 | |
| *** jayg|g0n3 is now known as jayg | 12:17 | |
| *** rain has joined #puppet-openstack | 12:24 | |
| *** rain is now known as Guest97758 | 12:24 | |
| *** Guest97758 is now known as leanderthal | 12:24 | |
| *** colonwq has joined #puppet-openstack | 12:28 | |
| *** trown|relocate is now known as trown | 12:35 | |
| *** fwdit has quit IRC | 12:39 | |
| *** fwdit has joined #puppet-openstack | 12:39 | |
| *** vinsh_ has quit IRC | 12:50 | |
| *** vinsh has joined #puppet-openstack | 12:50 | |
| *** karimb has joined #puppet-openstack | 12:52 | |
| *** ducttape_ has joined #puppet-openstack | 13:00 | |
| *** dprince has joined #puppet-openstack | 13:02 | |
| *** rcernin has joined #puppet-openstack | 13:03 | |
| *** morazi has joined #puppet-openstack | 13:03 | |
| *** dgurtner has joined #puppet-openstack | 13:04 | |
| *** dgurtner has quit IRC | 13:04 | |
| *** dgurtner has joined #puppet-openstack | 13:04 | |
| *** ducttape_ has quit IRC | 13:11 | |
| *** moravec has joined #puppet-openstack | 13:16 | |
| *** morazi has quit IRC | 13:17 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack-integration: Do not merge: test RabbitMQ with IPv6 https://review.openstack.org/293186 | 13:26 |
| *** morazi has joined #puppet-openstack | 13:30 | |
| *** ducttape_ has joined #puppet-openstack | 13:32 | |
| EmilienM | Hunner: chem and I updated puppetlabs-rabbitmq to work with ipv6, this time is the good one I think | 13:34 |
| EmilienM | mfisch: please look comment on https://review.openstack.org/#/c/283898/ from degorenko | 13:36 |
| degorenko | this is last comment from me :) otherwise lgtm | 13:37 |
| *** ducttape_ has quit IRC | 13:46 | |
| openstackgerrit | Michael Chapman proposed openstack/puppet-tripleo: Add keystone and db sync profiles https://review.openstack.org/292712 | 13:52 |
| *** richm has joined #puppet-openstack | 13:54 | |
| *** jpena is now known as jpena|lunch | 13:54 | |
| *** xarses has quit IRC | 13:59 | |
| *** mgagne has quit IRC | 14:07 | |
| *** mgagne has joined #puppet-openstack | 14:07 | |
| *** ducttape_ has joined #puppet-openstack | 14:10 | |
| openstackgerrit | guessi proposed openstack/puppet-nova: Extend support for systemd init system https://review.openstack.org/292450 | 14:17 |
| *** xarses has joined #puppet-openstack | 14:19 | |
| *** ducttape_ has quit IRC | 14:25 | |
| openstackgerrit | Michael Chapman proposed openstack/puppet-tripleo: Add keystone and db sync profiles https://review.openstack.org/292712 | 14:26 |
| openstackgerrit | guessi proposed openstack/puppet-keystone: Extend support for systemd init system https://review.openstack.org/292452 | 14:26 |
| *** ducttape_ has joined #puppet-openstack | 14:28 | |
| openstackgerrit | guessi proposed openstack/puppet-swift: Extend support for systemd init system https://review.openstack.org/292458 | 14:31 |
| iberezovskiy | EmilienM, hi. may you can help. some of our modules require 'kombu_ssl_certfile' and 'kombu_ssl_keyfile' params to be set together, some modules don't | 14:32 |
| iberezovskiy | I didn't find anu requirement for these params in oslo | 14:32 |
| EmilienM | afik they need to be together right? | 14:32 |
| *** ayoung has joined #puppet-openstack | 14:33 | |
| iberezovskiy | I also think so, but i'm not sure | 14:33 |
| EmilienM | if the constraint is not everywhere, let's drop the constraint so we keep backward compatibility | 14:33 |
| iberezovskiy | ok. thanks for help | 14:34 |
| *** ducttape_ has quit IRC | 14:35 | |
| *** guessi has quit IRC | 14:35 | |
| *** guessi has joined #puppet-openstack | 14:36 | |
| *** baker has joined #puppet-openstack | 14:38 | |
| *** saneax is now known as saneax_AFK | 14:40 | |
| *** chem has quit IRC | 14:41 | |
| *** baker has quit IRC | 14:42 | |
| *** baker has joined #puppet-openstack | 14:46 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/puppet-ironic: [inspector] create a separate class for configuring logging https://review.openstack.org/293366 | 14:55 |
| *** jpena|lunch is now known as jpena | 15:00 | |
| *** mdorman has joined #puppet-openstack | 15:00 | |
| *** Blah has joined #puppet-openstack | 15:03 | |
| openstackgerrit | Ivan Berezovskiy proposed openstack/puppet-oslo: Rework oslo-messaging-rabbit class https://review.openstack.org/292378 | 15:07 |
| iberezovskiy | skolekonov, please take a look ^ | 15:07 |
| openstackgerrit | Michael Chapman proposed openstack/puppet-tripleo: Add keystone and db sync profiles https://review.openstack.org/292712 | 15:08 |
| openstackgerrit | Dmitry Tantsur proposed openstack/puppet-ironic: [inspector] create a separate class for configuring logging https://review.openstack.org/293366 | 15:09 |
| skolekonov | iberezovskiy, +1, thanks | 15:19 |
| iberezovskiy | good, thanks you as well | 15:20 |
| *** Blah has quit IRC | 15:26 | |
| *** dhollinger has joined #puppet-openstack | 15:26 | |
| *** karimb has quit IRC | 15:27 | |
| mjblack | looks like problem with puppet 4.3 and puppet-horizon | 15:30 |
| openstackgerrit | guessi proposed openstack/puppet-swift: Extend support for systemd init system https://review.openstack.org/292458 | 15:30 |
| EmilienM | mjblack: that's possible, we don't functionaly test it. _ody was working on it last night. | 15:32 |
| mjblack | I'm seeng a stupid error though | 15:32 |
| mjblack | Error: Could not find dependent Package[apache2] for Package[horizon] at /etc/puppetlabs/code/environments/production/modules/horizon/manifests/init.pp:362 | 15:32 |
| mjblack | but I'll submit a bug report :D | 15:33 |
| EmilienM | mjblack: show your manifest please | 15:33 |
| mjblack | its all hiera data driven | 15:33 |
| mjblack | my default node does a hiera_include | 15:33 |
| mjblack | thats it | 15:33 |
| EmilienM | mjblack: do you declare ::apache? | 15:33 |
| mjblack | nope, it's declared in horizon though | 15:34 |
| EmilienM | mjblack: mhh, please run puppet with --trace and file a bug. We'll see what happens. | 15:34 |
| *** dgurtner has quit IRC | 15:34 | |
| mjblack | EmilienM: https://gist.github.com/mjblack/6ee6abc3406d2c177924 | 15:36 |
| mjblack | its getting the package dependency for apache in a different manifest though | 15:36 |
| *** jaosorior has quit IRC | 15:36 | |
| mjblack | under wsgi | 15:36 |
| EmilienM | mjblack: show me the classes you declare | 15:38 |
| EmilienM | I'm not sure that's a puppet4 thing. | 15:38 |
| mjblack | ::memcached, ::horizon | 15:38 |
