| *** Seb-Solon has quit IRC | 03:37 | |
| *** Seb-Solon has joined #openstack-chef | 03:43 | |
| *** eroux has quit IRC | 06:31 | |
| *** os-chef-bot has quit IRC | 09:33 | |
| *** os-chef-bot has joined #openstack-chef | 09:52 | |
| jklare | sc` hi | 14:00 |
|---|---|---|
| jklare | sc` was just moving though the release patches and wondered if the removal of the if/else block in the Berksfile is intended? | 14:00 |
| jklare | sc` if so, we should probably do this change in master and backport it to stable/pike | 14:02 |
| sc` | jklare: removal of the if/else block was the only way to generate the Berksfile.lock | 14:19 |
| sc` | keeping it in meant berks followed the paths and splatted out a lock file with those local paths | 14:19 |
| sc` | to me, the alternative is to not commit a lock file, letting chef dep solving choose the dependencies | 14:20 |
| sc` | the overall importance of berkshelf is waning, with other chef-native tooling sprouting up since | 14:21 |
| jklare | sc` ah ok, so you generated the Berksfile locally on your machine where these folders exist | 14:22 |
| sc` | when generating the release patches, i pulled all the cookbooks to a single directory and iterated across them. the act of doing that, with the if/else block, gave an invalid lock file to commit | 14:23 |
| jklare | sc` i see | 14:23 |
| jklare | sc` i think we can omit the Berksfile.lock | 14:24 |
| sc` | sounds good to me | 14:24 |
| jklare | sc` but that means we have no real way of checking which was a working set of cookbooks | 14:24 |
| sc` | i have a possible solution for that: policyfile | 14:24 |
| jklare | sc` ? | 14:24 |
| sc` | its intention is to replace berkshelf | 14:25 |
| jklare | but why? | 14:25 |
| jklare | i like berkshelf :) | 14:25 |
| sc` | :) | 14:26 |
| sc` | the output of that policyfile is a more concise way of pinning to cookbooks, which itself includes a lock file that, i believe, is safe for this exact use case | 14:27 |
| sc` | https://github.com/chef/chef-rfc/blob/master/rfc097-policyfile-includes.md | 14:28 |
| sc` | we can keep using berks. that's fine. | 14:29 |
| sc` | as far as i know, it's still supported | 14:30 |
| sc` | removing the lock will mean that the if/else block can live upstream, though | 14:30 |
| sc` | i'll knock out the Berksfile.lock from the release patches, and restore the if/else blocks | 14:31 |
| sc` | from a dev perspective, it's easier to pull in changes that way, as it may be | 14:32 |
| sc` | jklare: when i get supermarket fully populated, i'll push up a change to remove the github callouts for the Berksfiles, so it'll default to supermarket | 15:23 |
| sc` | that may not be for another several weeks | 15:23 |
| jklare | sc` ok | 15:23 |
| sc` | once all the gerrit changes are merged, i'll push 16.x to supermarket | 15:24 |
| sc` | that'll make changing berks easier. the goal is to have github not as a focal point, but just another mirror | 15:25 |
| sc` | i'm still working on the guide. i have a skeleton building with some content | 15:26 |
| sc` | got the locks removed. just running through a series of local rake jobs before pushing the new patchsets | 16:14 |
| sc` | jklare: the lock has to go from openstack-chef-repo as well, to keep the if/else block | 16:17 |
| *** openstack has joined #openstack-chef | 17:15 | |
| *** ChanServ sets mode: +o openstack | 17:15 | |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-block-storage stable/pike: stable/pike release https://review.openstack.org/547487 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-common stable/pike: stable/pike release https://review.openstack.org/547490 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-compute stable/pike: stable/pike release https://review.openstack.org/547491 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-dashboard stable/pike: stable/pike release https://review.openstack.org/547492 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-identity stable/pike: stable/pike release https://review.openstack.org/547493 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-image stable/pike: stable/pike release https://review.openstack.org/547494 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-integration-test master: Run discover_hosts after computes are added, before running Tempest https://review.openstack.org/547679 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-network stable/pike: stable/pike release https://review.openstack.org/547504 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-ops-database stable/pike: stable/pike release https://review.openstack.org/547505 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-ops-messaging stable/pike: stable/pike release https://review.openstack.org/547507 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-orchestration stable/pike: stable/pike release https://review.openstack.org/547508 | 19:54 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-telemetry stable/pike: stable/pike release https://review.openstack.org/547509 | 19:55 |
| openstackgerrit | Samuel Cassiba proposed openstack/openstack-chef-repo stable/pike: stable/pike release https://review.openstack.org/547510 | 19:55 |
| openstackgerrit | Samuel Cassiba proposed openstack/cookbook-openstack-integration-test stable/pike: stable/pike release https://review.openstack.org/547495 | 19:58 |
| sc` | all the patches are up. https://review.openstack.org/547694 needs to go in before the release patches can safely go | 20:38 |
| sc` | the -1s are because of that | 20:38 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!