)]}'
{"specs/victoria/tripleo-ceph.rst":[{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"b29f27824b88793f06db7840aa6ddc847ec604f9","unresolved":false,"context_lines":[{"line_number":83,"context_line":"- Configure clients quickly, e.g. generate the key in one place,"},{"line_number":84,"context_line":"  without a container, and copy it efficiently"},{"line_number":85,"context_line":"- This is a standalone role which is reusable to configure OpenStack"},{"line_number":86,"context_line":"  against an externally managed Ceph cluster"},{"line_number":87,"context_line":""},{"line_number":88,"context_line":"Why propose openstack/tripleo-ceph as a separate repository instead"},{"line_number":89,"context_line":"of a set of roles within tripleo-ansible?"}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_9b3c1ae2","line":86,"range":{"start_line":86,"start_character":44,"end_line":86,"end_character":44},"updated":"2020-06-03 03:25:52.000000000","message":"expect that the same proposal works for all nodes types and deployments including DCN and edge sites.\nNeed to document that while the proposal does cover Dashboard deployment for core it does not deploy/manage ceph dashboard for the edge sites.\n\nAlso it is good to cover multiple Ceph cluster deployments by one TripelO which is what happens with edge sites as well as multiple Ceph clusters currently supported by TripelO.\nFor example handling scope of sshkeys and rings.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"aa13e6d4a21ec2fc275b5966c1a456aa8b3c64d4","unresolved":false,"context_lines":[{"line_number":83,"context_line":"- Configure clients quickly, e.g. generate the key in one place,"},{"line_number":84,"context_line":"  without a container, and copy it efficiently"},{"line_number":85,"context_line":"- This is a standalone role which is reusable to configure OpenStack"},{"line_number":86,"context_line":"  against an externally managed Ceph cluster"},{"line_number":87,"context_line":""},{"line_number":88,"context_line":"Why propose openstack/tripleo-ceph as a separate repository instead"},{"line_number":89,"context_line":"of a set of roles within tripleo-ansible?"}],"source_content_type":"text/x-rst","patch_set":17,"id":"9f560f44_73cc4562","line":86,"range":{"start_line":86,"start_character":44,"end_line":86,"end_character":44},"in_reply_to":"ff570b3c_9b3c1ae2","updated":"2020-10-01 19:21:56.000000000","message":"\u003e expect that the same proposal works for all nodes types and\n \u003e deployments including DCN and edge sites.\n\nyes, I\u0027ll add that explicitly. \n\n \u003e Need to document that while the proposal does cover Dashboard\n \u003e deployment for core it does not deploy/manage ceph dashboard for\n \u003e the edge sites.\n\nyes, I\u0027ll add that explicitly. \n\n \u003e Also it is good to cover multiple Ceph cluster deployments by one\n \u003e TripelO which is what happens with edge sites as well as multiple\n \u003e Ceph clusters currently supported by TripelO.\n \u003e For example handling scope of sshkeys and rings.\n\nyes, I\u0027ll add that explicitly.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"b29f27824b88793f06db7840aa6ddc847ec604f9","unresolved":false,"context_lines":[{"line_number":85,"context_line":"- This is a standalone role which is reusable to configure OpenStack"},{"line_number":86,"context_line":"  against an externally managed Ceph cluster"},{"line_number":87,"context_line":""},{"line_number":88,"context_line":"Why propose openstack/tripleo-ceph as a separate repository instead"},{"line_number":89,"context_line":"of a set of roles within tripleo-ansible?"},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"This is similar to how openstack/puppet-tripleo used to call"},{"line_number":92,"context_line":"openstack/puppet-ceph and follows the same pattern as"}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_9b425a0e","line":89,"range":{"start_line":88,"start_character":0,"end_line":89,"end_character":41},"updated":"2020-06-03 03:25:52.000000000","message":"How will a node define what Ceph services are needed for a node role?","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"a5816196d9256c1667d7eb5af9d8e5075ad0e6ad","unresolved":false,"context_lines":[{"line_number":85,"context_line":"- This is a standalone role which is reusable to configure OpenStack"},{"line_number":86,"context_line":"  against an externally managed Ceph cluster"},{"line_number":87,"context_line":""},{"line_number":88,"context_line":"Why propose openstack/tripleo-ceph as a separate repository instead"},{"line_number":89,"context_line":"of a set of roles within tripleo-ansible?"},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"This is similar to how openstack/puppet-tripleo used to call"},{"line_number":92,"context_line":"openstack/puppet-ceph and follows the same pattern as"}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_05ed3fb6","line":89,"range":{"start_line":88,"start_character":0,"end_line":89,"end_character":41},"in_reply_to":"ff570b3c_9b425a0e","updated":"2020-06-03 10:41:46.000000000","message":"Roles like \"Controller\" in TripleO have this services defined\nin their environment definition [1], so nothing is going to \nchange here.\nSame for DCN related roles [2] [3] [4].\nAccording to the \u0027ServicesDefault\u0027 we\u0027re supposed to find that\nservice in that node, following the classic TripleO deployment\nflow (as well as in DCN context).\nIMHO the flow here is the same that works with ceph-ansible.\nWe continue to use the same inventory, and from tht, where\nthe variables are generated, we continue triggering a role\nwhich is able to deploy Ceph [5].\nThe difference is that we\u0027re going to include a playbook defined\nin the \u0027tripleo-ceph\u0027 project instead of ceph-ansible.\n\n[1] https://github.com/openstack/tripleo-heat-templates/blob/master/roles/Controller.yaml#L52-L57\n[2] https://github.com/openstack/tripleo-heat-templates/blob/master/roles/DistributedComputeHCI.yaml\n[3] https://github.com/openstack/tripleo-heat-templates/blob/master/roles/DistributedComputeHCIScaleOut.yaml\n[4] https://github.com/openstack/tripleo-heat-templates/blob/master/roles/DistributedComputeScaleOut.yaml\n[5] https://github.com/openstack/tripleo-heat-templates/blob/master/deployment/ceph-ansible/ceph-base.yaml#L653-L655","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"b29f27824b88793f06db7840aa6ddc847ec604f9","unresolved":false,"context_lines":[{"line_number":88,"context_line":"Why propose openstack/tripleo-ceph as a separate repository instead"},{"line_number":89,"context_line":"of a set of roles within tripleo-ansible?"},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"This is similar to how openstack/puppet-tripleo used to call"},{"line_number":92,"context_line":"openstack/puppet-ceph and follows the same pattern as"},{"line_number":93,"context_line":"openstack/tripleo-validations. We also hope it makes things easier for"},{"line_number":94,"context_line":"openstack/tripleo-ansible maintainers."},{"line_number":95,"context_line":""},{"line_number":96,"context_line":""},{"line_number":97,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_fb73967e","line":94,"range":{"start_line":91,"start_character":0,"end_line":94,"end_character":38},"updated":"2020-06-03 03:25:52.000000000","message":"What is release mechanism for openstack/tripleo-ceph release? is it still part of TripleO or a separately released entity?","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"aa13e6d4a21ec2fc275b5966c1a456aa8b3c64d4","unresolved":false,"context_lines":[{"line_number":88,"context_line":"Why propose openstack/tripleo-ceph as a separate repository instead"},{"line_number":89,"context_line":"of a set of roles within tripleo-ansible?"},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"This is similar to how openstack/puppet-tripleo used to call"},{"line_number":92,"context_line":"openstack/puppet-ceph and follows the same pattern as"},{"line_number":93,"context_line":"openstack/tripleo-validations. We also hope it makes things easier for"},{"line_number":94,"context_line":"openstack/tripleo-ansible maintainers."},{"line_number":95,"context_line":""},{"line_number":96,"context_line":""},{"line_number":97,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":17,"id":"9f560f44_d306f185","line":94,"range":{"start_line":91,"start_character":0,"end_line":94,"end_character":38},"in_reply_to":"ff570b3c_fb73967e","updated":"2020-10-01 19:21:56.000000000","message":"Still a part of TripleO but in a separate repository.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"b29f27824b88793f06db7840aa6ddc847ec604f9","unresolved":false,"context_lines":[{"line_number":99,"context_line":""},{"line_number":100,"context_line":"We could ask deployers to do this:"},{"line_number":101,"context_line":""},{"line_number":102,"context_line":"- Deploy hardware with metalsmith"},{"line_number":103,"context_line":"- Use cephadm and orchestrator directly to configure that hardware"},{"line_number":104,"context_line":"  with Ceph and create OpenStack pools accessible by CephX clients"},{"line_number":105,"context_line":"- Use TripleO to configure OpenStack"}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_9b109a02","line":102,"range":{"start_line":102,"start_character":0,"end_line":102,"end_character":33},"updated":"2020-06-03 03:25:52.000000000","message":"isn\u0027t that required for both proposals: original and alternative? And how HW is configured maybe be different depending on HW components and  node role?","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"aa13e6d4a21ec2fc275b5966c1a456aa8b3c64d4","unresolved":false,"context_lines":[{"line_number":99,"context_line":""},{"line_number":100,"context_line":"We could ask deployers to do this:"},{"line_number":101,"context_line":""},{"line_number":102,"context_line":"- Deploy hardware with metalsmith"},{"line_number":103,"context_line":"- Use cephadm and orchestrator directly to configure that hardware"},{"line_number":104,"context_line":"  with Ceph and create OpenStack pools accessible by CephX clients"},{"line_number":105,"context_line":"- Use TripleO to configure OpenStack"}],"source_content_type":"text/x-rst","patch_set":17,"id":"9f560f44_1336a952","line":102,"range":{"start_line":102,"start_character":0,"end_line":102,"end_character":33},"in_reply_to":"ff570b3c_9b109a02","updated":"2020-10-01 19:21:56.000000000","message":"The servers could also be pre-provisioned. Because metalsmith is not necessary I\u0027ll simply change this part to \"deploy hardware\".","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"b29f27824b88793f06db7840aa6ddc847ec604f9","unresolved":false,"context_lines":[{"line_number":128,"context_line":"cluster nodes in order to execute remote commands and deploy the"},{"line_number":129,"context_line":"specified services. This command will need to be installed on one of"},{"line_number":130,"context_line":"the overcloud nodes which will host the composable CephMon service."},{"line_number":131,"context_line":"From the cephadm point of view, that node will be a bootstrap node"},{"line_number":132,"context_line":"on which the Ceph cluster is created."},{"line_number":133,"context_line":""},{"line_number":134,"context_line":"For this reason the Ceph cluster nodes must be SSH accessible and"},{"line_number":135,"context_line":"provide a user with root privileges to perform some tasks."}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_bb10fe5a","line":132,"range":{"start_line":131,"start_character":0,"end_line":132,"end_character":37},"updated":"2020-06-03 03:25:52.000000000","message":"what happens with bootstrap nodes after deployment?\nWhat happens on failure of the bootstrap node?","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"aa13e6d4a21ec2fc275b5966c1a456aa8b3c64d4","unresolved":false,"context_lines":[{"line_number":128,"context_line":"cluster nodes in order to execute remote commands and deploy the"},{"line_number":129,"context_line":"specified services. This command will need to be installed on one of"},{"line_number":130,"context_line":"the overcloud nodes which will host the composable CephMon service."},{"line_number":131,"context_line":"From the cephadm point of view, that node will be a bootstrap node"},{"line_number":132,"context_line":"on which the Ceph cluster is created."