)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"151218db3e8bed1affe9d083dcbf210fbf5f36fa","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"4640300f_ca3fe0ae","updated":"2022-04-25 18:06:48.000000000","message":"Failure of tripleo-ansible-centos-8-molecule-tripleo_cephadm was during instance creation, i.e. before changed code could run.\n\nFAILED - RETRYING: Wait for instance(s) creation to complete (300 retries left).\n\nhttps://13e30d73ec5b0855d4d8-3df34f04e18be629eb587340c626577b.ssl.cf2.rackcdn.com/837998/2/check/tripleo-ansible-centos-8-molecule-tripleo_cephadm/996d370/reports.html\n\n","commit_id":"78e2b6e386ad46fbbdb2d839d3d08bfed2c3530e"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e1f97e5f2e463fc6ae597043aa4961e2531f0183","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"98141e49_360cde57","updated":"2022-04-27 11:11:59.000000000","message":"recheck","commit_id":"78e2b6e386ad46fbbdb2d839d3d08bfed2c3530e"}],"tripleo_ansible/roles/tripleo_cephadm/tasks/pools.yaml":[{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"a3ba74afe4d87569bdb558aa7f68afb4651256d9","unresolved":true,"context_lines":[{"line_number":42,"context_line":""},{"line_number":43,"context_line":"- name: Configure the RBD trash purge scheduler"},{"line_number":44,"context_line":"  when:"},{"line_number":45,"context_line":"    - tripleo_enabled_services | default([]) | intersect([\u0027cinder_volume\u0027])"},{"line_number":46,"context_line":"  block:"},{"line_number":47,"context_line":"    - name: Get the RBD ceph_cli"},{"line_number":48,"context_line":"      include_tasks: ceph_cli.yaml"}],"source_content_type":"text/x-yaml","patch_set":1,"id":"37d790d2_e4562be6","line":45,"range":{"start_line":45,"start_character":33,"end_line":45,"end_character":44},"updated":"2022-04-18 06:32:07.000000000","message":"++ makes sense in this context where that variable might not exist","commit_id":"610d4e7eeb568d5494aa237378384a1ac2049a32"}],"tripleo_ansible/roles/tripleo_cephadm/templates/deployed_ceph.yaml.j2":[{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"a3ba74afe4d87569bdb558aa7f68afb4651256d9","unresolved":true,"context_lines":[{"line_number":10,"context_line":"  ContainerCephDaemonImage: {{ tripleo_cephadm_container_ns }}/{{ tripleo_cephadm_container_image }}:{{ tripleo_cephadm_container_tag | string }}"},{"line_number":11,"context_line":"  CephClusterName: {{ tripleo_cephadm_cluster }}"},{"line_number":12,"context_line":"  {% if tripleo_cephadm_pools -%}"},{"line_number":13,"context_line":"  CephPools: {{ tripleo_cephadm_pools }}"},{"line_number":14,"context_line":"  {% endif -%}"},{"line_number":15,"context_line":"  {% if tripleo_ceph_client_vars -%}"},{"line_number":16,"context_line":"  CephClientConfigVars: {{ tripleo_ceph_client_vars }}"}],"source_content_type":"text/x-jinja2","patch_set":1,"id":"355d3069_2e457ed5","line":13,"range":{"start_line":13,"start_character":2,"end_line":13,"end_character":12},"updated":"2022-04-18 06:32:07.000000000","message":"We need to be careful when the overcloud deploy cmd is built because the order of the redefinition of these variables (especially pools and keys) might cause unexpected behaviors or failures during the overcloud deployment.\n\nMaybe this is just a topic we should highlight in the documentation change","commit_id":"610d4e7eeb568d5494aa237378384a1ac2049a32"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"2da249be33386c2ff81b496457e8c59bd708da7a","unresolved":true,"context_lines":[{"line_number":10,"context_line":"  ContainerCephDaemonImage: {{ tripleo_cephadm_container_ns }}/{{ tripleo_cephadm_container_image }}:{{ tripleo_cephadm_container_tag | string }}"},{"line_number":11,"context_line":"  CephClusterName: {{ tripleo_cephadm_cluster }}"},{"line_number":12,"context_line":"  {% if tripleo_cephadm_pools -%}"},{"line_number":13,"context_line":"  CephPools: {{ tripleo_cephadm_pools }}"},{"line_number":14,"context_line":"  {% endif -%}"},{"line_number":15,"context_line":"  {% if tripleo_ceph_client_vars -%}"},{"line_number":16,"context_line":"  CephClientConfigVars: {{ tripleo_ceph_client_vars }}"}],"source_content_type":"text/x-jinja2","patch_set":1,"id":"bcdcf749_4b25808d","line":13,"range":{"start_line":13,"start_character":2,"end_line":13,"end_character":12},"in_reply_to":"355d3069_2e457ed5","updated":"2022-04-18 14:50:18.000000000","message":"I share your concern here. Let\u0027s talk about the full cycle and see if there would be any issues. \n\n- User follows [1] and vms, volumes, images pools are created. Assume they used the same names as the defaults for CinderRbdPoolName, GlanceRbdPoolName, NovaRbdPoolName because we will them to in the doc. But we can\u0027t enforce this and overrides for CinderRbdPoolName, GlanceRbdPoolName, NovaRbdPoolName are not in this template.