| mjblack | I tried adding ::apache | 15:38 |
| mjblack | oh yeah and openstack_extras::repo::debian::ubuntu | 15:39 |
| EmilienM | mjblack: https://github.com/openstack/puppet-openstack-integration/blob/master/manifests/horizon.pp#L3-L12 | 15:39 |
| EmilienM | this is how we deploy | 15:39 |
| EmilienM | degorenko: if you see missing reno notes, feel free to add them | 15:41 |
| mjblack | this is what I'm passing in https://gist.github.com/mjblack/325610781c31926ad986 | 15:41 |
| EmilienM | I'm trying to add a good number of them this week, but I might missed some of them | 15:41 |
| degorenko | EmilienM, ack | 15:42 |
| mjblack | EmilienM: I'm not sure what happened but maybe it worked because the lookup for the resource looked at both title and namevar? | 15:48 |
| openstackgerrit | Denis Egorenko proposed openstack/puppet-glance: WIP Add Glance Glare API service https://review.openstack.org/293524 | 15:48 |
| mjblack | apache module defines the package title as httpd but sets namevar to whatever resolves from params | 15:49 |
| mjblack | I forced my horizon to use httpd and now it's not having dependency issues | 15:49 |
| EmilienM | degorenko: -1 ! acceptance ! lol | 15:49 |
| *** rahulait has joined #puppet-openstack | 15:49 | |
| degorenko | EmilienM, okaaaay :( | 15:49 |
| EmilienM | degorenko: only for centos please | 15:50 |
| degorenko | EmilienM, sure | 15:50 |
| degorenko | EmilienM, 1443 lines lol | 15:50 |
| EmilienM | mjblack: oh maybe let me check code | 15:50 |
| EmilienM | degorenko: yeah... | 15:50 |
| degorenko | it will be splitted | 15:50 |
| degorenko | this is wip yet | 15:50 |
| EmilienM | degorenko: no need, I think | 15:50 |
| EmilienM | mjblack: what version of puppet-horizon? | 15:51 |
| mjblack | stable/liberty | 15:51 |
| EmilienM | ah | 15:51 |
| mfisch | degorenko: ping | 15:51 |
| degorenko | EmilienM, 1443 - so many lines. It's hard to review. But if it is ok - then ok :) | 15:51 |
| degorenko | mfisch, pong | 15:51 |
| mfisch | replying to your comment | 15:52 |
| * degorenko looking | 15:52 | |
| EmilienM | mjblack: I don't see where the dep is set at L362 | 15:52 |
| EmilienM | mjblack: https://github.com/openstack/puppet-horizon/blob/stable/liberty/manifests/init.pp#L362-L366 | 15:52 |
| mjblack | EmilienM: its not there, it's in manifests/wsgi/apache.pp | 15:52 |
| EmilienM | are you running a fork? | 15:52 |
| EmilienM | ah | 15:52 |
| mjblack | https://github.com/openstack/puppet-horizon/blob/stable/liberty/manifests/wsgi/apache.pp#L150 | 15:53 |
| mjblack | which translates to apache2 | 15:53 |
| mjblack | https://github.com/openstack/puppet-horizon/blob/stable/liberty/manifests/params.pp#L24 | 15:53 |
| EmilienM | you found a bug I guess? | 15:54 |
| EmilienM | but why our CI does not have it, mhh | 15:54 |
| mjblack | looks like it | 15:54 |
| mjblack | what version of puppet 4 is it running? | 15:54 |
| mjblack | it's quite possible that puppetlabs "fixed" a bug | 15:54 |
| EmilienM | let's see what is resource name is puppetlabs-apache | 15:54 |
| mjblack | https://github.com/puppetlabs/puppetlabs-apache/blob/master/manifests/package.pp#L60 | 15:55 |
| mjblack | its httpd | 15:55 |
| EmilienM | Package['httpd'] in master | 15:55 |
| mjblack | it hasnt changed | 15:55 |
| mjblack | I checked at like 1.3.0 tag | 15:55 |
| EmilienM | mjblack: I guess you're good to submit a patch :) | 15:55 |
| mjblack | already did :D | 15:55 |
| mjblack | might be easy to solve | 15:56 |
| mfisch | degorenko: does my answer make sense? | 15:56 |
| EmilienM | mjblack: where is patch? it's not in my queue | 15:56 |
| mjblack | oh sorr | 15:56 |
| mjblack | I submitted bug report | 15:56 |
| EmilienM | ah | 15:56 |
| EmilienM | so send patch | 15:56 |
| mjblack | I'll test out an idea to fix it | 15:56 |
| mjblack | wll do | 15:56 |
| mjblack | *will do | 15:56 |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-openstack_spec_helper: Add support for spec. https://review.openstack.org/293529 | 15:58 |
| mjblack | ah ok, rename it to just httpd should work | 16:00 |
| degorenko | mfisch, +2 from me | 16:00 |
| mfisch | ty | 16:00 |
| degorenko | good work! | 16:00 |
| mfisch | EmilienM: your turn | 16:00 |
| degorenko | that's really helpful | 16:00 |
| EmilienM | mfisch: yeah | 16:01 |
| EmilienM | mfisch: boom | 16:01 |
| mfisch | thanks | 16:02 |
| openstackgerrit | guessi proposed openstack/puppet-swift: Extend support for systemd init system https://review.openstack.org/292458 | 16:03 |
| EmilienM | mjblack: I renamed https://bugs.launchpad.net/puppet-horizon/+bug/1558131 | 16:03 |
| openstack | Launchpad bug 1558131 in puppet-horizon "package dependency issue on Ubuntu" [Undecided,New] | 16:03 |
| mjblack | ok | 16:03 |
| EmilienM | mjblack: our CI did not see it because we install other components in WSGI that already contain the resource I think | 16:03 |
| EmilienM | so good catch man | 16:03 |
| mjblack | it shouldnt work even then | 16:04 |
| mjblack | it has to probably be puppet version specific | 16:04 |
| mjblack | because it will never see Package['apache2 | 16:04 |
| mjblack | when it does lookup | 16:04 |
| mjblack | it wll only find httpd | 16:04 |
| mjblack | in puppet 4.3.2 | 16:05 |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-keystone: Remove matcher definition. https://review.openstack.org/292563 | 16:05 |
| mjblack | the fix I'm testing right now will hopefully solve the issue for all versions | 16:05 |
| openstackgerrit | guessi proposed openstack/puppet-keystone: Extend support for systemd init system https://review.openstack.org/292452 | 16:06 |
| *** chem has joined #puppet-openstack | 16:11 | |
| EmilienM | mjblack: good candidate for backport | 16:12 |
| openstackgerrit | guessi proposed openstack/puppet-nova: Extend support for systemd init system https://review.openstack.org/292450 | 16:13 |
| openstackgerrit | guessi proposed openstack/puppet-keystone: Extend support for systemd init system https://review.openstack.org/292452 | 16:16 |
| EmilienM | guessi: I commented on your patches, I think it breaks unit test jobs for recent version of puppet | 16:17 |
| EmilienM | if you're not familiar with it, let me know | 16:17 |
| openstackgerrit | guessi proposed openstack/puppet-swift: Extend support for systemd init system https://review.openstack.org/292458 | 16:17 |
| guessi | EmilienM: should be fixed now, $::var --> ${::var} | 16:19 |
