)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":28048,"name":"Will Szumski","email":"will@stackhpc.com","username":"jovial"},"change_message_id":"d25784cdb8e3e5b62aeee9d73cef6b223177249c","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"aa81d29c_cd212d0a","updated":"2024-02-16 14:58:23.000000000","message":"Thanks for the fix. I guess the slight change in behaviour is that we will no longer remove the docker-registry:\n\nhttps://github.com/openstack/kayobe/blob/24d65075f6130e98dc7c0379f726f03ac9729925/ansible/roles/docker-registry/tasks/deploy.yml#L15\n\nI think the idea there was that it would be cleaned up if you started with enabled and then disabled it. It is currently possible to stop kayobe managing the docker registry by removing the host from the docker-registry group. The way I think about it is that there is a distinction between kayobe managing the service and whether or not it is enabled. Would it make sense to add this to the documentation?","commit_id":"2e57bf6955d9501965823b4ea1038aa4d9ae1ff5"},{"author":{"_account_id":14200,"name":"Maksim Malchuk","email":"maksim.malchuk@gmail.com","username":"mmalchuk"},"change_message_id":"997b8bb53c00bef7350bc87488c8a14a2d8fc7a9","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"a5afea56_d302eeb5","in_reply_to":"1e0ea034_042879ba","updated":"2024-02-28 18:54:39.000000000","message":"Think no, because it would be a new feature, not fix of the issue. No problem to make it as a separate patchset.","commit_id":"2e57bf6955d9501965823b4ea1038aa4d9ae1ff5"},{"author":{"_account_id":28048,"name":"Will Szumski","email":"will@stackhpc.com","username":"jovial"},"change_message_id":"5d7187664b6c93b72dee866e9f59ca29c75fd553","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"6ca9a283_928b8321","in_reply_to":"41bcd5e5_c251e284","updated":"2024-02-23 17:00:00.000000000","message":"Ahh... realised we don\u0027t have seed service destroy (which is probably why it was implemented like that).","commit_id":"2e57bf6955d9501965823b4ea1038aa4d9ae1ff5"},{"author":{"_account_id":28048,"name":"Will Szumski","email":"will@stackhpc.com","username":"jovial"},"change_message_id":"61fc8bb225ee37dbc3d15e7ff46b38926f22890c","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"1e0ea034_042879ba","in_reply_to":"6ca9a283_928b8321","updated":"2024-02-28 10:31:34.000000000","message":"Does it make sense to first add seed service destroy so that we retain the ability to remove the registry?","commit_id":"2e57bf6955d9501965823b4ea1038aa4d9ae1ff5"},{"author":{"_account_id":28048,"name":"Will Szumski","email":"will@stackhpc.com","username":"jovial"},"change_message_id":"dbad9ac01e02e7d59794a92ebac9e8e17cbfbcc7","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"41bcd5e5_c251e284","in_reply_to":"6f7421ea_acb0774d","updated":"2024-02-23 13:05:02.000000000","message":"I see your point about removing it in deploy, but I still think we need to at least flag this change in behaviour in case some was relying on it. Possibly add an upgrade note to run kayobe seed service destroy -t docker-registry -kt none instead of just setting the flag to false. What do you think?","commit_id":"2e57bf6955d9501965823b4ea1038aa4d9ae1ff5"},{"author":{"_account_id":14200,"name":"Maksim Malchuk","email":"maksim.malchuk@gmail.com","username":"mmalchuk"},"change_message_id":"68e04a4738bf8889cd42bf7a644790ef0a4bda04","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"49f11246_2e95be3e","in_reply_to":"91f1ed6b_0726042b","updated":"2024-06-19 14:17:29.000000000","message":"this is the fix for the issue. to cleanup registry there should be command \u0027kayobe seed containers destroy\u0027 or whatever.","commit_id":"2e57bf6955d9501965823b4ea1038aa4d9ae1ff5"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"c7e6d420476bd5de3c3355cbc4cf8d8c56527bf2","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"91f1ed6b_0726042b","in_reply_to":"a5afea56_d302eeb5","updated":"2024-03-07 13:50:10.000000000","message":"As Will has pointed out, the intention here was to remove the container if the flag is set to false.\n\nThis is generally not how Ansible or Kolla Ansible work - they don\u0027t clean things up. But this would be normal in something like Terraform or Puppet.\n\nWould it help if the task name was based on the target state of the container?","commit_id":"2e57bf6955d9501965823b4ea1038aa4d9ae1ff5"},{"author":{"_account_id":14200,"name":"Maksim Malchuk","email":"maksim.malchuk@gmail.com","username":"mmalchuk"},"change_message_id":"1f4a820631c7a9e72fa77e99c6b901705171773d","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"6f7421ea_acb0774d","in_reply_to":"aa81d29c_cd212d0a","updated":"2024-02-16 15:12:47.000000000","message":"Thanks for comment, but I think we can\u0027t rely on the state when user decided to change \u0027switch\u0027 and execute worng action. If registry was enabled, and user decided to destroy registry he should first destroy, then change the configuration and set \u0027switch\u0027 to the False value.\n\nAbout how to manage the service can be added to the documentation but feel free to do this in the separate patchset. What do you think?","commit_id":"2e57bf6955d9501965823b4ea1038aa4d9ae1ff5"},{"author":{"_account_id":28048,"name":"Will Szumski","email":"will@stackhpc.com","username":"jovial"},"change_message_id":"dbd86955ce6bdffe11add454f2a71c1c71bd1c31","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"2ea3cd91_1f4d4252","updated":"2024-06-20 09:20:54.000000000","message":"Now that we have seed service destroy. I think we can land this one if we document the change in behaviour and the need to run:\n\n    kayobe seed service destroy -t docker-registry -kt none\n\nTo remove the registry.","commit_id":"9049b59dbc395738f7a20c7da17b890876b394e2"},{"author":{"_account_id":14200,"name":"Maksim Malchuk","email":"maksim.malchuk@gmail.com","username":"mmalchuk"},"change_message_id":"c440ce308028c395d99a45be845895a165c7b75f","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"aef714fd_7eabe019","updated":"2024-02-23 18:33:31.000000000","message":"yep.","commit_id":"9049b59dbc395738f7a20c7da17b890876b394e2"},{"author":{"_account_id":35263,"name":"Matt Crees","email":"mattc@stackhpc.com","username":"mattcrees"},"change_message_id":"0c2a0fd0bee57c7716e08e32dc83b79e01fd8d68","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":3,"id":"8f7c29f1_7d9b0d24","in_reply_to":"14804d29_fd42e019","updated":"2024-06-27 14:31:41.000000000","message":"+1 from me for this change in behaviour, imo it makes more sense to handle destruction from the new command. This patch will need at least an upgrade note, and preferably also docs included :)","commit_id":"9049b59dbc395738f7a20c7da17b890876b394e2"},{"author":{"_account_id":14200,"name":"Maksim Malchuk","email":"maksim.malchuk@gmail.com","username":"mmalchuk"},"change_message_id":"4fdac91299cb0f969dfb339a4a411d74e97e2a9c","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"14804d29_fd42e019","in_reply_to":"2ea3cd91_1f4d4252","updated":"2024-06-20 09:32:59.000000000","message":"yep. that is what I want to say)","commit_id":"9049b59dbc395738f7a20c7da17b890876b394e2"}]}