},{"line_number":133,"context_line":""},{"line_number":134,"context_line":"For this reason the Ceph cluster nodes must be SSH accessible and"},{"line_number":135,"context_line":"provide a user with root privileges to perform some tasks."}],"source_content_type":"text/x-rst","patch_set":17,"id":"9f560f44_b38eddce","line":132,"range":{"start_line":131,"start_character":0,"end_line":132,"end_character":37},"in_reply_to":"ff570b3c_bb10fe5a","updated":"2020-10-01 19:21:56.000000000","message":"That bootstap node is the first ceph monitor. After deployment it stays in service. A ceph-ansible deployment works the same way. One of the hosts with the CephMon service is used as the first ceph-mon and the ceph cluster is started on it and then other nodes are added.\n\nIf the bootstrap process fails then the deployment stops the same way it does if ceph-ansible has an error.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"b29f27824b88793f06db7840aa6ddc847ec604f9","unresolved":false,"context_lines":[{"line_number":175,"context_line":""},{"line_number":176,"context_line":".. note::"},{"line_number":177,"context_line":""},{"line_number":178,"context_line":"    The Upgrade process for future releases can be subject of slight"},{"line_number":179,"context_line":"    modifications according to the OpenStack requirements."},{"line_number":180,"context_line":""},{"line_number":181,"context_line":""},{"line_number":182,"context_line":"Other End User Impact"}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_5e043015","line":179,"range":{"start_line":178,"start_character":0,"end_line":179,"end_character":58},"updated":"2020-06-03 03:25:52.000000000","message":"assume there is no dependency on original bootstrap node.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"aa13e6d4a21ec2fc275b5966c1a456aa8b3c64d4","unresolved":false,"context_lines":[{"line_number":175,"context_line":""},{"line_number":176,"context_line":".. note::"},{"line_number":177,"context_line":""},{"line_number":178,"context_line":"    The Upgrade process for future releases can be subject of slight"},{"line_number":179,"context_line":"    modifications according to the OpenStack requirements."},{"line_number":180,"context_line":""},{"line_number":181,"context_line":""},{"line_number":182,"context_line":"Other End User Impact"}],"source_content_type":"text/x-rst","patch_set":17,"id":"9f560f44_737285e6","line":179,"range":{"start_line":178,"start_character":0,"end_line":179,"end_character":58},"in_reply_to":"ff570b3c_5e043015","updated":"2020-10-01 19:21:56.000000000","message":"No. It\u0027s just the first ceph mon and ceph mons can be replaced.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"b29f27824b88793f06db7840aa6ddc847ec604f9","unresolved":false,"context_lines":[{"line_number":205,"context_line":"with TripleO at all:"},{"line_number":206,"context_line":""},{"line_number":207,"context_line":"1. Monitor replacement"},{"line_number":208,"context_line":"2. OSDs replacement"},{"line_number":209,"context_line":""},{"line_number":210,"context_line":".. note::"},{"line_number":211,"context_line":""}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_3e197c79","line":208,"range":{"start_line":208,"start_character":0,"end_line":208,"end_character":19},"updated":"2020-06-03 03:25:52.000000000","message":"If it OSD as a service then yes.\nIf is physical device that OSD uses then Ironic is involved as it track health of nodes and any parts replacements.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"aa13e6d4a21ec2fc275b5966c1a456aa8b3c64d4","unresolved":false,"context_lines":[{"line_number":205,"context_line":"with TripleO at all:"},{"line_number":206,"context_line":""},{"line_number":207,"context_line":"1. Monitor replacement"},{"line_number":208,"context_line":"2. OSDs replacement"},{"line_number":209,"context_line":""},{"line_number":210,"context_line":".. note::"},{"line_number":211,"context_line":""}],"source_content_type":"text/x-rst","patch_set":17,"id":"9f560f44_f347b579","line":208,"range":{"start_line":208,"start_character":0,"end_line":208,"end_character":19},"in_reply_to":"ff570b3c_3e197c79","updated":"2020-10-01 19:21:56.000000000","message":"I\u0027ll include this distinction in the next draft.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"b29f27824b88793f06db7840aa6ddc847ec604f9","unresolved":false,"context_lines":[{"line_number":411,"context_line":"  that can be applied to the ceph cluster and are orchestrated by the"},{"line_number":412,"context_line":"  tripleo-ceph project."},{"line_number":413,"context_line":""},{"line_number":414,"context_line":"The two items described above are pretty similar to the test suite that"},{"line_number":415,"context_line":"today is maintained in the TripleO CI, and they can be implemented"},{"line_number":416,"context_line":"reworking the existing scenarios, adding the proper support to the"},{"line_number":417,"context_line":"cephadm deployment model."},{"line_number":418,"context_line":"A WIP patch can be created and submitted with the purpose of testing"},{"line_number":419,"context_line":"and gating the tripleo-ceph project, and, when it becomes stable"},{"line_number":420,"context_line":"enough, the scenario001 will be able to be officially merged."}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_7e9bf4c7","line":417,"range":{"start_line":414,"start_character":0,"end_line":417,"end_character":25},"updated":"2020-06-03 03:25:52.000000000","message":"Let\u0027s cover multiple Ceph clusters deploymentsby single TripleO to be on par of currently supported configurations.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"a5816196d9256c1667d7eb5af9d8e5075ad0e6ad","unresolved":false,"context_lines":[{"line_number":411,"context_line":"  that can be applied to the ceph cluster and are orchestrated by the"},{"line_number":412,"context_line":"  tripleo-ceph project."},{"line_number":413,"context_line":""},{"line_number":414,"context_line":"The two items described above are pretty similar to the test suite that"},{"line_number":415,"context_line":"today is maintained in the TripleO CI, and they can be implemented"},{"line_number":416,"context_line":"reworking the existing scenarios, adding the proper support to the"},{"line_number":417,"context_line":"cephadm deployment model."},{"line_number":418,"context_line":"A WIP patch can be created and submitted with the purpose of testing"},{"line_number":419,"context_line":"and gating the tripleo-ceph project, and, when it becomes stable"},{"line_number":420,"context_line":"enough, the scenario001 will be able to be officially merged."}],"source_content_type":"text/x-rst","patch_set":17,"id":"ff570b3c_85eb0f76","line":417,"range":{"start_line":414,"start_character":0,"end_line":417,"end_character":25},"in_reply_to":"ff570b3c_7e9bf4c7","updated":"2020-06-03 10:41:46.000000000","message":"AFAIK today CI \"covers\" multiple ceph cluster deployment\nconfiguration, and the generated config through [1]\nwhich triggers [2] [3] is executed with scenario001 [4].\nThis should continue to work with cephadm model.\n\n\n[1] https://github.com/openstack/tripleo-heat-templates/blob/master/deployment/ceph-ansible/ceph-base.yaml#L273\n\n[2] https://github.com/openstack/tripleo-ansible/blob/e4d9c5cbb3b911b379d1dbbd95eb9fff08a0ed88/tripleo_ansible/roles/tripleo_ceph_work_dir/tasks/prepare.yml#L73\n\n[3] https://github.com/openstack/tripleo-ansible/blob/64313d1f3534d20b4ba54ad68629c912a1ed2d9c/tripleo_ansible/roles/tripleo_ceph_run_ansible/tasks/main.yml#L94\n\n[4] https://github.com/openstack/tripleo-heat-templates/blob/master/ci/environments/scenario001-standalone.yaml","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"aa13e6d4a21ec2fc275b5966c1a456aa8b3c64d4","unresolved":false,"context_lines":[{"line_number":411,"context_line":"  that can be applied to the ceph cluster and are orchestrated by the"},{"line_number":412,"context_line":"  tripleo-ceph project."},{"line_number":413,"context_line":""},{"line_number":414,"context_line":"The two items described above are pretty similar to the test suite that"},{"line_number":415,"context_line":"today is maintained in the TripleO CI, and they can be implemented"},{"line_number":416,"context_line":"reworking the existing scenarios, adding the proper support to the"},{"line_number":417,"context_line":"cephadm deployment model."},{"line_number":418,"context_line":"A WIP patch can be created and submitted with the purpose of testing"},{"line_number":419,"context_line":"and gating the tripleo-ceph project, and, when it becomes stable"},{"line_number":420,"context_line":"enough, the scenario001 will be able to be officially merged."}],"source_content_type":"text/x-rst","patch_set":17,"id":"9f560f44_b3037d10","line":417,"range":{"start_line":414,"start_character":0,"end_line":417,"end_character":25},"in_reply_to":"ff570b3c_7e9bf4c7","updated":"2020-10-01 19:21:56.000000000","message":"This proposal will not break DCN deployment support. I\u0027ve added an explicit statement about that on line 70 of the next patch set.","commit_id":"2b03fff68612c9a53a2147e56576db25d8fb5757"},{"author":{"_account_id":16643,"name":"Goutham Pacha Ravi","email":"gouthampravi@gmail.com","username":"gouthamr"},"change_message_id":"542eca7429d35e2d19025fd7f14099899af8dab6","unresolved":false,"context_lines":[{"line_number":64,"context_line":"  OpenStack and Ceph"},{"line_number":65,"context_line":"- Configure cephx keyrings and pools for OpenStack on deployed Ceph"},{"line_number":66,"context_line":"  cluster"},{"line_number":67,"context_line":"- Support Ganesha deployment for scenarios where Ceph and CephFS are"},{"line_number":68,"context_line":"  external but Ganesha is TripleO managed"},{"line_number":69,"context_line":"- Support collocation (hyperconvergence) of OpenStack/Ceph containers"},{"line_number":70,"context_line":"  on same host"},{"line_number":71,"context_line":"  - cephadm reconciliation loop must not break OpenStack configuration"}],"source_content_type":"text/x-rst","patch_set":19,"id":"ff570b3c_7bbe6665","line":68,"range":{"start_line":67,"start_character":0,"end_line":68,"end_character":41},"updated":"2020-06-03 03:12:20.000000000","message":"Currently, Ganesha is tripleo managed in both cases - when ceph cluster is external as well as \"internal\"/tripleo-deployed. \nDoes the point here, and line 55 collectively mean that tripleo-ceph will orchestrate ganesha via the ceph orchestrator in case of \"internal\" ceph as well as \"external\" ceph?","commit_id":"e711998169e3d213191c35c37eb80821186c91a0"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"aa13e6d4a21ec2fc275b5966c1a456aa8b3c64d4","unresolved":false,"context_lines":[{"line_number":64,"context_line":"  OpenStack and Ceph"},{"line_number":65,"context_line":"- Configure cephx keyrings and pools for OpenStack on deployed Ceph"},{"line_number":66,"context_line":"  cluster"},{"line_number":67,"context_line":"- Support Ganesha deployment for scenarios where Ceph and CephFS are"},{"line_number":68,"context_line":"  external but Ganesha is TripleO managed"},{"line_number":69,"context_line":"- Support collocation (hyperconvergence) of OpenStack/Ceph containers"},{"line_number":70,"context_line":"  on same host"},{"line_number":71,"context_line":"  - cephadm reconciliation loop must not break OpenStack configuration"}],"source_content_type":"text/x-rst","patch_set":19,"id":"9f560f44_bf1e9f5c","line":68,"range":{"start_line":67,"start_character":0,"end_line":68,"end_character":41},"in_reply_to":"ff570b3c_7bbe6665","updated":"2020-10-01 19:21:56.000000000","message":"I think we should update this spec for Wallaby to move certain problems into separate specs. So we\u0027ll have a separate spec to address the client work and a separate spec for ganesha.\n\nThus, I\u0027m going to remove this statement in the next patch set.","commit_id":"e711998169e3d213191c35c37eb80821186c91a0"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e68af231f5fc597bb1aaa30d43d61570d4de92be","unresolved":false,"context_lines":[{"line_number":71,"context_line":"  - cephadm reconciliation loop must not break OpenStack configuration"},{"line_number":72,"context_line":"  - TripleO configuration updates must not break Ceph configuration"},{"line_number":73,"context_line":""},{"line_number":74,"context_line":"tripleo-ansible/roles/ceph_client goals:"},{"line_number":75,"context_line":""},{"line_number":76,"context_line":"- Configure OpenStack services as clients of an external Ceph cluster"},{"line_number":77,"context_line":"  (in the case of collocation, the ceph cluster is still logically"}],"source_content_type":"text/x-rst","patch_set":19,"id":"9f560f44_62b00c8d","line":74,"range":{"start_line":74,"start_character":0,"end_line":74,"end_character":40},"updated":"2020-08-13 14:51:43.000000000","message":"Can we re-use this?