| EmilienM | excellent | 16:19 |
| mwhahaha | no the correct fix was to actually define the var in tests | 16:21 |
| mwhahaha | it was failing because that fact is empty in the tests | 16:21 |
| EmilienM | mwhahaha: and recent version of rspec is requiring it? | 16:22 |
| mwhahaha | recent version of puppet's versioncmp function | 16:23 |
| guessi | how do i get jenkins log for previous patch set? don't know why is the fact ::operatingsystemmajrelease resolved as "Undef" in tests? | 16:27 |
| *** fwdit has quit IRC | 16:30 | |
| mwhahaha | it's a limitation because those modules haven't been switched to use rspec-puppet-facts which would populate that fact | 16:30 |
| *** astupnikov has quit IRC | 16:31 | |
| mwhahaha | It also points to is a deficiency in test coverage around that as well | 16:31 |
| mwhahaha | so wrapping it in "" stops the error but we aren't properly testing that condition | 16:31 |
| _ody | EmilienM: Yeah. To my surprise last night, I found that not a single scenario is running under Puppet 4...and I have no idea why. I'll need to consult with Hunner so see if I am totatlly missing something obvious. | 16:33 |
| EmilienM | that's not a good news :( | 16:33 |
| _ody | We're running the kilo modules in production on puppet 4 so I am confused. | 16:33 |
| EmilienM | oh really?? | 16:33 |
| _ody | yeah. | 16:33 |
| mjblack | EmilienM: looks like this is the issue here... https://tickets.puppetlabs.com/browse/PUP-5021 | 16:34 |
| EmilienM | mjblack: nice catch, Hunner found it! | 16:34 |
| mjblack | oh? | 16:34 |
| EmilienM | well, he found the bug | 16:35 |
| EmilienM | but no fix :( | 16:35 |
| *** trown is now known as trown|lunch | 16:36 | |
| mjblack | EmilienM: well I just submitted the fix that works for me :D | 16:38 |
| mjblack | https://review.openstack.org/293551 | 16:38 |
| EmilienM | mjblack: nice | 16:39 |
| EmilienM | mjblack: we'll need to backport it to liberty | 16:40 |
| mjblack | should be easy to do, I'll submit that as well | 16:40 |
| *** severion has joined #puppet-openstack | 16:41 | |
| EmilienM | mjblack: after it merges | 16:41 |
| degorenko | EmilienM, can you please take a look on: https://review.openstack.org/#/c/293557/ :) | 16:42 |
| *** leanderthal is now known as leanderthal|afk | 16:43 | |
| EmilienM | degorenko: wait, ${name}/volume_clear was the wrong place, right? | 16:43 |
| degorenko | this is probably incorrect, because options, that you added is for default section | 16:43 |
| EmilienM | it's in DEFEAULT | 16:43 |
| degorenko | yes :) | 16:44 |
| EmilienM | ok | 16:44 |
| EmilienM | +2 | 16:44 |
| degorenko | EmilienM, i will ask also cinder guys to confirm that | 16:44 |
| Hunner | I'll ping for PUP-5021 | 16:44 |
| EmilienM | Hunner: thx | 16:44 |
| degorenko | and as we don't release ever bdd it is ok to just remove this option | 16:44 |
| EmilienM | yeah | 16:45 |
| guessi | just found a related topic for the "Undef" problem, https://tickets.puppetlabs.com/browse/PUP-5683?focusedCommentId=252877 | 16:45 |
| Hunner | mjblack: Is there a way to do it with the title instead of the collector? | 16:46 |
| mjblack | Hunner: only way is to explicitly say 'httpd' | 16:46 |
| guessi | looks like it just happens to puppet > v4.3 | 16:46 |
| Hunner | :/ | 16:46 |
| mjblack | or we create a new variable | 16:46 |
| mjblack | I went to the simpliest change | 16:47 |
| Hunner | both ick | 16:47 |
| openstackgerrit | Denis Egorenko proposed openstack/puppet-cinder: Remove volume_clear option from Cinder BDD backend https://review.openstack.org/293557 | 16:50 |
| degorenko | EmilienM, updated commit message can you re+2 please? mwhahaha please ^^ | 16:50 |
| mjblack | Hunner: we can always go with explicit naming since on the apache module side it is explicitly named title | 16:52 |
| openstackgerrit | Merged openstack/puppet-vswitch: Complete ovs_port fix for Ubuntu https://review.openstack.org/292390 | 16:53 |
| guessi | EmilienM, _ody, mwhahaha: sorry guys, it's 01:00 am here, got to sleep, see you! | 16:55 |
| mjblack | Hunner: I wonder if they fixed it as part of this? https://tickets.puppetlabs.com/browse/PUP-1073 | 16:58 |
| *** guessi has quit IRC | 17:01 | |
| openstackgerrit | Merged openstack/puppet-openstack-integration: Add vs_port to provision template https://review.openstack.org/291811 | 17:02 |
| openstackgerrit | Dmitry Tantsur proposed openstack/puppet-ironic: [inspector] create a separate class for configuring logging https://review.openstack.org/293366 | 17:05 |
| openstackgerrit | Liao Penghui proposed openstack/puppet-keystone: Make request by service token if got unauth error https://review.openstack.org/293579 | 17:09 |
| openstackgerrit | Merged openstack/puppet-tripleo: Allow enabling authentication on haproxy.stats https://review.openstack.org/290896 | 17:09 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-tripleo: Allow enabling authentication on haproxy.stats https://review.openstack.org/293581 | 17:10 |
| openstackgerrit | Liao Penghui proposed openstack/puppet-keystone: Make request by service token if got unauth error https://review.openstack.org/293579 | 17:11 |
| *** jistr has quit IRC | 17:12 | |
| openstackgerrit | Merged openstack/puppet-openstack_spec_helper: Add support for spec. https://review.openstack.org/293529 | 17:14 |
| EmilienM | iberezovskiy: https://review.openstack.org/#/c/292378/ looks solid to me, I'll let mwhahaha / degorenko doing a last review to make sure it's correct | 17:16 |
| degorenko | already looking at this | 17:16 |
| iberezovskiy | EmilienM, thanks. I've also prepared nova patch to switch on oslo :) | 17:17 |
| degorenko | EmilienM, that's totally good idea to me! ^^ To have one point for configuring oslo. Because now we have different settings and checks across all modules | 17:17 |
| degorenko | we must to switch all modules (if it possible) | 17:18 |
| degorenko | mwhahaha, you probably also want to review oslo, so, i give you ability to +A it :D | 17:19 |
| mwhahaha | k | 17:20 |
| *** trown|lunch is now known as trown | 17:25 | |
| mwhahaha | found issue, commented | 17:26 |
| openstackgerrit | Andrew Woodward proposed openstack/puppet-ceph: Change profile to not use $name https://review.openstack.org/291875 | 17:27 |
| EmilienM | mwhahaha: your comment is valid | 17:30 |
| *** paramite has quit IRC | 17:32 | |
| *** foxhunt72 is now known as foxhunt_away | 17:36 | |