\n\n- During overcloud deploy, CephPools is overridden by this line [2]\n\n- CephPools will behave like extra_pools [3]\n\n- We are already encourage overriding CephPools because the PG numbers are rarely all the same [4] so I don\u0027t think that will be a problem. Even if they use it on stack updates to add extra pools.\n\n- Because we overrode the client username and key [5], calls to the ceph_key module  will be idempotent.\n\n- RGW pools are created automatically by RGW and not on the THT layer. Manila pools are handled by a different variable on the THT layer [6].\n\nSo the RBD pools should be OK via CephPools based on my understanding, though we need to test it to be sure.\n\n\n[1] https://review.opendev.org/c/openstack/tripleo-docs/+/838003/2/deploy-guide/source/features/deployed_ceph.rst#979\n\n[2] https://review.opendev.org/c/openstack/tripleo-ansible/+/837998/1/tripleo_ansible/roles/tripleo_cephadm/templates/deployed_ceph.yaml.j2#13\n\n[3] https://github.com/openstack/tripleo-heat-templates/blob/master/deployment/cephadm/ceph-base.yaml#L610\n\n[4] https://docs.openstack.org/project-deploy-guide/tripleo-docs/latest/features/ceph_config.html#customize-ceph-placement-groups-per-openstack-pool\n\n[5] https://github.com/openstack/tripleo-heat-templates/blob/master/deployment/cephadm/ceph-base.yaml#L617-L620\n\n[6] https://github.com/openstack/tripleo-heat-templates/blob/master/deployment/cephadm/ceph-base.yaml#L612-L616","commit_id":"610d4e7eeb568d5494aa237378384a1ac2049a32"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"a165b847bd14350676023b6512f041b450e2d613","unresolved":true,"context_lines":[{"line_number":10,"context_line":"  ContainerCephDaemonImage: {{ tripleo_cephadm_container_ns }}/{{ tripleo_cephadm_container_image }}:{{ tripleo_cephadm_container_tag | string }}"},{"line_number":11,"context_line":"  CephClusterName: {{ tripleo_cephadm_cluster }}"},{"line_number":12,"context_line":"  {% if tripleo_cephadm_pools -%}"},{"line_number":13,"context_line":"  CephPools: {{ tripleo_cephadm_pools }}"},{"line_number":14,"context_line":"  {% endif -%}"},{"line_number":15,"context_line":"  {% if tripleo_ceph_client_vars -%}"},{"line_number":16,"context_line":"  CephClientConfigVars: {{ tripleo_ceph_client_vars }}"}],"source_content_type":"text/x-jinja2","patch_set":1,"id":"814c61f7_8c9ac6d0","line":13,"range":{"start_line":13,"start_character":2,"end_line":13,"end_character":12},"in_reply_to":"bcdcf749_4b25808d","updated":"2022-04-19 08:12:33.000000000","message":"\u003e I share your concern here. Let\u0027s talk about the full cycle and see if there would be any issues. \n\u003e \n\u003e - User follows [1] and vms, volumes, images pools are created. Assume they used the same names as the defaults for CinderRbdPoolName, GlanceRbdPoolName, NovaRbdPoolName because we will them to in the doc. But we can\u0027t enforce this and overrides for CinderRbdPoolName, GlanceRbdPoolName, NovaRbdPoolName are not in this template.\n\u003e \n\u003e - During overcloud deploy, CephPools is overridden by this line [2]\n\u003e \n\u003e - CephPools will behave like extra_pools [3]\n++ good point, so they can be different and added to the list defined in tht or overrides in case of redefinition.\n\n\u003e \n\u003e - We are already encourage overriding CephPools because the PG numbers are rarely all the same [4] so I don\u0027t think that will be a problem. Even if they use it on stack updates to add extra pools.\n\u003e \n\u003e - Because we overrode the client username and key [5], calls to the ceph_key module  will be idempotent.