\n\nhttps://github.com/openstack/openstack-ansible-ceph_client","commit_id":"e711998169e3d213191c35c37eb80821186c91a0"}],"specs/victoria/tripleo-cephadm.rst":[{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"0a143be5a5cc9bf3cb8ed69630449e7bc29a03f5","unresolved":false,"context_lines":[{"line_number":239,"context_line":"    * Does this feature require any new library dependencies or code otherwise not"},{"line_number":240,"context_line":"      included in OpenStack? Or does it depend on a specific version of library?"},{"line_number":241,"context_line":""},{"line_number":242,"context_line":"Testing"},{"line_number":243,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":244,"context_line":""},{"line_number":245,"context_line":"..  Please discuss how the change will be tested."}],"source_content_type":"text/x-rst","patch_set":2,"id":"1f493fa4_e9c90926","line":242,"range":{"start_line":242,"start_character":0,"end_line":242,"end_character":7},"updated":"2020-04-29 09:56:28.000000000","message":"This project should be tested against at least on two different scenarios (like we have today in the upstream CI). These scenarios should test different features that can be enabled at day1.\nI\u0027m +1 to maintain the line of testing we have today, adapting the scenario00{1,4} parameters to the new tool.\n\nWe can:\n\n1. create 1 or 2 new scenarios to deploy using cephadm and start gating this project (and the involved tht) using them;\n2. deprecate scenario00{1,4} in favour of the new tool when it will be default in the W cycle.\n\nAnother important aspect could be related to standalone vs multinode. \nAs seen in the past, multinode can help catching issues that are not visible in a standalone environment, but of course the jobs can be improved later and we can start with standalone testing.\n\nGiulio, John, wdyt about testing? Do you have other ideas that can be useful on this section?","commit_id":"456c1d9d2acc0ed42cfa2977c2c1cd03e096f663"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"3b58419b6538dd412f0f60754ff3a851d8a71a22","unresolved":false,"context_lines":[{"line_number":239,"context_line":"    * Does this feature require any new library dependencies or code otherwise not"},{"line_number":240,"context_line":"      included in OpenStack? Or does it depend on a specific version of library?"},{"line_number":241,"context_line":""},{"line_number":242,"context_line":"Testing"},{"line_number":243,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":244,"context_line":""},{"line_number":245,"context_line":"..  Please discuss how the change will be tested."}],"source_content_type":"text/x-rst","patch_set":2,"id":"1f493fa4_92b98a16","line":242,"range":{"start_line":242,"start_character":0,"end_line":242,"end_character":7},"in_reply_to":"1f493fa4_e9c90926","updated":"2020-04-29 11:52:21.000000000","message":"In general I agree. Let\u0027s get this under the Testing header. \n\nI was thinking we could start with a WIP patch which changes scenario001 and depends-on our other changes and if it\u0027s sufficiently stable we could merge it (or revert if there are issues). When we\u0027re then happy with that we could do the same to scenario004. We\u0027d have to make the call at the time if we could merge it for V. We might run ceph-ansible and cephadm jobs concurrently for a little while too.","commit_id":"456c1d9d2acc0ed42cfa2977c2c1cd03e096f663"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"22dc49e3bb487a7225c9de82100817635c8965f5","unresolved":false,"context_lines":[{"line_number":257,"context_line":"References"},{"line_number":258,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":259,"context_line":""},{"line_number":260,"context_line":".. [1] `cephadm \u003chttps://github.com/ceph/ceph-ansible\u003e`_"},{"line_number":261,"context_line":".. [2] `orchestrator \u003chttps://docs.ceph.com/docs/octopus/mgr/orchestrator/\u003e`_"},{"line_number":262,"context_line":".. [3] `ceph-ansible \u003chttps://github.com/ceph/ceph-ansible\u003e`_"},{"line_number":263,"context_line":".. [4] `Provisioning Baremetal Before Overcloud Deploy \u003chttps://docs.openstack.org/project-deploy-guide/tripleo-docs/latest/provisioning/baremetal_provision.html\u003e`_"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1f493fa4_5271c275","line":260,"range":{"start_line":260,"start_character":0,"end_line":260,"end_character":56},"updated":"2020-04-29 11:54:10.000000000","message":"Freudian slip! lol","commit_id":"456c1d9d2acc0ed42cfa2977c2c1cd03e096f663"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"43da2e95b38c36edb7e61f9edf83e0fe1b75b59e","unresolved":false,"context_lines":[{"line_number":225,"context_line":"  Victoria will release with Nautilus support as deployed by"},{"line_number":226,"context_line":"  ceph-ansible just like Ussuri."},{"line_number":227,"context_line":""},{"line_number":228,"context_line":"- OpenStack W: tripleo-cephadm default, tripleo-ansible/roles/ceph_*"},{"line_number":229,"context_line":"  deprecated except ceph_client"},{"line_number":230,"context_line":""},{"line_number":231,"context_line":"Dependencies"},{"line_number":232,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1f493fa4_d704a417","line":229,"range":{"start_line":228,"start_character":2,"end_line":229,"end_character":31},"updated":"2020-04-29 10:48:53.000000000","message":"We should also deprecate and rename [1] in favour of a new set of cephadm templates, where we need to rework the existing template logic to support the new way the ceph cluster can be described [2] [3], and:\n\n1. trigger the ceph_client role which is still available in tripleo-ansible\n\n2. trigger the cephadm roles in this new project for all the remaining config to be applied to the cluster \n\n\n[1] https://github.com/openstack/tripleo-heat-templates/tree/master/deployment/ceph-ansible\n\n[2] https://tracker.ceph.com/issues/44205\n[3] https://docs.ceph.com/docs/master/cephadm/drivegroups/","commit_id":"9c165ef5ce7c6774569e31a2e886d1b1215cd0d7"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":225,"context_line":"  Victoria will release with Nautilus support as deployed by"},{"line_number":226,"context_line":"  ceph-ansible just like Ussuri."},{"line_number":227,"context_line":""},{"line_number":228,"context_line":"- OpenStack W: tripleo-cephadm default, tripleo-ansible/roles/ceph_*"},{"line_number":229,"context_line":"  deprecated except ceph_client"},{"line_number":230,"context_line":""},{"line_number":231,"context_line":"Dependencies"},{"line_number":232,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":3,"id":"ff570b3c_86a7f90b","line":229,"range":{"start_line":228,"start_character":2,"end_line":229,"end_character":31},"in_reply_to":"1f493fa4_b2b7663a","updated":"2020-05-28 02:30:24.000000000","message":"I guess I said deprecate but I meant we\u0027d maintain support for both until we can deprecate the ceph-ansible templates. The next patchset of the spec will state this.","commit_id":"9c165ef5ce7c6774569e31a2e886d1b1215cd0d7"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"3b58419b6538dd412f0f60754ff3a851d8a71a22","unresolved":false,"context_lines":[{"line_number":225,"context_line":"  Victoria will release with Nautilus support as deployed by"},{"line_number":226,"context_line":"  ceph-ansible just like Ussuri."},{"line_number":227,"context_line":""},{"line_number":228,"context_line":"- OpenStack W: tripleo-cephadm default, tripleo-ansible/roles/ceph_*"},{"line_number":229,"context_line":"  deprecated except ceph_client"},{"line_number":230,"context_line":""},{"line_number":231,"context_line":"Dependencies"},{"line_number":232,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1f493fa4_b2b7663a","line":229,"range":{"start_line":228,"start_character":2,"end_line":229,"end_character":31},"in_reply_to":"1f493fa4_d704a417","updated":"2020-04-29 11:52:21.000000000","message":"I agree. We should create a Header under Implementation (\u003d\u003d\u003d\u003d) on line 190 with these under it\u0027s own sub header (-----) called Heat.","commit_id":"9c165ef5ce7c6774569e31a2e886d1b1215cd0d7"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"058809aa6e4f6450886d33b70fccb8ab824b87e2","unresolved":false,"context_lines":[{"line_number":219,"context_line":"deployment stages, generate a yaml file that represents the ceph"},{"line_number":220,"context_line":"cluster definition according to the cephadm spec [6]_ and:"},{"line_number":221,"context_line":""},{"line_number":222,"context_line":"- trigger the `ceph_client` role which will be still available in"},{"line_number":223,"context_line":"  tripleo-ansible"},{"line_number":224,"context_line":"- trigger the cephadm roles in this new project for all the remaining"},{"line_number":225,"context_line":"  config that should be applied to the cluster [7]_"},{"line_number":226,"context_line":""}],"source_content_type":"text/x-rst","patch_set":6,"id":"1f493fa4_17dce344","line":223,"range":{"start_line":222,"start_character":2,"end_line":223,"end_character":17},"updated":"2020-05-05 11:47:03.000000000","message":"John, do you think could be useful having a separated (-----) section under Implementation to describe the ceph_client role that should be added in tripleo-ansible (and mentioning that the remaining roles will be deprecated) ?","commit_id":"f53dc177c72853f43ec64c4871a1851705738410"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"dba629e44f31658b61a053c3517cbb7eabc3e6ce","unresolved":false,"context_lines":[{"line_number":257,"context_line":"- OpenStack W: tripleo-cephadm default, tripleo-ansible/roles/ceph_*"},{"line_number":258,"context_line":"  deprecated except ceph_client"},{"line_number":259,"context_line":""},{"line_number":260,"context_line":"Dependencies"},{"line_number":261,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":262,"context_line":""},{"line_number":263,"context_line":"..  * Include specific references to specs and/or blueprints in tripleo, or in other"}],"source_content_type":"text/x-rst","patch_set":6,"id":"1f493fa4_378e3a27","line":260,"range":{"start_line":260,"start_character":0,"end_line":260,"end_character":12},"updated":"2020-05-05 14:51:21.000000000","message":"I\u0027d like to add here as a dependency the cephadm tool itself and the related tracker that allow us to deploy the ceph cluster in a declarative way:\n\n1. https://tracker.ceph.com/issues/44205\n2. https://tracker.ceph.com/issues/44873\n\nBefore drafting this section I\u0027ll look for more trackers I\u0027m not aware of (or useful to improve this section).","commit_id":"f53dc177c72853f43ec64c4871a1851705738410"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"38a9e4db8ce75f2f83c4fad550e1cc530144cf9e","unresolved":false,"context_lines":[{"line_number":270,"context_line":"- gfidente"},{"line_number":271,"context_line":"- jmolmo"},{"line_number":272,"context_line":""},{"line_number":273,"context_line":"Work Items"},{"line_number":274,"context_line":"----------"},{"line_number":275,"context_line":""},{"line_number":276,"context_line":"Proposed Schedule"}],"source_content_type":"text/x-rst","patch_set":10,"id":"ff570b3c_e29ee852","line":273,"range":{"start_line":273,"start_character":0,"end_line":273,"end_character":10},"updated":"2020-05-11 12:50:43.000000000","message":"Since we\u0027re putting the involved components under \"implementation\", should we remove this?","commit_id":"21f6ad4e4dfcfb6b2a3ee7c0c8494a68f27e46fa"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"d497ea40809ffcfe9103c895727d527faa72ed0f","unresolved":false,"context_lines":[{"line_number":55,"context_line":"- For day2 Ceph operations use cephadm, orchestrator, or Dashboard"},{"line_number":56,"context_line":"  directly unless the day2 Ceph change involves adding hardware and"},{"line_number":57,"context_line":"  other exceptions as needed"},{"line_number":58,"context_line":"- TripleO stack updates do not trigger tripleo-cephadm by default"},{"line_number":59,"context_line":"- Provide