| degorenko | mjblack, i left a few comments on your horizon change | 17:37 |
| degorenko | pelase take a look | 17:37 |
| degorenko | please* | 17:37 |
| iberezovskiy | EmilienM, mwhahaha yep, you are right. will be updated in a moment | 17:37 |
| mjblack | degorenko: will do | 17:38 |
| mjblack | just got back from a reboot | 17:38 |
| *** skolekonov has quit IRC | 17:39 | |
| *** paramite has joined #puppet-openstack | 17:40 | |
| _ody | Hunner: It is Wednesday and you're not at your desk. Where be you? | 17:41 |
| mjblack | degorenko: the issue is that puppet used to accept either title or namevar for dependency resolution but in puppet 4.3 it is not working for namevar | 17:42 |
| degorenko | mjblack, ah, it's not working for tittle at all? | 17:42 |
| mjblack | it doesnt affect centos because title and namevar will always be the same | 17:42 |
| mjblack | it works for title but on ubuntu the package name is apache2 not httpd | 17:42 |
| mjblack | so to get around it, that's why name parameter is set to $::horizon::params::http_service | 17:43 |
| degorenko | $::horizon::params::http_service for ubuntu != for centos | 17:43 |
| degorenko | and for package too | 17:43 |
| mjblack | yup | 17:43 |
| degorenko | yup | 17:43 |
| degorenko | but | 17:43 |
| degorenko | apache creates resource called 'httpd' exactly | 17:43 |
| degorenko | and we are trying to manage apache | 17:43 |
| degorenko | of course that's not correct | 17:44 |
| mjblack | so I can easily change it to be Package['httpd'] | 17:44 |
| mjblack | hard code it | 17:44 |
| degorenko | well, i'm just want to have same vision of problem that you | 17:45 |
| degorenko | so, my way is to rely on apache module, your way is to relay on package real name | 17:45 |
| degorenko | that's two valid points | 17:45 |
| degorenko | i can remove my -1 | 17:45 |
| openstackgerrit | Ivan Berezovskiy proposed openstack/puppet-oslo: Rework oslo-messaging-rabbit class https://review.openstack.org/292378 | 17:45 |
| mjblack | I have to check and make sure that service is not affected too | 17:45 |
| mjblack | since its the same deal there too | 17:46 |
| degorenko | EmilienM, we need you :) | 17:46 |
| degorenko | mjblack, yes please, my -1 is mostly for service, not for your way | 17:46 |
| degorenko | mwhahaha, your thoughts? | 17:46 |
| mjblack | I think the reason its "broken" is that there was one or two bugs relating to having namevar be the same package name but title different and also different providers. Like for example uninstall an apt package and install a pip package | 17:47 |
| mjblack | it was that if two resources had the same namevar it would fail with dup resource declaration | 17:48 |
| mjblack | yup, package looks like the only one to be affected | 17:50 |
| *** derekh has quit IRC | 17:50 | |
| mjblack | Hunner: I think the bug you found might be a feature instead :) | 17:50 |
| EmilienM | degorenko: how can I help? I was about leaving for lunch | 17:51 |
| degorenko | EmilienM, we are talking about https://review.openstack.org/#/c/293551/1/manifests/wsgi/apache.pp | 17:51 |
| degorenko | mjblack, also i've just noticed | 17:52 |
| EmilienM | degorenko: right, we can use Package['httpd'], I think it would be fine | 17:52 |
| degorenko | that we have Package resource with http_service name | 17:52 |
| degorenko | i think that's real problem | 17:52 |
| degorenko | :) | 17:52 |
| EmilienM | both Service & Package are 'httpd' | 17:52 |
| *** rcernin has quit IRC | 17:53 | |
| degorenko | EmilienM, nope :) | 17:53 |
| EmilienM | they should be I mean | 17:53 |
| degorenko | https://github.com/openstack/puppet-horizon/blob/master/manifests/params.pp | 17:53 |
| EmilienM | right | 17:53 |
| Hunner | EmilienM: For rabbitmq, the staging namespace is "puppet" not "voxpupuli"; sorry | 17:53 |
| EmilienM | degorenko: let's update the patch to have Service['httpd'] and Package['httpd'] right? | 17:53 |
| Hunner | voxpupuli used to be puppet community, so has the name "puppet" on the forge, and currently there is no way to move modules on the forge so it'll be that way for a while | 17:53 |
| degorenko | mjblack, i think main problem that we are using package resource with service_name parameter | 17:54 |
| Hunner | mjblack: orly? | 17:54 |
| degorenko | EmilienM, that's correct for me | 17:54 |
| mjblack | Hunner: it only affects packages, not services | 17:54 |
| EmilienM | degorenko: good. I go lunch. | 17:54 |
| EmilienM | Hunner: ok, feel free to revert my patch if needed. | 17:54 |
| Hunner | mjblack: Can you actually declare multiple services with the same name? | 17:55 |
| mjblack | degorenko: we can make http_service no distro dependent :P | 17:55 |
| mjblack | *not | 17:55 |
| degorenko | mjblack, yes we can | 17:55 |
| Hunner | EmilienM: I'm not going to revert it; when I get to a machine with github access I can update the metadata to say puppet though | 17:55 |
| degorenko | but i don't think, that using package with title from service name is correct :) | 17:55 |
| mjblack | Hunner: we do, $http_service is set to apache2 on ubuntu | 17:55 |
| degorenko | on any distros | 17:55 |
| mjblack | its almost the same setup as the package resource thats failing | 17:55 |
| Hunner | mjblack: Why do you have two apache2 services? | 17:55 |
| mjblack | but its working correctly | 17:55 |
| mjblack | no no | 17:55 |
| mjblack | puppet 4.3 no longer allows resource lookup on both title and name | 17:56 |
| mjblack | it only works on title | 17:56 |
| mjblack | for packages | 17:56 |
| mjblack | it works correctly for both with service rsources | 17:56 |
| Hunner | Because only packages has the property of being able to have duplicate types+titles without conflicting | 17:56 |
| mjblack | no | 17:56 |
| mjblack | title has to always be unique | 17:57 |
| Hunner | package { 'foo': name => 'foo', } package { 'foo': name => 'bar', } | 17:57 |
| mjblack | name was changed to not require uniqeness | 17:57 |
| Hunner | err, yeah, other way around | 17:57 |
| Hunner | package { 'foo': name => 'foo', } package { 'bar': name => 'foo', } | 17:57 |
| Hunner | but if you do that with service, it'll break, yeah? | 17:57 |
| mjblack | some bug reports I read were asking for it so you can uninstall a deb package and install a pip package | 17:57 |
| mjblack | nope it doesnt break with service | 17:57 |
| Hunner | orly | 17:58 |
