)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":3009,"name":"Vladimir Kozhukalov","email":"kozhukalov@gmail.com","username":"kozhukalov"},"change_message_id":"aacac2ff243b83edcb4fda6811ba940463e474da","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"cc617cd3_a6ac5020","updated":"2026-03-26 20:15:59.000000000","message":"Frankly, the logic looks a bit fragile. Also it adds more implicit logic. IMO we can go this way as a short term but we could also discuss a more explicit and CRD like way to improve our dependencies framework during PTG.\n\nI don\u0027t remember, maybe you already suggested this earlier. We could probably introduce a CRD let\u0027s say \"Dependency\" and update the kubernetes-entrypoint to lookup for such CR\u0027s. Let\u0027s say we have a template (or similar)\n\n```\napiVersion: osh.openstack.org/v1\nkind: Dependency\nmetadata:\n  name: nova-api\n  labels:\n    application: nova\n    component: db-sync\nspec:\n  requires:\n...\n{%- if .Values.manifests.job_db_sync %}\n    - type: job\n      name: nova-db-sync\n{%- end %}\n...\n```\n\nAnd then this template is rendered together with the nova api deployment where the kubernetes-entrypoint init container waits for this dependency is resolved. Also some dependencies could have skip_if_not_found flag. This way we could throw away the barely understandable template logic of static and dynamic dependencies. Also this would open a way for users to have extra dependencies even not managed by charts. Debugging would be also easier, like you could `kubectl get dependencies`.","commit_id":"26650f0f2606e46ff1fc1629edc99672840fd850"},{"author":{"_account_id":5890,"name":"Doug Goldstein","email":"cardoe@cardoe.com","username":"cardoe"},"change_message_id":"c957c1a3158c43d684c1e3d3b6fb5220e408c5d8","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"fad02d21_707ac544","updated":"2026-04-01 02:45:07.000000000","message":"Recheck","commit_id":"26650f0f2606e46ff1fc1629edc99672840fd850"},{"author":{"_account_id":5890,"name":"Doug Goldstein","email":"cardoe@cardoe.com","username":"cardoe"},"change_message_id":"77d874e7e1da1b5b525ce331bd19f30baf41ade7","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"df55075d_1c111640","updated":"2026-03-30 03:13:28.000000000","message":"Sorry. When I signed in somehow I redirected to this issue when I had been looking at another one.","commit_id":"26650f0f2606e46ff1fc1629edc99672840fd850"},{"author":{"_account_id":5890,"name":"Doug Goldstein","email":"cardoe@cardoe.com","username":"cardoe"},"change_message_id":"9c4c20a0477cd83d04d1655affca69b0fe9a2692","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"77e2e3ce_7d290caa","updated":"2026-03-25 14:01:35.000000000","message":"```\n❯ helm template . | grep -A 3 DEPENDENCY_JOBS\n        - name: DEPENDENCY_JOBS\n          value: nova-db-sync,nova-rabbit-init\n        - name: DEPENDENCY_DAEMONSET\n          value: \"\"\n```\n\nWith this change:\n\n```\n❯ helm template . --set manifests.job_rabbit_init\u003dfalse | grep -A 3 DEPENDENCY_JOBS\n        - name: DEPENDENCY_JOBS\n          value: nova-db-sync\n        - name: DEPENDENCY_DAEMONSET\n          value: \"\"\n```","commit_id":"26650f0f2606e46ff1fc1629edc99672840fd850"},{"author":{"_account_id":34520,"name":"Sergiy Markin","email":"smarkin@mirantis.com","username":"sm515x"},"change_message_id":"8b4284044ee8d0ef0dd1c59ccdda6811decea995","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"9f2b312c_3ef1455e","updated":"2026-03-31 22:15:26.000000000","message":"recheck","commit_id":"26650f0f2606e46ff1fc1629edc99672840fd850"},{"author":{"_account_id":5890,"name":"Doug Goldstein","email":"cardoe@cardoe.com","username":"cardoe"},"change_message_id":"852d28871312ecbc1c0512713a0e5d4b986286d6","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"b89d5309_7ce94a52","in_reply_to":"cc617cd3_a6ac5020","updated":"2026-03-27 00:11:01.000000000","message":"Yeah this would be awesome. We should ask to have that DNS namespace reserved.","commit_id":"26650f0f2606e46ff1fc1629edc99672840fd850"},{"author":{"_account_id":5890,"name":"Doug Goldstein","email":"cardoe@cardoe.com","username":"cardoe"},"change_message_id":"50e09985beaf09e92d0695b2a4e26461b9e26db3","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"b1a72be7_2cb8c76e","updated":"2026-04-01 17:52:16.000000000","message":"Recheck","commit_id":"be235310d1842d21c52a550e2dd5a2c6dd613233"},{"author":{"_account_id":3009,"name":"Vladimir Kozhukalov","email":"kozhukalov@gmail.com","username":"kozhukalov"},"change_message_id":"b5876d83a0877773f8fa5d3a029c9a8d99b44b8b","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"12800372_8b3cb87d","updated":"2026-04-01 20:28:00.000000000","message":"I set this in the nova values to make the cell setup pod wait until all computes are up before discovering them.\n\n```\njobs:\n  cell_setup:\n    extended_wait:\n      enabled: true\n```","commit_id":"1b0415189d0f1cb746e0c77dac9bfd1a917ca49e"},{"author":{"_account_id":3009,"name":"Vladimir Kozhukalov","email":"kozhukalov@gmail.com","username":"kozhukalov"},"change_message_id":"1600079b13a2f71da8504ded6f8633c47610cce3","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"6af06063_b84429d9","updated":"2026-04-01 20:25:41.000000000","message":"Sorry for rebasing. I updated the nova chart together with Ldap fix to make the nova bootstrap job reliable. It has been failing often recently and it is due to cell setup discovered computes too early before they all are up.","commit_id":"1b0415189d0f1cb746e0c77dac9bfd1a917ca49e"}]}