an opinionated Ceph installation based on parameters from"},{"line_number":60,"context_line":"  TripleO and Ironic"},{"line_number":61,"context_line":"- Provide Ceph integration but maximize orthogonality between"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_8c75dafc","line":58,"updated":"2020-05-20 16:18:37.000000000","message":"this might be necessary for scale up/down though","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":55,"context_line":"- For day2 Ceph operations use cephadm, orchestrator, or Dashboard"},{"line_number":56,"context_line":"  directly unless the day2 Ceph change involves adding hardware and"},{"line_number":57,"context_line":"  other exceptions as needed"},{"line_number":58,"context_line":"- TripleO stack updates do not trigger tripleo-cephadm by default"},{"line_number":59,"context_line":"- Provide an opinionated Ceph installation based on parameters from"},{"line_number":60,"context_line":"  TripleO and Ironic"},{"line_number":61,"context_line":"- Provide Ceph integration but maximize orthogonality between"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_90ce13b2","line":58,"in_reply_to":"ff570b3c_8c75dafc","updated":"2020-05-28 02:30:24.000000000","message":"Done","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"d497ea40809ffcfe9103c895727d527faa72ed0f","unresolved":false,"context_lines":[{"line_number":65,"context_line":"- Support collocation (hyperconvergence) of OpenStack/Ceph containers"},{"line_number":66,"context_line":"  on same host"},{"line_number":67,"context_line":"  - cephadm reconciliation loop must not break OpenStack configuration"},{"line_number":68,"context_line":"  - TripleO configuration updates must not break Ceph configuration"},{"line_number":69,"context_line":""},{"line_number":70,"context_line":"tripleo-ansible/roles/ceph_client goals:"},{"line_number":71,"context_line":""}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_6c4966b5","line":68,"updated":"2020-05-20 16:18:37.000000000","message":"we might have to add here ganesha deployment, for the scenario where ceph and cephfs are external but ganesha is tripleo managed","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":65,"context_line":"- Support collocation (hyperconvergence) of OpenStack/Ceph containers"},{"line_number":66,"context_line":"  on same host"},{"line_number":67,"context_line":"  - cephadm reconciliation loop must not break OpenStack configuration"},{"line_number":68,"context_line":"  - TripleO configuration updates must not break Ceph configuration"},{"line_number":69,"context_line":""},{"line_number":70,"context_line":"tripleo-ansible/roles/ceph_client goals:"},{"line_number":71,"context_line":""}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_10e1632b","line":68,"in_reply_to":"ff570b3c_6c4966b5","updated":"2020-05-28 02:30:24.000000000","message":"Done","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"d497ea40809ffcfe9103c895727d527faa72ed0f","unresolved":false,"context_lines":[{"line_number":77,"context_line":"- Full multiclient support, e.g. one OpenStack deployment may use"},{"line_number":78,"context_line":"  multiple Ceph clusters, e.g. multibackend Glance"},{"line_number":79,"context_line":"- Configure clients quickly, e.g. generate the key in one place,"},{"line_number":80,"context_line":"  without a container, and copy it efficiently"},{"line_number":81,"context_line":""},{"line_number":82,"context_line":"Why propose openstack/tripleo-cephadm as a separate repository instead"},{"line_number":83,"context_line":"of a set of roles within tripleo-ansible?"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_4c58e286","line":80,"updated":"2020-05-20 16:18:37.000000000","message":"should also be standalone; reusable to configure openstack against an externally managed ceph cluster","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":77,"context_line":"- Full multiclient support, e.g. one OpenStack deployment may use"},{"line_number":78,"context_line":"  multiple Ceph clusters, e.g. multibackend Glance"},{"line_number":79,"context_line":"- Configure clients quickly, e.g. generate the key in one place,"},{"line_number":80,"context_line":"  without a container, and copy it efficiently"},{"line_number":81,"context_line":""},{"line_number":82,"context_line":"Why propose openstack/tripleo-cephadm as a separate repository instead"},{"line_number":83,"context_line":"of a set of roles within tripleo-ansible?"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_b0f23705","line":80,"in_reply_to":"ff570b3c_4c58e286","updated":"2020-05-28 02:30:24.000000000","message":"Done","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"d497ea40809ffcfe9103c895727d527faa72ed0f","unresolved":false,"context_lines":[{"line_number":115,"context_line":"project with a smaller scope."},{"line_number":116,"context_line":""},{"line_number":117,"context_line":""},{"line_number":118,"context_line":"Security Impact"},{"line_number":119,"context_line":"---------------"},{"line_number":120,"context_line":""},{"line_number":121,"context_line":"Cephadm is an imperative tool and requires ssh access to the Ceph"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_8c915af4","line":118,"updated":"2020-05-20 16:18:37.000000000","message":"it might useful to mention here that cephadm apparently needs to be installed on one of the bootstrap nodes","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":115,"context_line":"project with a smaller scope."},{"line_number":116,"context_line":""},{"line_number":117,"context_line":""},{"line_number":118,"context_line":"Security Impact"},{"line_number":119,"context_line":"---------------"},{"line_number":120,"context_line":""},{"line_number":121,"context_line":"Cephadm is an imperative tool and requires ssh access to the Ceph"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_90a7330c","line":118,"in_reply_to":"ff570b3c_8c915af4","updated":"2020-05-28 02:30:24.000000000","message":"Done","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"d497ea40809ffcfe9103c895727d527faa72ed0f","unresolved":false,"context_lines":[{"line_number":198,"context_line":"    Even though cephadm standalone supports all the commands required to"},{"line_number":199,"context_line":"    the other day2 operations, tripleo-cephadm continues to manage and"},{"line_number":200,"context_line":"    orchestrate all the other actions that can be taken by an operator"},{"line_number":201,"context_line":"    (e.g., nodes scale-up and scale-down)"},{"line_number":202,"context_line":""},{"line_number":203,"context_line":"Performance Impact"},{"line_number":204,"context_line":"------------------"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_f1942e8d","line":201,"updated":"2020-05-20 16:18:37.000000000","message":"we don\u0027t know how though, I guess this is a section where feedback from PTG will be very useful/important","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":198,"context_line":"    Even though cephadm standalone supports all the commands required to"},{"line_number":199,"context_line":"    the other day2 operations, tripleo-cephadm continues to manage and"},{"line_number":200,"context_line":"    orchestrate all the other actions that can be taken by an operator"},{"line_number":201,"context_line":"    (e.g., nodes scale-up and scale-down)"},{"line_number":202,"context_line":""},{"line_number":203,"context_line":"Performance Impact"},{"line_number":204,"context_line":"------------------"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_b05997ff","line":201,"in_reply_to":"ff570b3c_f1942e8d","updated":"2020-05-28 02:30:24.000000000","message":"I changed the language a little based on what we expect. I agree with that you\u0027re saying and we\u0027ll see what feedback we get here.","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"a016cd207a8d5da0b207c4952d14c7c1c66763b7","unresolved":false,"context_lines":[{"line_number":228,"context_line":"will be triggered and we can see performances improved since no additional"},{"line_number":229,"context_line":"layers are involved here."},{"line_number":230,"context_line":""},{"line_number":231,"context_line":"Other Deployer Impact"},{"line_number":232,"context_line":"---------------------"},{"line_number":233,"context_line":""},{"line_number":234,"context_line":".. Discuss things that will affect how you deploy and configure OpenStack"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_8208652c","line":231,"range":{"start_line":231,"start_character":0,"end_line":231,"end_character":21},"updated":"2020-05-15 10:11:55.000000000","message":"Everything start from tht: you should include the cephadm templates which include the role from this new project.\nThe workflow doesn\u0027t change, so this item will be mostly unchanged.\nJohn/Giulio, do you see other impacts we should mention here?","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":228,"context_line":"will be triggered and we can see performances improved since no additional"},{"line_number":229,"context_line":"layers are involved here."},{"line_number":230,"context_line":""},{"line_number":231,"context_line":"Other Deployer Impact"},{"line_number":232,"context_line":"---------------------"},{"line_number":233,"context_line":""},{"line_number":234,"context_line":".. Discuss things that will affect how you deploy and configure OpenStack"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_a17dc76f","line":231,"range":{"start_line":231,"start_character":0,"end_line":231,"end_character":21},"in_reply_to":"ff570b3c_8208652c","updated":"2020-05-28 02:30:24.000000000","message":"No, I think that\u0027s accurate. I\u0027ll add some language there about what will change in the next patchset.","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"a016cd207a8d5da0b207c4952d14c7c1c66763b7","unresolved":false,"context_lines":[{"line_number":241,"context_line":"   * Is this a change that takes immediate effect after its merged, or is it"},{"line_number":242,"context_line":"     something that has to be explicitly enabled?"},{"line_number":243,"context_line":""},{"line_number":244,"context_line":"Developer Impact"},{"line_number":245,"context_line":"----------------"},{"line_number":246,"context_line":""},{"line_number":247,"context_line":".. Discuss things that will affect other developers working on OpenStack."}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_5d3112ec","line":244,"range":{"start_line":244,"start_character":0,"end_line":244,"end_character":16},"updated":"2020-05-15 10:11:55.000000000","message":"Same here, if delevelopers want to add ceph on their overcloud, they should just include the tht cephadm templates, which is the same flow we have for tripleo-ansible.\nJohn/Giulio, am I missing other parts in the dev workflow that should be impacted?","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":241,"context_line":"   * Is this a change that takes immediate effect after its merged, or is it"},{"line_number":242,"context_line":"     something that has to be explicitly enabled?"},{"line_number":243,"context_line":""},{"line_number":244,"context_line":"Developer Impact"},{"line_number":245,"context_line":"----------------"},{"line_number":246,"context_line":""},{"line_number":247,"context_line":".. Discuss things that will affect other developers working on OpenStack."}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_01e1132b","line":244,"range":{"start_line":244,"start_character":0,"end_line":244,"end_character":16},"in_reply_to":"ff570b3c_5d3112ec","updated":"2020-05-28 02:30:24.000000000","message":"Yes, they\u0027ll need to switch to the new templates. In theory both templates should install the same ceph key and conf files for the clients. I\u0027ll add that in the next update.","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"11c5f87274d1b1e546719a173e0f5ad4f31d2bd7","unresolved":false,"context_lines":[{"line_number":253,"context_line":"In general we will defer as much as possible to the following pattern"},{"line_number":254,"context_line":"(though the order might vary for scaling up or down)."