\n\u003e \n\u003e - RGW pools are created automatically by RGW and not on the THT layer. Manila pools are handled by a different variable on the THT layer [6].\n\u003e \n\u003e So the RBD pools should be OK via CephPools based on my understanding, though we need to test it to be sure.\n\u003e \nYeah, I was thinking about running a few tests to make sure the entire flow you described works as expected, but in general I agree with you: we just need to make sure the keys are updated with the right access to the pools.\n\u003e \n\u003e [1] https://review.opendev.org/c/openstack/tripleo-docs/+/838003/2/deploy-guide/source/features/deployed_ceph.rst#979\n\u003e \n\u003e [2] https://review.opendev.org/c/openstack/tripleo-ansible/+/837998/1/tripleo_ansible/roles/tripleo_cephadm/templates/deployed_ceph.yaml.j2#13\n\u003e \n\u003e [3] https://github.com/openstack/tripleo-heat-templates/blob/master/deployment/cephadm/ceph-base.yaml#L610\n\u003e \n\u003e [4] https://docs.openstack.org/project-deploy-guide/tripleo-docs/latest/features/ceph_config.html#customize-ceph-placement-groups-per-openstack-pool\n\u003e \n\u003e [5] https://github.com/openstack/tripleo-heat-templates/blob/master/deployment/cephadm/ceph-base.yaml#L617-L620\n\u003e \n\u003e [6] https://github.com/openstack/tripleo-heat-templates/blob/master/deployment/cephadm/ceph-base.yaml#L612-L616","commit_id":"610d4e7eeb568d5494aa237378384a1ac2049a32"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"ef354034361f134e8dc76401803a1b28fc5c77af","unresolved":true,"context_lines":[{"line_number":16,"context_line":"  CephClientConfigVars: {{ tripleo_ceph_client_vars }}"},{"line_number":17,"context_line":"  {% endif -%}"},{"line_number":18,"context_line":"  {% if ceph_client_username -%}"},{"line_number":19,"context_line":"  CephClientKey: {{ ceph_client_username }}"},{"line_number":20,"context_line":"  {% endif -%}"},{"line_number":21,"context_line":"  {% if ceph_client_key -%}"},{"line_number":22,"context_line":"  CephClientUserName: {{ ceph_client_key }}"},{"line_number":23,"context_line":"  {% endif -%}"},{"line_number":24,"context_line":"  # Recommendation: set the following to false after overcloud is deployed"},{"line_number":25,"context_line":"  ApplyCephConfigOverridesOnUpdate: true"}],"source_content_type":"text/x-jinja2","patch_set":1,"id":"b73b8e49_aa1de9ef","line":22,"range":{"start_line":19,"start_character":0,"end_line":22,"end_character":43},"updated":"2022-04-22 18:37:44.000000000","message":"THT var and Ansible var are crossed and need update. I.e. you would get this:\n\n  CephClientKey: openstack\n  CephClientUserName: AQBNFF9iAAAAABAA/oeWHfB+AmahQM5/jOyzIg\u003d\u003d\n\nInstead you want this:\n\n  CephClientUserName: openstack\n  CephClientKey: AQBNFF9iAAAAABAA/oeWHfB+AmahQM5/jOyzIg\u003d\u003d","commit_id":"610d4e7eeb568d5494aa237378384a1ac2049a32"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"6abed4c55c600d226a1c18eff9a744f81cdad11f","unresolved":false,"context_lines":[{"line_number":16,"context_line":"  CephClientConfigVars: {{ tripleo_ceph_client_vars }}"},{"line_number":17,"context_line":"  {% endif -%}"},{"line_number":18,"context_line":"  {% if ceph_client_username -%}"},{"line_number":19,"context_line":"  CephClientKey: {{ ceph_client_username }}"},{"line_number":20,"context_line":"  {% endif -%}"},{"line_number":21,"context_line":"  {% if ceph_client_key -%}"},{"line_number":22,"context_line":"  CephClientUserName: {{ ceph_client_key }}"},{"line_number":23,"context_line":"  {% endif -%}"},{"line_number":24,"context_line":"  # Recommendation: set the following to false after overcloud is deployed"},{"line_number":25,"context_line":"  ApplyCephConfigOverridesOnUpdate: true"}],"source_content_type":"text/x-jinja2","patch_set":1,"id":"99606aea_744c0b5a","line":22,"range":{"start_line":19,"start_character":0,"end_line":22,"end_character":43},"in_reply_to":"b73b8e49_aa1de9ef","updated":"2022-04-22 18:43:32.000000000","message":"fixed \nhttps://review.opendev.org/c/openstack/tripleo-ansible/+/837998/1..2/tripleo_ansible/roles/tripleo_cephadm/templates/deployed_ceph.yaml.j2","commit_id":"610d4e7eeb568d5494aa237378384a1ac2049a32"}]}