| mjblack | Debug: Executing: '/etc/init.d/apache2 start' | 17:58 |
| mjblack | Notice: /Stage[main]/Apache::Service/Service[httpd]: Triggered 'refresh' from 1 events | 17:58 |
| mjblack | and thats occurring from horizon::wsgi::apache | 17:59 |
| degorenko | mjblack, that's correct for service, because we have proper names for service resource | 17:59 |
| mjblack | degorenko: no, thats not the issue | 17:59 |
| degorenko | why not? | 17:59 |
| mjblack | it's exactly the same as package with regards to title/name | 17:59 |
| mjblack | package is apache2, service is apache2 | 18:00 |
| degorenko | package in debian is not apache2 | 18:00 |
| degorenko | oops | 18:00 |
| degorenko | wait | 18:00 |
| degorenko | i'm looking at horizon params | 18:00 |
| degorenko | sec | 18:00 |
| mjblack | Package['horizon'] -> Package[$::horizon::params::http_service] | 18:01 |
| mjblack | File[$::horizon::params::config_file] ~> Service[$::horizon::params::http_service] | 18:01 |
| mjblack | the package title and service title in puppetlabs-apache is 'httpd' | 18:01 |
| Hunner | But you're using Apache::Service::Service[httpd] it says above | 18:01 |
| mjblack | thats because puppetlabs-apache has service resource defined with 'httpd' | 18:02 |
| mjblack | but the name is 'apache2' | 18:02 |
| degorenko | yes | 18:02 |
| degorenko | but resource name is httpd | 18:02 |
| mjblack | package { 'httpd': name => 'apache2' }, service { 'httpd': name => 'apache2' } | 18:03 |
| degorenko | you are using exactly resource Package['what_ever'] | 18:03 |
| mjblack | right | 18:03 |
| degorenko | such construction is exactly using title | 18:03 |
| degorenko | and you don't have resource with title apache3 | 18:03 |
| degorenko | apache2* | 18:03 |
| Hunner | mjblack: where is the other service with a name of apache2 declared? | 18:03 |
| mjblack | correct | 18:03 |
| degorenko | so, you have resource 'httpd' | 18:03 |
| degorenko | that comes from apache | 18:03 |
| degorenko | on any distros | 18:04 |
| mjblack | Hunner: there is only one service resource | 18:04 |
| Hunner | -.- | 18:04 |
| Hunner | That's what I've been saying; irc communication ftl | 18:04 |
| degorenko | guys, let me summarize a bit | 18:04 |
| mjblack | lol | 18:04 |
| Hunner | package{foo:name=>foo}package{bar:name=>foo} // is valid | 18:04 |
| Hunner | service{foo:name=>foo}service{bar:name=>foo} // is invalid | 18:05 |
| degorenko | so, we have package resource from apache module with title 'httpd'. In horizon module we are trying to use resource with title 'apache2' | 18:05 |
| degorenko | that's problem ^ | 18:05 |
| mjblack | Hunner: no, both are valid | 18:05 |
| mjblack | the issue is when you have package { 'httpd': name => 'apache2' } and then say Package['apache2'] -> File['some_resource'] | 18:06 |
| degorenko | package{foo:name=>foo}package{bar:name=>foo} - two different resources - valid, service{foo:name=>foo}service{bar:name=>foo} - two different resources, valid | 18:06 |
| mjblack | yes, I think it was intentionally done that way so you can do something like this | 18:07 |
| degorenko | mjblack, you can't do this package { 'httpd': name => 'apache2' } and then say Package['apache2'] -> File['some_resource'] | 18:07 |
| degorenko | mjblack, looks like we have one vision of problem :D | 18:07 |
| Hunner | Yeah, this seems like a big misunderstanding :/ | 18:07 |
| mjblack | package { 'remote-deb': name => 'python-keystoneclient', provider => 'apt' } package { 'install_keystone': name => 'python-keystoneclient, provider => 'pip' } | 18:08 |
| Hunner | Also, you're right that two services with the same $name is valid! I was wrong about that... | 18:08 |
| Hunner | Though I thought that change was only made for packages | 18:08 |
| mjblack | when you do the ~> on debian with service resource Service['apache2'] it will resolve to Service['httpd'] | 18:09 |
| mjblack | and that is how package used to work | 18:09 |
| mjblack | Package['apache2'] would resolve to Package['httpd'] cause the name var on ubuntu was apache2 | 18:10 |
| mjblack | maybe I'm being confusing :D | 18:10 |
| degorenko | :D lol, well, actually, for package provide title can be package name, if name property is undef | 18:11 |
| mjblack | right | 18:11 |
| degorenko | package { 'vim': ensure => present } == package { 'ololo': ensure => present, name => 'vim' } - they both will install one package | 18:12 |
| mjblack | but with regards to my change, we can add a new var for httpd package, hard code package name, or what I have proposed in the change. | 18:12 |
| degorenko | so long discussion on so small fix :) | 18:14 |
| mjblack | degorenko: people wanted the ability to uninstall packages from one provider and install the same package name but from a different provider | 18:14 |
| degorenko | mjblack, that's ok, yes | 18:14 |
| mjblack | here is what I found https://tickets.puppetlabs.com/browse/PUP-1073 | 18:14 |
| degorenko | hmmm | 18:16 |
| degorenko | that's really bug in puppet | 18:16 |
| degorenko | it should be two different providers | 18:16 |
| degorenko | resources* | 18:16 |
| degorenko | Referring to Package resources by Package[$namevar] in Puppet code doesn't work for a composite namevar, so it no longer works for Package. Querying by Package[$title] still works. | 18:17 |
| degorenko | so, looks like solution with title will work anyway | 18:17 |
| mjblack | yeah | 18:17 |
| mjblack | the change I made works too | 18:18 |
| degorenko | but in this case we will have duplication error, if some one wants to use another provider, that's your point? | 18:18 |
| mjblack | no, my point is that it used to work with either title or name with resolving dependencies | 18:18 |
| *** tesseract- has quit IRC | 18:19 | |
| mjblack | now it doesnt for packages | 18:19 |
| mjblack | it only works on title | 18:19 |
| mjblack | unless you do a resource collector | 18:19 |
| degorenko | why do we need at all this dependency? package['horizon'] -> package['apache'] ? | 18:21 |
| degorenko | the important here is restarting of service | 18:22 |
| colonwq | xarses, Are you around? I think I found an other puppet-ceph design issue. | 18:23 |
| *** degorenko is now known as _degorenko|afk | 18:23 | |
| *** Blah has joined #puppet-openstack | 18:25 | |
| mjblack | degorenko,_degorenko|afk: this might be better at describing the problem https://gist.github.com/mjblack/3aebc4f294f41d50a36d | 18:27 |