},{"line_number":255,"context_line":""},{"line_number":256,"context_line":"- Provision Baremetal [4]_"},{"line_number":257,"context_line":"- Ceph orchestrator [2]_"},{"line_number":258,"context_line":"- Stack updates"},{"line_number":259,"context_line":""},{"line_number":260,"context_line":"In the past stack updates did everything, but the split for Metalsmith"},{"line_number":261,"context_line":"(described in Provisioning Baremetal Before Overcloud Deploy [4]_)"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_d0a2569a","line":258,"range":{"start_line":256,"start_character":0,"end_line":258,"end_character":15},"updated":"2020-05-13 13:15:49.000000000","message":"we need to make sure the storage and storage management networks are configured (including any firewall changes) before using a ceph tool which expects them to be there. maybe we can use ansible tags.","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":253,"context_line":"In general we will defer as much as possible to the following pattern"},{"line_number":254,"context_line":"(though the order might vary for scaling up or down)."},{"line_number":255,"context_line":""},{"line_number":256,"context_line":"- Provision Baremetal [4]_"},{"line_number":257,"context_line":"- Ceph orchestrator [2]_"},{"line_number":258,"context_line":"- Stack updates"},{"line_number":259,"context_line":""},{"line_number":260,"context_line":"In the past stack updates did everything, but the split for Metalsmith"},{"line_number":261,"context_line":"(described in Provisioning Baremetal Before Overcloud Deploy [4]_)"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_cd1e3473","line":258,"range":{"start_line":256,"start_character":0,"end_line":258,"end_character":15},"in_reply_to":"ff570b3c_d0a2569a","updated":"2020-05-28 02:30:24.000000000","message":"My next update to the spec will contain this idea based on Francesco\u0027s POC.","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"d497ea40809ffcfe9103c895727d527faa72ed0f","unresolved":false,"context_lines":[{"line_number":291,"context_line":"represents the use case supported only for the deployment stage; the"},{"line_number":292,"context_line":"stack update will have no effect on the cluster because the day2 can"},{"line_number":293,"context_line":"be done using Ceph\u0027s native tools, with no interaction between"},{"line_number":294,"context_line":"operators and TripleO."},{"line_number":295,"context_line":""},{"line_number":296,"context_line":""},{"line_number":297,"context_line":"Assignee(s)"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_b1f476f2","line":294,"updated":"2020-05-20 16:18:37.000000000","message":"do we want to mention explicitly that for at least one release we\u0027re aiming at having intree support for both ceph-ansible and cephadm?","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":291,"context_line":"represents the use case supported only for the deployment stage; the"},{"line_number":292,"context_line":"stack update will have no effect on the cluster because the day2 can"},{"line_number":293,"context_line":"be done using Ceph\u0027s native tools, with no interaction between"},{"line_number":294,"context_line":"operators and TripleO."},{"line_number":295,"context_line":""},{"line_number":296,"context_line":""},{"line_number":297,"context_line":"Assignee(s)"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_669aa5c5","line":294,"in_reply_to":"ff570b3c_b1f476f2","updated":"2020-05-28 02:30:24.000000000","message":"yes, I will work that into an update to the spec","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"d497ea40809ffcfe9103c895727d527faa72ed0f","unresolved":false,"context_lines":[{"line_number":382,"context_line":"References"},{"line_number":383,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":384,"context_line":""},{"line_number":385,"context_line":".. [1] `cephadm \u003chttps://github.com/ceph/ceph/tree/master/src/cephadm\u003e`_"},{"line_number":386,"context_line":".. [2] `orchestrator \u003chttps://docs.ceph.com/docs/octopus/mgr/orchestrator/\u003e`_"},{"line_number":387,"context_line":".. [3] `ceph-ansible \u003chttps://github.com/ceph/ceph-ansible\u003e`_"},{"line_number":388,"context_line":".. [4] `Provisioning Baremetal Before Overcloud Deploy \u003chttps://docs.openstack.org/project-deploy-guide/tripleo-docs/latest/provisioning/baremetal_provision.html\u003e`_"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_b11f596a","line":385,"updated":"2020-05-20 16:18:37.000000000","message":"maybe we want this to link to cephadm docs vs code","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":382,"context_line":"References"},{"line_number":383,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":384,"context_line":""},{"line_number":385,"context_line":".. [1] `cephadm \u003chttps://github.com/ceph/ceph/tree/master/src/cephadm\u003e`_"},{"line_number":386,"context_line":".. [2] `orchestrator \u003chttps://docs.ceph.com/docs/octopus/mgr/orchestrator/\u003e`_"},{"line_number":387,"context_line":".. [3] `ceph-ansible \u003chttps://github.com/ceph/ceph-ansible\u003e`_"},{"line_number":388,"context_line":".. [4] `Provisioning Baremetal Before Overcloud Deploy \u003chttps://docs.openstack.org/project-deploy-guide/tripleo-docs/latest/provisioning/baremetal_provision.html\u003e`_"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_4d07c4ca","line":385,"in_reply_to":"ff570b3c_b11f596a","updated":"2020-05-28 02:30:24.000000000","message":"Done","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":14270,"name":"Dimitri Savineau","email":"dsavinea@redhat.com","username":"dsavineau"},"change_message_id":"c3242f66732be072addfd0a15bf5cbf5da162bae","unresolved":false,"context_lines":[{"line_number":390,"context_line":".. [6] `cephadm config.yml spec \u003chttps://tracker.ceph.com/issues/44205\u003e`_"},{"line_number":391,"context_line":".. [7] `cephadm bootstrap \u003chttps://tracker.ceph.com/issues/44873\u003e`_"},{"line_number":392,"context_line":".. [8] `cephadm new host \u003chttps://docs.ceph.com/docs/master/cephadm/install/#add-hosts-to-the-cluster\u003e`_"},{"line_number":393,"context_line":".. [9] `cephadn cluster upgrade \u003chttps://docs.ceph.com/docs/master/cephadm/upgrade\u003e`_"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_2b460351","line":393,"range":{"start_line":393,"start_character":8,"end_line":393,"end_character":15},"updated":"2020-05-22 16:17:36.000000000","message":"s/cephadn/cephadm/","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e97baf23897732d32af4a2132f2d61355c033472","unresolved":false,"context_lines":[{"line_number":390,"context_line":".. [6] `cephadm config.yml spec \u003chttps://tracker.ceph.com/issues/44205\u003e`_"},{"line_number":391,"context_line":".. [7] `cephadm bootstrap \u003chttps://tracker.ceph.com/issues/44873\u003e`_"},{"line_number":392,"context_line":".. [8] `cephadm new host \u003chttps://docs.ceph.com/docs/master/cephadm/install/#add-hosts-to-the-cluster\u003e`_"},{"line_number":393,"context_line":".. [9] `cephadn cluster upgrade \u003chttps://docs.ceph.com/docs/master/cephadm/upgrade\u003e`_"}],"source_content_type":"text/x-rst","patch_set":13,"id":"ff570b3c_7a19a086","line":393,"range":{"start_line":393,"start_character":8,"end_line":393,"end_character":15},"in_reply_to":"ff570b3c_2b460351","updated":"2020-05-28 02:30:24.000000000","message":"Done","commit_id":"4f369247f2d4959a9b7da7c7a90d2ec19ccc3dae"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"880a61710f2f990c3c7388892199f417bc87100b","unresolved":false,"context_lines":[{"line_number":41,"context_line":"Overview"},{"line_number":42,"context_line":"--------"},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"Create a new OpenStack repository called tripleo-cephadm for Ceph"},{"line_number":45,"context_line":"server deployment and some Ceph server configuration and introduce a"},{"line_number":46,"context_line":"new role into tripleo-ansible for Ceph client configuration. These two"},{"line_number":47,"context_line":"projects will have the following goals."}],"source_content_type":"text/x-rst","patch_set":15,"id":"ff570b3c_ada2d598","line":44,"updated":"2020-05-28 11:44:55.000000000","message":"maybe we could name this tripleo-ceph given the amount of tasks in there which will be operating directly against the ceph orchestrator and the other ceph tools (not cephadm)\n\nnot sure what other people think about it though because the bootstrap itself is still happening using cephadm; just leaving it here for discussion","commit_id":"a2163c70390d87b13b054fadad107820949e33ef"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"d2c93cb8f1821f8968495fd345edbf69e8986eee","unresolved":false,"context_lines":[{"line_number":41,"context_line":"Overview"},{"line_number":42,"context_line":"--------"},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"Create a new OpenStack repository called tripleo-cephadm for Ceph"},{"line_number":45,"context_line":"server deployment and some Ceph server configuration and introduce a"},{"line_number":46,"context_line":"new role into tripleo-ansible for Ceph client configuration. These two"},{"line_number":47,"context_line":"projects will have the following goals."}],"source_content_type":"text/x-rst","patch_set":15,"id":"ff570b3c_8d50b160","line":44,"in_reply_to":"ff570b3c_ada2d598","updated":"2020-05-28 11:57:48.000000000","message":"OK, I changed the name since I agree it\u0027s going to be more than just tripleo-cephadm.","commit_id":"a2163c70390d87b13b054fadad107820949e33ef"}],"specs/wallaby/tripleo-ceph-client.rst":[{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"66960f5ff22b5e8b24d3e9f10d8fd7e827be0787","unresolved":false,"context_lines":[{"line_number":1,"context_line":".."},{"line_number":2,"context_line":" This work is licensed under a Creative Commons Attribution 3.0 Unported"},{"line_number":3,"context_line":" License."},{"line_number":4,"context_line":""}],"source_content_type":"text/x-rst","patch_set":25,"id":"3f65232a_69c2c3eb","side":"PARENT","line":1,"updated":"2020-10-21 21:48:44.000000000","message":"we shouldn\u0027t be deleting the other spec :D","commit_id":"5dd223fb2d1f1cdc1460324b465a1fdd71e9346e"}],"specs/wallaby/tripleo-ceph.rst":[{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"7b1f64ddd42c087caf6e760c9182732fdabd4513","unresolved":false,"context_lines":[{"line_number":72,"context_line":"  - cephadm reconciliation loop must not break OpenStack configuration"},{"line_number":73,"context_line":"  - TripleO configuration updates must not break Ceph configuration"},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"tripleo-ansible/roles/ceph_client goals:"},{"line_number":76,"context_line":""},{"line_number":77,"context_line":"- Configure OpenStack services as clients of an external Ceph cluster"},{"line_number":78,"context_line":"  (in the case of collocation, the ceph cluster is still logically"},{"line_number":79,"context_line":"  external)"},{"line_number":80,"context_line":"- Provide Ceph configuration files and cephx keys for OpenStack"},{"line_number":81,"context_line":"  clients of RBD, RGW, CephFS (Nova, Cinder, Glance, Manila)"},{"line_number":82,"context_line":"- Full multiclient support, e.g. one OpenStack deployment may use"},{"line_number":83,"context_line":"  multiple Ceph clusters, e.g. multibackend Glance"},{"line_number":84,"context_line":"- Configure clients quickly, e.g. generate the key in one place,"},{"line_number":85,"context_line":"  without a container, and copy it efficiently"},{"line_number":86,"context_line":"- This is a standalone role which is reusable to configure OpenStack"},{"line_number":87,"context_line":"  against an externally managed Ceph cluster"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"Why propose openstack/tripleo-ceph as a separate repository instead"},{"line_number":90,"context_line":"of a set of roles within tripleo-ansible?"}],"source_content_type":"text/x-rst","patch_set":20,"id":"9f560f44_dfa55303","line":87,"range":{"start_line":75,"start_character":0,"end_line":87,"end_character":44},"updated":"2020-10-01 19:24:57.000000000","message":"@Giulio,\n\nDo you want to cut this (and similar here) out and put it in a separate spec and then we can update this spec to link to the separate one?