| openstackgerrit | Sergey Kolekonov proposed openstack/puppet-neutron: Add support for ovsdb_interface option https://review.openstack.org/293639 | 18:28 |
| *** karimb has joined #puppet-openstack | 18:28 | |
| *** Blah has quit IRC | 18:39 | |
| *** jpena is now known as jpena|off | 18:41 | |
| EmilienM | mfisch: it would be great if you could patch puppet-keystone with a reno for the big patch we merged today | 19:07 |
| mfisch | sure | 19:08 |
| mfisch | is there a wiki on how to do it? | 19:09 |
| EmilienM | mfisch: http://docs.openstack.org/developer/reno/usage.html | 19:10 |
| *** ayoung has quit IRC | 19:10 | |
| EmilienM | mfisch: pip install reno && reno new <name> | 19:10 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-designate: releasenotes: add notes for Mitaka cycle https://review.openstack.org/293663 | 19:13 |
| *** karimb has quit IRC | 19:26 | |
| openstackgerrit | Merged openstack/puppet-oslo: Rework oslo-messaging-rabbit class https://review.openstack.org/292378 | 19:27 |
| EmilienM | I'm writting release notes | 19:31 |
| EmilienM | you folks work too much, so much patches. | 19:31 |
| mfisch | EmilienM: uploading now | 19:43 |
| openstackgerrit | Matt Fischer proposed openstack/puppet-keystone: Release notes for Keystone external deps https://review.openstack.org/293677 | 19:43 |
| mfisch | ^ | 19:43 |
| EmilienM | mfisch: you rocks | 19:44 |
| mfisch | je suis roche? | 19:44 |
| EmilienM | lol | 19:44 |
| mfisch | je suis une roche? | 19:44 |
| mfisch | un or une | 19:44 |
| mfisch | une because roche ends with e I figure | 19:44 |
| EmilienM | you speak french very well | 19:45 |
| mfisch | about as well as a Trump voter speaks English | 19:46 |
| mfisch | 2nd grade level | 19:46 |
| mfisch | lol | 19:46 |
| *** foxhunt_away is now known as foxhunt72 | 19:48 | |
| EmilienM | mfisch: my IRC bouncer blocked at 15:46:24 - sorry I've missed your words :-) | 19:51 |
| mfisch | Supreme Leader Trump is having me report to re-eductaion | 19:52 |
| openstackgerrit | Merged openstack/puppet-ceph: Change profile to not use $name https://review.openstack.org/291875 | 19:53 |
| openstackgerrit | Merged openstack/puppet-cinder: releasenotes: add notes for Mitaka cycle https://review.openstack.org/293201 | 19:53 |
| *** ericksonsantos has joined #puppet-openstack | 19:58 | |
| colonwq | horray for 291875. | 20:00 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack-integration: neutron: drop depreacted parameters https://review.openstack.org/293690 | 20:06 |
| *** brandon has joined #puppet-openstack | 20:06 | |
| _ody | Can someone please remind me where the list of ouf "big tent" project is documented? | 20:08 |
| openstackgerrit | Merged openstack/puppet-keystone: Release notes for Keystone external deps https://review.openstack.org/293677 | 20:11 |
| openstackgerrit | Keith Schincke proposed openstack/puppet-ceph: Separate Package['ceph'] from ::ceph init https://review.openstack.org/293693 | 20:11 |
| *** ayoung has joined #puppet-openstack | 20:11 | |
| EmilienM | _ody: http://governance.openstack.org/reference/projects/index.html | 20:12 |
| _ody | EmilienM: Thanks. | 20:13 |
| _ody | I tried clicking my way there and kept getting lost, somehow. | 20:13 |
| *** paramite has quit IRC | 20:21 | |
| *** johnmilton has quit IRC | 20:32 | |
| *** imcsk8 has joined #puppet-openstack | 20:42 | |
| *** imcsk8_ has quit IRC | 20:42 | |
| openstackgerrit | Ricardo Rocha proposed openstack/puppet-magnum: Add conductor install and configuration. https://review.openstack.org/275223 | 20:45 |
| *** jayg is now known as jayg|g0n3 | 20:46 | |
| mwhahaha | why is it whenever there's a new puppet release stuff breaks http://logs.openstack.org/79/293579/2/gate/gate-puppet-keystone-puppet-beaker-rspec-dsvm-trusty/1a59fc5/console.html#_2016-03-16_20_43_49_603 | 20:55 |
| EmilienM | mwhahaha: which patch? | 20:57 |
| EmilienM | _ody: ^ | 20:57 |
| mwhahaha | https://review.openstack.org/#/c/293579/2 | 20:57 |
| EmilienM | mhh I'm not sure what is happenning here | 20:58 |
| EmilienM | _ody: experimental puppet4 jobs are ready. | 20:58 |
| EmilienM | if you see NOT_REGISTERED in Zuul, that's "normal", I renamed the jobs to have the Puppet version in scenarios | 20:59 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-glance: releasenotes: add notes for Mitaka cycle https://review.openstack.org/293707 | 21:03 |
| _ody | Doesn | 21:05 |
| _ody | oops. | 21:05 |
| _ody | Doesn't infra run a gem mirror? | 21:05 |
| EmilienM | not yet | 21:05 |
| EmilienM | we are working on it for a year almost | 21:05 |
| xarses | colonwq: I'm around now | 21:08 |
| colonwq | xarses, cool. I am having an issue using ceph::key to add a key to an existing/running cluster. | 21:08 |
| _ody | I don't see anything wrong installing the gem that mwhahaha saw failures on. Might have to chalk it up to rubygems.org. | 21:09 |
| mwhahaha | yea it might have been, just seems suspicious | 21:09 |
| mwhahaha | i rechecked | 21:09 |
| xarses | colonwq: got some more information than that? | 21:10 |
| colonwq | xarses, see the commit message of https://review.openstack.org/#/c/293693/ | 21:10 |
| colonwq | xarses, I should not have to use exec to add a key to a running cluster or re-declare all of the values used to create the cluster. | 21:11 |
| mwhahaha | EmilienM: has rdo switched out the script for keystone WSGI? There seems to be a new file as the one in /usr/lib/cgibin/keystone/admin has a deprecation warning | 21:19 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-gnocchi: releasenotes: add notes for Mitaka cycle https://review.openstack.org/293714 | 21:19 |
| mwhahaha | this was a mitaka change | 21:19 |
| mwhahaha | according to the warning | 21:19 |
| EmilienM | mwhahaha: it should, we run openstack code from trunk | 21:19 |
| EmilienM | ayoung can confirm ^ | 21:19 |
| mwhahaha | just wondering if it's /usr/bin/keystone-wsgi-{admin,main} like on ubuntu | 21:20 |
| ayoung | what is the question? | 21:23 |
| mwhahaha | does the keystone package drop /usr/bin/keystone-wsgi-{admin,main} like on ubuntu (for use with wsgi) | 21:23 |
| ayoung | you run Keystone by using systemctl for httpd on a Fedora and Centos based system | 21:23 |
| mwhahaha | right but the config needs to be pointed to a script | 21:23 |
| ayoung | we put it in /var/www/cgi-bin | 21:24 |
| ayoung | mwhahaha, there is not a start/stop script | 21:24 |
| mwhahaha | right i'm not asking that | 21:24 |
| ayoung | the config is done via files in /etc/httpd/conf.g | 21:25 |
| mwhahaha | the /usr/bin/keystone-wsgi-{admin,main} is apython script that is called via apache | 21:25 |
| ayoung | the config is done via files in /etc/httpd/conf.d | 21:25 |
| mwhahaha | WSGIScriptAlias / "/usr/lib/cgi-bin/keystone/admin" | 21:25 |
| xarses | colonwq: ya, I see the problem we have it elsewhere too | 21:25 |
| ayoung | right, as I said that dgoes in /var/www/cgi-bing | 21:25 |
| mwhahaha | right there's a new script | 21:25 |
| mwhahaha | for mitaka | 21:25 |
| mwhahaha | what script do you put there | 21:25 |
| ayoung | did they chagne the location? | 21:25 |
| mwhahaha | that's what i'm trying to determine | 21:25 |
| mwhahaha | on ubuntu, yes there are new scripts | 21:26 |
| xarses | colonwq: if we move it to it's own class, then we have to keep the $::ceph::ensure ability to be set to purge to un-install ceph | 21:26 |
| xarses | so we either need to propgate that to this new class, or we can probably use ensure_resource | 21:26 |
| ayoung | mwhahaha, /var/www/cgi-bin/keystone/admin from my latest Tripleo deploy | 21:26 |
| colonwq | xarses, Could the Package{} checks in ceph::key be replaced with exec checks to see if the cluster is up? ceph health or ceph status. | 21:27 |
| * mwhahaha bangs head on desk | 21:27 | |
| mwhahaha | https://github.com/openstack/puppet-keystone/blob/master/manifests/params.pp#L34 | 21:27 |
| ayoung | mwhahaha, | 21:27 |
| xarses | colonwq: the Package require isn't even being used to check cluster status, its used to ensue you have the required tools to complete the command | 21:27 |
| mwhahaha | https://github.com/openstack/puppet-keystone/blob/master/manifests/wsgi/apache.pp#L238-L246 | 21:28 |
| ayoung | mwhahaha, http://paste.openstack.org/show/490802/ | 21:28 |
| xarses | colonwq: if you want to check cluster status, which we probably should do, we need to do something else | 21:28 |
| ayoung | mwhahaha, ah...right | 21:28 |
| mwhahaha | http://paste.openstack.org/show/490803/ | 21:28 |
| mwhahaha | see deprecation warning | 21:28 |
| xarses | colonwq: in addition to having the packages easier to ensure in the individual resources. | 21:29 |
| mwhahaha | we are copying in puppet-keystone the wsgi.py that is deprecated | 21:29 |
| mwhahaha | they have new scripts | 21:29 |
| mwhahaha | i want to know where they are on RDO | 21:29 |
| mwhahaha | so i can udpate puppet-keystone with the correct files | 21:29 |
| mwhahaha | :) | 21:29 |
| ayoung | /usr/bin/keystone-wsgi-admin | 21:29 |
| ayoung | /usr/bin/keystone-wsgi-public | 21:29 |
| mwhahaha | k that's what i needed | 21:29 |
| ayoung | I...I don't know who decided that was right | 21:29 |
| ayoung | I never would have put it in bin | 21:29 |
| ayoung | Cowboys | 21:29 |
| mwhahaha | asshattery | 21:29 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-heat: releasenotes: add notes for Mitaka cycle https://review.openstack.org/293718 | 21:29 |
| ayoung | They put a wsgi script in bin? | 21:29 |
| mwhahaha | anyway on ubuntu they are keystone-wsgi-{admin,main} | 21:29 |
| mwhahaha | so we'll just need to copy them correctly based on OS | 21:30 |
| ayoung | blame zigo for that | 21:30 |
| mwhahaha | eh the wsgi.py was in a different spot for both already :D | 21:30 |
| zigo | ayoung: Why would you blame me for something done in a package produced by Canonical ?!? | 21:30 |
| mwhahaha | zigo: i'm using MOS packages | 21:31 |
| * mwhahaha blames zigo | 21:31 | |
| zigo | Or did I miss something? | 21:31 |
| ayoung | zigo, I blame you for most things that are out of your control | 21:31 |
| zigo | Ah... | 21:31 |
| zigo | Right ! :) | 21:31 |
| zigo | Thanks. | 21:31 |
| colonwq | xarses, then we need to be able to check the package and check the cluster status. How can we do this while getting ceph::ensure without needing fsid and the rest | 21:31 |
| zigo | So what do I need to fix? :) | 21:31 |
| mwhahaha | nothing | 21:31 |
| ayoung | zigo, Its OK, people blame me for Keystone decisions I've been stuyck with for years | 21:31 |
| mwhahaha | just arguing over the location of the wsgi scripts | 21:31 |
| zigo | Ah ok. | 21:31 |
| mwhahaha | and names | 21:32 |
| zigo | If that's annoying, I can get it fixed in the way Ubuntu guys do. | 21:32 |
| zigo | Would that help? | 21:32 |
| ayoung | mwhahaha, does not look like it has changed since Kilo on Fedora...http://koji.fedoraproject.org/koji/buildinfo?buildID=715583 | 21:32 |
| *** fvollero has quit IRC | 21:32 | |
| mwhahaha | might be provided by a different package, not sure | 21:33 |
| mwhahaha | just looking into something, i'll figure it out | 21:33 |
| zigo | mwhahaha: FYI, I'm currently packaging both puppet-module-aodh and that Galera module that fuel-library needs... | 21:33 |
| mwhahaha | zigo: did you already package puppet-module-olso? | 21:33 |
| mwhahaha | we'll be needing that soon | 21:34 |
| *** dprince has quit IRC | 21:34 | |
| *** fvollero has joined #puppet-openstack | 21:34 | |
| zigo | mwhahaha: Nop, but I can do it too, if that's needed. | 21:34 |
| mwhahaha | it will be, iberezovskiy is working on updating puppet-nova to use i think | 21:34 |
| zigo | mwhahaha: Compared to 2 months ago, we only have galera, aodh and mysql as new dependencies of fuel-library, right? | 21:34 |
| colonwq | xarses, I'll be back later. I need to relocate. | 21:34 |
| zigo | Or am I missing one? | 21:34 |
| xarses | colonwq: ok | 21:34 |
| *** colonwq has quit IRC | 21:34 | |
| mwhahaha | zigo: did you get rsyslog and mongodb? | 21:35 |
| mwhahaha | https://github.com/openstack/fuel-library/commits/master/deployment/Puppetfile | 21:35 |
| zigo | mwhahaha: I have mongo, but not rsyslog. | 21:35 |
| zigo | Will do it now. | 21:35 |
| zigo | So, on my todo: oslo, rsyslog, galera and aodh. | 21:36 |
| zigo | Anything else? | 21:36 |
| mwhahaha | think that's it | 21:36 |
| mwhahaha | oh no we have a different pacemaker one | 21:36 |
| zigo | mwhahaha: FYI, I have made a "pkgos-puppet2deb" shell script that does most of the work for me! :) | 21:37 |
| *** johnmilton has joined #puppet-openstack | 21:37 | |
| * ayoung still staggered....bin? | 21:37 | |
| mwhahaha | we changed out the pacemaker module for one written by dilyin | 21:38 |