\n\nThanks,\n  John","commit_id":"97362cc4b37970d5c5e85462a9ed3995c58af942"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"92ba74f8aade2ad50a9d9af6db17845c4d06afe3","unresolved":false,"context_lines":[{"line_number":72,"context_line":"  - cephadm reconciliation loop must not break OpenStack configuration"},{"line_number":73,"context_line":"  - TripleO configuration updates must not break Ceph configuration"},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"tripleo-ansible/roles/ceph_client goals:"},{"line_number":76,"context_line":""},{"line_number":77,"context_line":"- Configure OpenStack services as clients of an external Ceph cluster"},{"line_number":78,"context_line":"  (in the case of collocation, the ceph cluster is still logically"},{"line_number":79,"context_line":"  external)"},{"line_number":80,"context_line":"- Provide Ceph configuration files and cephx keys for OpenStack"},{"line_number":81,"context_line":"  clients of RBD, RGW, CephFS (Nova, Cinder, Glance, Manila)"},{"line_number":82,"context_line":"- Full multiclient support, e.g. one OpenStack deployment may use"},{"line_number":83,"context_line":"  multiple Ceph clusters, e.g. multibackend Glance"},{"line_number":84,"context_line":"- Configure clients quickly, e.g. generate the key in one place,"},{"line_number":85,"context_line":"  without a container, and copy it efficiently"},{"line_number":86,"context_line":"- This is a standalone role which is reusable to configure OpenStack"},{"line_number":87,"context_line":"  against an externally managed Ceph cluster"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"Why propose openstack/tripleo-ceph as a separate repository instead"},{"line_number":90,"context_line":"of a set of roles within tripleo-ansible?"}],"source_content_type":"text/x-rst","patch_set":20,"id":"9f560f44_f392befb","line":87,"range":{"start_line":75,"start_character":0,"end_line":87,"end_character":44},"in_reply_to":"9f560f44_dfa55303","updated":"2020-10-08 09:40:50.000000000","message":"+1","commit_id":"97362cc4b37970d5c5e85462a9ed3995c58af942"},{"author":{"_account_id":14985,"name":"Alex Schultz","email":"aschultz@next-development.com","username":"mwhahaha"},"change_message_id":"6dd4eaff9d6e5a6d1c9aecda030be58ebb068f6e","unresolved":false,"context_lines":[{"line_number":57,"context_line":"- For day2 Ceph operations use cephadm, orchestrator, or Dashboard"},{"line_number":58,"context_line":"  directly unless the day2 Ceph change involves adding or removing"},{"line_number":59,"context_line":"  hardware (scale up or scale down) and other exceptions as needed"},{"line_number":60,"context_line":"- TripleO stack updates do not trigger tripleo-ceph by default"},{"line_number":61,"context_line":"- TripleO stack updates will have an option trigger tripleo-ceph"},{"line_number":62,"context_line":"  when appropriate, e.g. during scale up or scale down."},{"line_number":63,"context_line":"- Provide an opinionated Ceph installation based on parameters from"}],"source_content_type":"text/x-rst","patch_set":25,"id":"3f65232a_058940c9","line":60,"range":{"start_line":60,"start_character":2,"end_line":60,"end_character":62},"updated":"2020-10-26 20:36:22.000000000","message":"This seems wrong by the way the deployment works. Maybe it should trigger but it may be up to tripleo-ceph to determine if something should update?  If this is a desired thing, it makes my request to move it to a completely external process even more of the correct solution.","commit_id":"3d9fe49dd65c349badc48f8c6a9044ba650356fa"},{"author":{"_account_id":14985,"name":"Alex Schultz","email":"aschultz@next-development.com","username":"mwhahaha"},"change_message_id":"6dd4eaff9d6e5a6d1c9aecda030be58ebb068f6e","unresolved":false,"context_lines":[{"line_number":59,"context_line":"  hardware (scale up or scale down) and other exceptions as needed"},{"line_number":60,"context_line":"- TripleO stack updates do not trigger tripleo-ceph by default"},{"line_number":61,"context_line":"- TripleO stack updates will have an option trigger tripleo-ceph"},{"line_number":62,"context_line":"  when appropriate, e.g. during scale up or scale down."},{"line_number":63,"context_line":"- Provide an opinionated Ceph installation based on parameters from"},{"line_number":64,"context_line":"  TripleO and Ironic"},{"line_number":65,"context_line":"- Provide Ceph integration but maximize orthogonality between"}],"source_content_type":"text/x-rst","patch_set":25,"id":"3f65232a_258efccf","line":62,"updated":"2020-10-26 20:36:22.000000000","message":"We have scale down/scale up tasks in the framework so this is likely something that may work.","commit_id":"3d9fe49dd65c349badc48f8c6a9044ba650356fa"},{"author":{"_account_id":6926,"name":"Bogdan Dobrelya","email":"bdobreli@redhat.com","username":"bogdando"},"change_message_id":"cee0e0722f892aa9049ff3520ab82b45fb17fe65","unresolved":false,"context_lines":[{"line_number":89,"context_line":"We could ask deployers to do this:"},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"- Deploy hardware and configure networking with the step1 tasks from"},{"line_number":92,"context_line":"  config-download"},{"line_number":93,"context_line":"- Use cephadm and orchestrator directly to configure that hardware"},{"line_number":94,"context_line":"  with Ceph and create OpenStack pools accessible by CephX clients"},{"line_number":95,"context_line":"- Use TripleO to configure OpenStack by using steps 2-5 from"},{"line_number":96,"context_line":"  config-download."},{"line_number":97,"context_line":""}],"source_content_type":"text/x-rst","patch_set":25,"id":"7f6b1bfe_54ae682e","line":94,"range":{"start_line":92,"start_character":2,"end_line":94,"end_character":66},"updated":"2020-10-15 10:35:40.000000000","message":"while this option looks the most viable, the 2nd spec (for the ceph-client work) outlines a few downsides of cephadm.\n\nBut personally I\u0027d prefer we to use already existing tools supported with Ceph community instead of going into customized ansible for TripleO...","commit_id":"3d9fe49dd65c349badc48f8c6a9044ba650356fa"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"5a88e7944fa6fcccb9a9ed76dc10df1de3922e10","unresolved":false,"context_lines":[{"line_number":89,"context_line":"We could ask deployers to do this:"},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"- Deploy hardware and configure networking with the step1 tasks from"},{"line_number":92,"context_line":"  config-download"},{"line_number":93,"context_line":"- Use cephadm and orchestrator directly to configure that hardware"},{"line_number":94,"context_line":"  with Ceph and create OpenStack pools accessible by CephX clients"},{"line_number":95,"context_line":"- Use TripleO to configure OpenStack by using steps 2-5 from"},{"line_number":96,"context_line":"  config-download."},{"line_number":97,"context_line":""}],"source_content_type":"text/x-rst","patch_set":25,"id":"3f65232a_bae9b781","line":94,"range":{"start_line":92,"start_character":2,"end_line":94,"end_character":66},"in_reply_to":"7f6b1bfe_54ae682e","updated":"2020-10-22 06:01:04.000000000","message":"you mean for the clients configuration too?\n\ncephadm can\u0027t do clients configuration for now and with ceph-ansible not being supported from pacific, we\u0027d have to \"import\" the existing ceph-ansible roles in tripleo but would remain the sole group responsible for it","commit_id":"3d9fe49dd65c349badc48f8c6a9044ba650356fa"},{"author":{"_account_id":14985,"name":"Alex Schultz","email":"aschultz@next-development.com","username":"mwhahaha"},"change_message_id":"6dd4eaff9d6e5a6d1c9aecda030be58ebb068f6e","unresolved":false,"context_lines":[{"line_number":88,"context_line":""},{"line_number":89,"context_line":"We could ask deployers to do this:"},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"- Deploy hardware and configure networking with the step1 tasks from"},{"line_number":92,"context_line":"  config-download"},{"line_number":93,"context_line":"- Use cephadm and orchestrator directly to configure that hardware"},{"line_number":94,"context_line":"  with Ceph and create OpenStack pools accessible by CephX clients"},{"line_number":95,"context_line":"- Use TripleO to configure OpenStack by using steps 2-5 from"},{"line_number":96,"context_line":"  config-download."},{"line_number":97,"context_line":""},{"line_number":98,"context_line":"We have completed a POC of the above using Ussuri but would prefer to"},{"line_number":99,"context_line":"offer an option to automate step two. The TripleO project has already"}],"source_content_type":"text/x-rst","patch_set":25,"id":"3f65232a_a5bb2c70","line":96,"range":{"start_line":91,"start_character":0,"end_line":96,"end_character":18},"updated":"2020-10-26 20:36:22.000000000","message":"As mentioned @PTG, I think the alternative to creating roles that are invoked via THT (e.g. the existing method), would be to move the management of ceph outside of the `openstack overcloud deploy` process.  As outlined, it would be something like...\n\n1) `openstack overcloud node provision`\n2) network configuration needs to be executed. (Currently proposed effort as part of network config v2)\n3) ceph provision executed against provisioned nodes\n4) `openstack overcloud deploy` run with inputs from node provisioning and ceph processes.\n\nThis allows the management of ceph to continue to be collocated with the tripleo nodes, but allows for proper separation for updates/upgrades/etc.","commit_id":"3d9fe49dd65c349badc48f8c6a9044ba650356fa"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"32001d82291909777789f2374ceb7f12dc61af60","unresolved":false,"context_lines":[{"line_number":88,"context_line":""},{"line_number":89,"context_line":"We could ask deployers to do this:"},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"- Deploy hardware and configure networking with the step1 tasks from"},{"line_number":92,"context_line":"  config-download"},{"line_number":93,"context_line":"- Use cephadm and orchestrator directly to configure that hardware"},{"line_number":94,"context_line":"  with Ceph and create OpenStack pools accessible by CephX clients"},{"line_number":95,"context_line":"- Use TripleO to configure OpenStack by using steps 2-5 from"},{"line_number":96,"context_line":"  config-download."},{"line_number":97,"context_line":""},{"line_number":98,"context_line":"We have completed a POC of the above using Ussuri but would prefer to"},{"line_number":99,"context_line":"offer an option to automate step two. The TripleO project has already"}],"source_content_type":"text/x-rst","patch_set":25,"id":"3f65232a_ab493d8c","line":96,"range":{"start_line":91,"start_character":0,"end_line":96,"end_character":18},"in_reply_to":"3f65232a_a5bb2c70","updated":"2020-10-27 02:37:53.000000000","message":"I updated the spec to reflect this suggestion.","commit_id":"3d9fe49dd65c349badc48f8c6a9044ba650356fa"},{"author":{"_account_id":16643,"name":"Goutham Pacha Ravi","email":"gouthampravi@gmail.com","username":"gouthamr"},"change_message_id":"a8ca874ad96a45aca0fd3561e0ae55dda33b77e6","unresolved":false,"context_lines":[{"line_number":213,"context_line":"    be taken by an operator when TripleO should be involved. E.g."},{"line_number":214,"context_line":"    a CephStorage node is added as a scale-up operation, then"},{"line_number":215,"context_line":"    tripleo-ceph should make calls to add the OSDs."},{"line_number":216,"context_line":""},{"line_number":217,"context_line":""},{"line_number":218,"context_line":"Performance Impact"},{"line_number":219,"context_line":"------------------"}],"source_content_type":"text/x-rst","patch_set":25,"id":"3f65232a_9a9ddb5b","line":216,"range":{"start_line":216,"start_character":0,"end_line":216,"end_character":0},"updated":"2020-10-22 04:59:48.000000000","message":"Currently, the cephadm tooling does not allow deploying/configuring the \"ceph-nfs\" (nfs-ganesha) service that is possible with ceph-ansible. I know you are working with the ceph community to have this added to cephadm. Perhaps we should call that out here, along with the possible alternative of:\nIn X, if cephadm still has no support to deploy ceph-nfs, we\u0027d have to keep portions of ceph-ansible around for nfs-ganesha to continue to work;","commit_id":"3d9fe49dd65c349badc48f8c6a9044ba650356fa"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"5a88e7944fa6fcccb9a9ed76dc10df1de3922e10","unresolved":false,"context_lines":[{"line_number":213,"context_line":"    be taken by an operator when TripleO should be involved. E.g."},{"line_number":214,"context_line":"    a CephStorage node is added as a scale-up operation, then"},{"line_number":215,"context_line":"    tripleo-ceph should make calls to add the OSDs."