| zigo | mwhahaha: pkgos-puppet2deb -u git://github.com/openstack/puppet-aodh.git -v 8.0.0b1 -n puppet-aodh aodh | 21:38 |
| zigo | Then I just need to basically edit debian/copyright ... :P | 21:38 |
| *** trown is now known as trown|outtypewww | 21:39 | |
| mwhahaha | oh i should have looked more carefully so we're using keystone-wsgi-admin and keystone-wsgi-public as well for the new scripts so i guess that reduces the OS specific differences | 21:41 |
| *** johnmilton has quit IRC | 21:42 | |
| zigo | EmilienM: There's no tag for oslo, could you please do a b1 tag for it? | 21:43 |
| EmilienM | no | 21:43 |
| EmilienM | puppet-oslo won't have a release. | 21:43 |
| EmilienM | since we don't use it yet | 21:43 |
| EmilienM | :) | 21:43 |
| zigo | EmilienM: Never?!? | 21:43 |
| EmilienM | in Newton | 21:43 |
| EmilienM | iberezovskiy is working on it | 21:43 |
| zigo | mwhahaha: So, I don't need to do puppet-oslo then? | 21:43 |
| EmilienM | but won't likely finish in time for Newton. | 21:43 |
| mwhahaha | well then we can't pull it in to any other modules | 21:44 |
| zigo | mwhahaha: I want to package fuel-library for Mitaka, does it use puppet-oslo? | 21:44 |
| mwhahaha | not yet, but it will if iberezovskiy introduces it to something like nova | 21:44 |
| mwhahaha | if we're not going to take oslo for mitaka then we cannot integrate it into any other modules until after mitaka is released | 21:45 |
| EmilienM | mwhahaha: does it need a tag? | 21:47 |
| zigo | mwhahaha: If the answer is "not yet", then I'm not packaging it yet either. | 21:47 |
| EmilienM | to me, it does not make sense to release something that we don't use yet | 21:47 |
| zigo | mwhahaha: I just want to be up-to-date to Fuel for *today* (all components). | 21:47 |
| zigo | EmilienM: But if it gets in use in Nova? | 21:47 |
| mwhahaha | it's not a today dep, but it becomes a compontent the instant it's pulled into the other modules | 21:48 |
| EmilienM | zigo: well, not sure we will merge the patch | 21:48 |
| zigo | Ok. | 21:48 |
| EmilienM | probably after Mitaka release | 21:48 |
| EmilienM | mwhahaha: ^ fyi | 21:48 |
| zigo | Makes sense. | 21:48 |
| EmilienM | we don't face official FF | 21:48 |
| EmilienM | but I don't want big changes like this in the end of cycle. | 21:48 |
| mwhahaha | ok | 21:48 |
| EmilienM | s/face/have/ | 21:48 |
| zigo | mwhahaha: Is this the rsyslog module which we use? https://github.com/saz/puppet-rsyslog | 21:48 |
| mwhahaha | zigo: yes | 21:49 |
| zigo | Ok, doing it. | 21:49 |
| EmilienM | ayoung: what is the use case of using rabbitmq for keystone? | 21:49 |
| ayoung | EmilienM, notifications of creates etc for Audit | 21:49 |
| ayoung | it is all notifications | 21:50 |
| EmilienM | for ceilometer? | 21:50 |
| ayoung | Keystone doesn't read off the queues, just writes | 21:50 |
| ayoung | I think that is the only consumer | 21:50 |
| ayoung | EmilienM, I really want to split notifications from control messages. We can do two vhosts. | 21:51 |
| EmilienM | I don't understand what you're saying | 21:51 |
| *** johnmilton has joined #puppet-openstack | 21:54 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-horizon: releasenotes: add notes for Mitaka cycle https://review.openstack.org/293736 | 22:01 |
| *** dhollinger has quit IRC | 22:04 | |
| EmilienM | mwhahaha: I'm done for today, feel free to review release notes if you have some time, I'll try to finish others tomorrow. So Friday I can submit last patches to modules (README, metadata, etc) | 22:06 |
| EmilienM | I updated https://docs.google.com/spreadsheets/d/1XVrmEiLrJSdxDo-S_vFB7ljxTdYg-pe8hiMUryRor5A/edit#gid=1036795949 | 22:06 |
| mwhahaha | sure | 22:06 |
| EmilienM | but notes are visible here: https://review.openstack.org/#/q/topic:puppet/mitaka | 22:06 |
| *** rahulait has quit IRC | 22:06 | |
| zigo | mwhahaha: galera, aodh and rsyslog puppet packages all done and uploaded to Debian. Let's see how long it takes to the FTP masters to accept them! :) | 22:08 |
| *** imcsk8 has quit IRC | 22:11 | |
| *** imcsk8_ has joined #puppet-openstack | 22:11 | |
| EmilienM | mwhahaha: you going to fix https://bugs.launchpad.net/puppet-keystone/+bug/1558290 ? | 22:13 |
| openstack | Launchpad bug 1558290 in puppet-keystone "wsgi scripts have changed for keystone in mitaka" [Medium,Confirmed] | 22:13 |
| EmilienM | just wondering if it will makes our Mitaka release or not | 22:13 |
| *** ilbot3 has quit IRC | 22:14 | |
| *** mattymo_ has joined #puppet-openstack | 22:15 | |
| EmilienM | if not, I'll see if it's a blocker or not and work on it asap | 22:16 |
| *** mattymo has quit IRC | 22:16 | |
| *** ilbot3 has joined #puppet-openstack | 22:18 | |
| *** baker has quit IRC | 22:18 | |
| mwhahaha | it's not a blocker but i'm goign to work on it tomorrow | 22:19 |
| mwhahaha | the old one will continue to work until at least O according to the warning | 22:19 |
| EmilienM | mwhahaha: perfect, worst case we do it in newton and backport. | 22:20 |
| *** colonwq has joined #puppet-openstack | 22:22 | |
| *** imcsk8_ is now known as imcsk8 | 22:30 | |
| *** v1k0d3n has quit IRC | 22:39 | |
| *** severion has quit IRC | 22:39 | |
| *** v1k0d3n has joined #puppet-openstack | 22:39 | |
| *** severion has joined #puppet-openstack | 22:40 | |
| *** foxhunt72 is now known as foxhunt_away | 22:40 | |
| *** mdorman has quit IRC | 22:42 | |
| *** v1k0d3n_ has joined #puppet-openstack | 22:43 | |
| *** severion has quit IRC | 22:44 | |
| *** severion has joined #puppet-openstack | 22:44 | |
| *** v1k0d3n has quit IRC | 22:44 | |
| *** gfidente has quit IRC | 22:51 | |
| *** johnmilton has quit IRC | 23:03 | |
| *** v1k0d3n_ has quit IRC | 23:04 | |
| *** richm has quit IRC | 23:07 | |
| *** karimb has joined #puppet-openstack | 23:13 | |
| *** baker has joined #puppet-openstack | 23:13 | |
| *** baker_ has joined #puppet-openstack | 23:15 | |
| *** johnmilton has joined #puppet-openstack | 23:16 | |
| *** baker has quit IRC | 23:17 | |
| *** v1k0d3n has joined #puppet-openstack | 23:18 | |
| *** baker_ has quit IRC | 23:19 | |
| *** baker has joined #puppet-openstack | 23:19 | |
| *** zino has quit IRC | 23:39 | |
| *** ducttape_ has joined #puppet-openstack | 23:41 | |
| *** ducttape_ has quit IRC | 23:47 | |
| *** zino has joined #puppet-openstack | 23:57 | |
| *** karimb has quit IRC | 23:58 | |
| *** guessi has joined #puppet-openstack | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!