},{"line_number":216,"context_line":""},{"line_number":217,"context_line":""},{"line_number":218,"context_line":"Performance Impact"},{"line_number":219,"context_line":"------------------"}],"source_content_type":"text/x-rst","patch_set":25,"id":"3f65232a_fa020f3e","line":216,"range":{"start_line":216,"start_character":0,"end_line":216,"end_character":0},"in_reply_to":"3f65232a_9a9ddb5b","updated":"2020-10-22 06:01:04.000000000","message":"ceph orchestrator can deploy ganesha but only via ceph orchestrator, while in tripleo we want to support ganesha deployment when ceph is externally managed, which we couldn\u0027t implement unless tripleo is given admin credentials of the \"remote\" ceph orchestrator\n\nhence, we\u0027re thinking to import in tripleo the existing ganesha role from ceph-ansible for this but don\u0027t have a spec for it yet","commit_id":"3d9fe49dd65c349badc48f8c6a9044ba650356fa"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"32001d82291909777789f2374ceb7f12dc61af60","unresolved":false,"context_lines":[{"line_number":213,"context_line":"    be taken by an operator when TripleO should be involved. E.g."},{"line_number":214,"context_line":"    a CephStorage node is added as a scale-up operation, then"},{"line_number":215,"context_line":"    tripleo-ceph should make calls to add the OSDs."},{"line_number":216,"context_line":""},{"line_number":217,"context_line":""},{"line_number":218,"context_line":"Performance Impact"},{"line_number":219,"context_line":"------------------"}],"source_content_type":"text/x-rst","patch_set":25,"id":"3f65232a_2b552d1e","line":216,"range":{"start_line":216,"start_character":0,"end_line":216,"end_character":0},"in_reply_to":"3f65232a_fa020f3e","updated":"2020-10-27 02:37:53.000000000","message":"That spec is at https://review.opendev.org/#/c/759395","commit_id":"3d9fe49dd65c349badc48f8c6a9044ba650356fa"},{"author":{"_account_id":14985,"name":"Alex Schultz","email":"aschultz@next-development.com","username":"mwhahaha"},"change_message_id":"b544c911d70279161a6261539792e2d42e2ae314","unresolved":false,"context_lines":[{"line_number":46,"context_line":""},{"line_number":47,"context_line":"- Provide Ansible roles which the TripleO client can use to deploy"},{"line_number":48,"context_line":"  Ceph by calling cephadm and Ceph orchestrator."},{"line_number":49,"context_line":"- The new Ansible roles will be triggered when the user runs a command"},{"line_number":50,"context_line":"  like `openstack overcloud ceph ...`; this command is meant"},{"line_number":51,"context_line":"  to be run after running `openstack overcloud node provision` [4]_"},{"line_number":52,"context_line":"  but before running `openstack overcloud deploy`."},{"line_number":53,"context_line":"- Focus on the day1 problem for Ceph RBD, RGW, CephFS, and Dashboard"},{"line_number":54,"context_line":"  deployment by leveraging `cephadm bootstrap --apply-spec` [7]_."},{"line_number":55,"context_line":"- Provide an opinionated Ceph installation based on parameters from"}],"source_content_type":"text/x-rst","patch_set":28,"id":"3f65232a_7186e8dd","line":52,"range":{"start_line":49,"start_character":0,"end_line":52,"end_character":50},"updated":"2020-10-27 18:13:27.000000000","message":"I think this is a stretch goal depending on the output of the network changes.  If this can\u0027t be reached, then we can continue with invoking ansible as an external_deploy_task in step1 which is after the network has been configured, after host_prep_tasks, and before deploy step tasks step 1.  It make make sense to explain this as a possibility somewhere.","commit_id":"4e3ebb95d8293ee9c5ef4cb0dd0b3de5d44cfe8b"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"646cb27ad4f9e130cf4ec1b8c363e7b4504ad57a","unresolved":false,"context_lines":[{"line_number":46,"context_line":""},{"line_number":47,"context_line":"- Provide Ansible roles which the TripleO client can use to deploy"},{"line_number":48,"context_line":"  Ceph by calling cephadm and Ceph orchestrator."},{"line_number":49,"context_line":"- The new Ansible roles will be triggered when the user runs a command"},{"line_number":50,"context_line":"  like `openstack overcloud ceph ...`; this command is meant"},{"line_number":51,"context_line":"  to be run after running `openstack overcloud node provision` [4]_"},{"line_number":52,"context_line":"  but before running `openstack overcloud deploy`."},{"line_number":53,"context_line":"- Focus on the day1 problem for Ceph RBD, RGW, CephFS, and Dashboard"},{"line_number":54,"context_line":"  deployment by leveraging `cephadm bootstrap --apply-spec` [7]_."},{"line_number":55,"context_line":"- Provide an opinionated Ceph installation based on parameters from"}],"source_content_type":"text/x-rst","patch_set":28,"id":"1f621f24_b7e1c4d3","line":52,"range":{"start_line":49,"start_character":0,"end_line":52,"end_character":50},"in_reply_to":"1f621f24_9ff2f605","updated":"2020-10-29 22:20:17.000000000","message":"done.","commit_id":"4e3ebb95d8293ee9c5ef4cb0dd0b3de5d44cfe8b"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"b4060c50bf9820531f31411c34262e7ce26edb10","unresolved":false,"context_lines":[{"line_number":46,"context_line":""},{"line_number":47,"context_line":"- Provide Ansible roles which the TripleO client can use to deploy"},{"line_number":48,"context_line":"  Ceph by calling cephadm and Ceph orchestrator."},{"line_number":49,"context_line":"- The new Ansible roles will be triggered when the user runs a command"},{"line_number":50,"context_line":"  like `openstack overcloud ceph ...`; this command is meant"},{"line_number":51,"context_line":"  to be run after running `openstack overcloud node provision` [4]_"},{"line_number":52,"context_line":"  but before running `openstack overcloud deploy`."},{"line_number":53,"context_line":"- Focus on the day1 problem for Ceph RBD, RGW, CephFS, and Dashboard"},{"line_number":54,"context_line":"  deployment by leveraging `cephadm bootstrap --apply-spec` [7]_."},{"line_number":55,"context_line":"- Provide an opinionated Ceph installation based on parameters from"}],"source_content_type":"text/x-rst","patch_set":28,"id":"1f621f24_9ff2f605","line":52,"range":{"start_line":49,"start_character":0,"end_line":52,"end_character":50},"in_reply_to":"3f65232a_7186e8dd","updated":"2020-10-28 15:26:34.000000000","message":"ack, i\u0027ll update spec as per 2nd part of ptg discussion... If network isn\u0027t ready at end of cycle, then we can have a new external deploy steps task which can execute the new playbook in tripleo-ceph and pick up the new configuration definition described in tripleo_ceph_config.yaml below.","commit_id":"4e3ebb95d8293ee9c5ef4cb0dd0b3de5d44cfe8b"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"d74d0c698202a69d068d053bc93491d42477a244","unresolved":false,"context_lines":[{"line_number":378,"context_line":"Work Items"},{"line_number":379,"context_line":"----------"},{"line_number":380,"context_line":""},{"line_number":381,"context_line":"- Create the repository openstack/tripleo-ceph to host the new code"},{"line_number":382,"context_line":"- Configure unit tests and molecule ansible tests (no need for"},{"line_number":383,"context_line":"  tripleo-standalone testing)"},{"line_number":384,"context_line":"- Work with RDO to get RPM built for tripleo-ceph"}],"source_content_type":"text/x-rst","patch_set":28,"id":"1f621f24_1f2ee6a4","line":381,"range":{"start_line":381,"start_character":0,"end_line":381,"end_character":67},"updated":"2020-10-28 16:18:14.000000000","message":"as per PTG discussion this could go in tripleo-ansible. It\u0027s needed to fully support tripleo python client anway.","commit_id":"4e3ebb95d8293ee9c5ef4cb0dd0b3de5d44cfe8b"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"646cb27ad4f9e130cf4ec1b8c363e7b4504ad57a","unresolved":false,"context_lines":[{"line_number":378,"context_line":"Work Items"},{"line_number":379,"context_line":"----------"},{"line_number":380,"context_line":""},{"line_number":381,"context_line":"- Create the repository openstack/tripleo-ceph to host the new code"},{"line_number":382,"context_line":"- Configure unit tests and molecule ansible tests (no need for"},{"line_number":383,"context_line":"  tripleo-standalone testing)"},{"line_number":384,"context_line":"- Work with RDO to get RPM built for tripleo-ceph"}],"source_content_type":"text/x-rst","patch_set":28,"id":"1f621f24_17e7d8e6","line":381,"range":{"start_line":381,"start_character":0,"end_line":381,"end_character":67},"in_reply_to":"1f621f24_1f2ee6a4","updated":"2020-10-29 22:20:17.000000000","message":"spec updated accordingly.","commit_id":"4e3ebb95d8293ee9c5ef4cb0dd0b3de5d44cfe8b"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"271dd028b97a5f8034d7e7093de48824cbce79c1","unresolved":false,"context_lines":[{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_."},{"line_number":94,"context_line":"Item 3 above is the end goal of this spec but it is not necessarily"},{"line_number":95,"context_line":"the only integration point. If it is not possible to configure the"},{"line_number":96,"context_line":"storage networks prior to deploying OpenStack, then the new method"}],"source_content_type":"text/x-rst","patch_set":29,"id":"1f621f24_36976d93","line":93,"range":{"start_line":93,"start_character":0,"end_line":93,"end_character":66},"updated":"2020-10-30 15:59:13.000000000","message":"This spec covers the format but may not include the work to move port management out of heat, and apply network configuration prior to the heat stack deployment. It\u0027s not clear that will be done during W so we need to have the 2-step process. We\u0027ll set up the 2-step process so we can move to the 4-step process as soon as we can however.","commit_id":"71631f87eb487a5646631caecc3d4c39ea436492"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"78de3316ba5ca23e8271c5604b8998ec74c0ab85","unresolved":false,"context_lines":[{"line_number":364,"context_line":"required to use them."},{"line_number":365,"context_line":""},{"line_number":366,"context_line":"Regardless of if the 2-step or 4-step process is followed, prior to"},{"line_number":367,"context_line":"deploying Ceph, `openstack overcloud admin authorize` should be run"},{"line_number":368,"context_line":"and it should pass options to enable a ceph-admin user which can be"},{"line_number":369,"context_line":"used by cephadm and to allow SSH access for the ansible roles"},{"line_number":370,"context_line":"described in this spec."},{"line_number":371,"context_line":""}],"source_content_type":"text/x-rst","patch_set":29,"id":"1f621f24_f7397c41","line":368,"range":{"start_line":367,"start_character":16,"end_line":368,"end_character":55},"updated":"2020-10-29 22:25:08.000000000","message":"This command will not always require the --stack option (we won\u0027t have a stack yet) and harold will need to solve the same problem for this network spec.\n\n https://review.opendev.org/#/c/760462/1","commit_id":"71631f87eb487a5646631caecc3d4c39ea436492"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"78de3316ba5ca23e8271c5604b8998ec74c0ab85","unresolved":false,"context_lines":[{"line_number":423,"context_line":"the RBD pools used by Glance, Cinder, and Nova, as well as the cephx"},{"line_number":424,"context_line":"keyrings to use to authenticate to use those pools."},{"line_number":425,"context_line":""},{"line_number":426,"context_line":"In the case of the 2-step-process, if external_deploy_steps_tasks"},{"line_number":427,"context_line":"calls the new playbook, then the same file will be written to it\u0027s"},{"line_number":428,"context_line":"default location (/home/stack/ceph_client.yaml) and the new client"},{"line_number":429,"context_line":"role will directly read the parameters from this file."},{"line_number":430,"context_line":""},{"line_number":431,"context_line":"Properties of the 2-step-process"},{"line_number":432,"context_line":"--------------------------------"}],"source_content_type":"text/x-rst","patch_set":29,"id":"1f621f24_77d72cef","line":429,"range":{"start_line":426,"start_character":0,"end_line":429,"end_character":54},"updated":"2020-10-29 22:25:08.000000000","message":"@fmount let\u0027s talk more about this part. It ties into things we were discussing in https://review.opendev.org/#/c/758039/ today","commit_id":"71631f87eb487a5646631caecc3d4c39ea436492"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"c0941640bc39cf596fb3ed5adaca29bd78ffffc6","unresolved":false,"context_lines":[{"line_number":423,"context_line":"the RBD pools used by Glance, Cinder, and Nova, as well as the cephx"},{"line_number":424,"context_line":"keyrings to use to authenticate to use those pools."},{"line_number":425,"context_line":""},{"line_number":426,"context_line":"In the case of the 2-step-process, if external_deploy_steps_tasks"},{"line_number":427,"context_line":"calls the new playbook, then the same file will be written to it\u0027s"},{"line_number":428,"context_line":"default location (/home/stack/ceph_client.yaml) and the new client"},{"line_number":429,"context_line":"role will directly read the parameters from this file."},{"line_number":430,"context_line":""},{"line_number":431,"context_line":"Properties of the 2-step-process"},{"line_number":432,"context_line":"--------------------------------"}],"source_content_type":"text/x-rst","patch_set":29,"id":"1f621f24_86ed78f1","line":429,"range":{"start_line":426,"start_character":0,"end_line":429,"end_character":54},"in_reply_to":"1f621f24_77d72cef","updated":"2020-10-30 08:17:49.000000000","message":"Yes, I agree, let\u0027s talk more about this to make sure we\u0027re able to cover all the use cases we\u0027re thinking about.\n\nIn both 2 and 4 step process, when cephadm is involved, the output produced and passed to the `overcloud deploy` command should be the same, with the same parameters, so I would imagine that we will have the same deprecation model on tht (e.g. no ceph_{mon,osd,rgw,grafana} services anymore), block [1] should be skipped but `ceph_mon_ips` is produced by this new process, right?\n\nIf we have a ceph-ansible deployment (this means we\u0027re going to backport the ceph_client to Train), the same condition [1] is satisfied, and ceph_mon_ips can only be found in hieradata; ceph-ansible is able to produce these variable but I\u0027m not sure they\u0027re available outside its execution, and if we want to keep this role w/ no interactions w/ the Ceph cluster, I don\u0027t see other ways (so far) to get monitor ip addresses (but I\u0027m open to suggestions of course :D).\n\nFinally, if the cluster is external (no matter if it\u0027s ceph-ansible or cephadm deployed), then we should maintain the same interface we have today in tht [2].\n\nNot sure if I\u0027m missing pieces here (probably yes), but this could be a good starting point to make sure all the use cases are covered by that role.  \n\n\n[1] https://review.opendev.org/#/c/758039/21/tripleo_ansible/roles/tripleo_ceph_client/tasks/create_ceph_conf.yml@28\n\n[2] https://github.com/openstack/tripleo-heat-templates/blob/master/environments/ceph-ansible/ceph-ansible-external.yaml","commit_id":"71631f87eb487a5646631caecc3d4c39ea436492"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"271dd028b97a5f8034d7e7093de48824cbce79c1","unresolved":false,"context_lines":[{"line_number":423,"context_line":"the RBD pools used by Glance, Cinder, and Nova, as well as the cephx"},{"line_number":424,"context_line":"keyrings to use to authenticate to use those pools."},{"line_number":425,"context_line":""},{"line_number":426,"context_line":"In the case of the 2-step-process, if external_deploy_steps_tasks"},{"line_number":427,"context_line":"calls the new playbook, then the same file will be written to it\u0027s"},{"line_number":428,"context_line":"default location (/home/stack/ceph_client.yaml) and the new client"},{"line_number":429,"context_line":"role will directly read the parameters from this file."},{"line_number":430,"context_line":""},{"line_number":431,"context_line":"Properties of the 2-step-process"},{"line_number":432,"context_line":"--------------------------------"}],"source_content_type":"text/x-rst","patch_set":29,"id":"1f621f24_79ab961c","line":429,"range":{"start_line":426,"start_character":0,"end_line":429,"end_character":54},"in_reply_to":"1f621f24_86ed78f1","updated":"2020-10-30 15:59:13.000000000","message":"\u003e Yes, I agree, let\u0027s talk more about this to make sure we\u0027re able to\n \u003e cover all the use cases we\u0027re thinking about.\n \u003e \n \u003e In both 2 and 4 step process, when cephadm is involved, the output\n \u003e produced and passed to the `overcloud deploy` command should be the\n \u003e same, with the same parameters, so I would imagine that we will\n \u003e have the same deprecation model on tht (e.g. no ceph_{mon,osd,rgw,grafana}\n \u003e services anymore), block [1] should be skipped but `ceph_mon_ips`\n \u003e is produced by this new process, right?\n \nYes. The new client role can require that the ceph_mon_ips always be passed to it in the new method. The exception is that the client role can gather this information on its own for backwards compatibility with ceph-ansible internal deployments.\n\n \u003e If we have a ceph-ansible deployment (this means we\u0027re going to\n \u003e backport the ceph_client to Train), the same condition [1] is\n \u003e satisfied, and ceph_mon_ips can only be found in hieradata;\n\nYes.\n\n \u003e ceph-ansible is able to produce these variable but I\u0027m not sure\n \u003e they\u0027re available outside its execution, and if we want to keep\n \u003e this role w/ no interactions w/ the Ceph cluster, I don\u0027t see other\n \u003e ways (so far) to get monitor ip addresses (but I\u0027m open to\n \u003e suggestions of course :D).\n\nRight, so we keep this part of the new client role to gather that information in that case. \n\nhttps://review.opendev.org/#/c/758039/23/tripleo_ansible/roles/tripleo_ceph_client/tasks/ceph_ansible.yml\n\n \u003e \n \u003e Finally, if the cluster is external (no matter if it\u0027s ceph-ansible\n \u003e or cephadm deployed), then we should maintain the same interface we\n \u003e have today in tht [2].\n\nYes, I agree that that will include passing the CephExternalMonHost. \n\n\n \u003e Not sure if I\u0027m missing pieces here (probably yes), but this could\n \u003e be a good starting point to make sure all the use cases are covered\n \u003e by that role.\n \nI think the question here is SHOULD the client role (for the new method only) have a task to look for /home/stack/ceph_client.yaml and use those paramters if it is present. That way for the 2-step process we can use the same interface as the 4-step process.\n  \n \u003e [1] https://review.opendev.org/#/c/758039/21/tripleo_ansible/roles/tripleo_ceph_client/tasks/create_ceph_conf.yml@28\n \u003e \n \u003e [2] https://github.com/openstack/tripleo-heat-templates/blob/master/environments/ceph-ansible/ceph-ansible-external.yaml","commit_id":"71631f87eb487a5646631caecc3d4c39ea436492"},{"author":{"_account_id":24245,"name":"Harald Jensås","email":"hjensas@redhat.com","username":"harald.jensas"},"change_message_id":"5fd53319c18d3bdb901a69ac28150f26ab10c750","unresolved":false,"context_lines":[{"line_number":87,"context_line":""},{"line_number":88,"context_line":"1. Deploy the hardware with metalsmith [4]_"},{"line_number":89,"context_line":"2. Configure networking (including storage networks)"},{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_"},{"line_number":94,"context_line":"and a subsequent network-related feature which moves port management"}],"source_content_type":"text/x-rst","patch_set":32,"id":"1f621f24_3698e419","line":91,"range":{"start_line":90,"start_character":0,"end_line":91,"end_character":53},"updated":"2020-11-03 14:51:28.000000000","message":"In this 4 step process, how is firewall configuration handled? Today firewall rules are defined in the THT service templates, and the tripleo_firewall role in tripleo-ansible configures the firewall. (I assume the \"Deploy Ceph\" tool will configure firewall rules, but will those rules be removed by the tripleo_firewall role?)","commit_id":"f8641f5c0f8204214cb79b7a115417b9fc5fe052"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"cb90ebcf16f7c163932ad8a445f77486c061bf64","unresolved":false,"context_lines":[{"line_number":87,"context_line":""},{"line_number":88,"context_line":"1. Deploy the hardware with metalsmith [4]_"},{"line_number":89,"context_line":"2. Configure networking (including storage networks)"},{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_"},{"line_number":94,"context_line":"and a subsequent network-related feature which moves port management"}],"source_content_type":"text/x-rst","patch_set":32,"id":"1f621f24_48eb2faf","line":91,"range":{"start_line":90,"start_character":0,"end_line":91,"end_character":53},"in_reply_to":"1f621f24_19c5145c","updated":"2020-11-04 13:16:40.000000000","message":"There will be no rules (iptables or firewalld) from cephadm to set. We shouldn\u0027t need to execute anything again. We\u0027ll use the rules defined in THT. No rules will be there before so there will be nothing to delete.\n\nI\u0027ll update the spec on line 327 to state this.","commit_id":"f8641f5c0f8204214cb79b7a115417b9fc5fe052"},{"author":{"_account_id":14985,"name":"Alex Schultz","email":"aschultz@next-development.com","username":"mwhahaha"},"change_message_id":"b27fb08613751c9f2c84515585cc22f92ca9aaee","unresolved":false,"context_lines":[{"line_number":87,"context_line":""},{"line_number":88,"context_line":"1. Deploy the hardware with metalsmith [4]_"},{"line_number":89,"context_line":"2. Configure networking (including storage networks)"},{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_"},{"line_number":94,"context_line":"and a subsequent network-related feature which moves port management"}],"source_content_type":"text/x-rst","patch_set":32,"id":"1f621f24_f98818c7","line":91,"range":{"start_line":90,"start_character":0,"end_line":91,"end_character":53},"in_reply_to":"1f621f24_19c5145c","updated":"2020-11-03 22:05:28.000000000","message":"tripleo_firewall should leave existing iptables rules in place if they not a managed rules. At least that\u0027s how it used to be handled for puppet-tripleo (probably needs verification). It should be noted that nftables is not supported so if that\u0027s how cephadm is setting them up today, that\u0027ll likely break once tripleo applies.","commit_id":"f8641f5c0f8204214cb79b7a115417b9fc5fe052"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"603c5b8109d78d9db1867110869025669a163ac9","unresolved":false,"context_lines":[{"line_number":87,"context_line":""},{"line_number":88,"context_line":"1. Deploy the hardware with metalsmith [4]_"},{"line_number":89,"context_line":"2. Configure networking (including storage networks)"},{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_"},{"line_number":94,"context_line":"and a subsequent network-related feature which moves port management"}],"source_content_type":"text/x-rst","patch_set":32,"id":"1f621f24_41379cb1","line":91,"range":{"start_line":90,"start_character":0,"end_line":91,"end_character":53},"in_reply_to":"1f621f24_3698e419","updated":"2020-11-03 15:07:53.000000000","message":"The new tool is supposed to deploy ceph, but the orchestration that is run on top of cephadm should be composed by a set of tripleo-ansible tasks that are able to execute actions against the ceph cluster.\nSince everything is within the same project, even though we deprecate the current tht we should be able to include the tripleo-firewall role, applying the rules we need for the involved services (mon, mgr, osds, rgw).\nWe can do something like:\n\n```\n    - include_role: \"tripleo_firewall\"\n      vars:\n        firewall_rules:\n        \u0027110 ceph_mon\u0027:\n          dport:\n            list_concat:\n              - - 6789\n              - - 3300\n```\nThat\u0027s an high level idea and it should be explicitly described here, @John do you have any concern on this?","commit_id":"f8641f5c0f8204214cb79b7a115417b9fc5fe052"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"14513d916994366800c30123508a06eebd9d4a4b","unresolved":false,"context_lines":[{"line_number":87,"context_line":""},{"line_number":88,"context_line":"1. Deploy the hardware with metalsmith [4]_"},{"line_number":89,"context_line":"2. Configure networking (including storage networks)"},{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_"},{"line_number":94,"context_line":"and a subsequent network-related feature which moves port management"}],"source_content_type":"text/x-rst","patch_set":32,"id":"1f621f24_81353449","line":91,"range":{"start_line":90,"start_character":0,"end_line":91,"end_character":53},"in_reply_to":"1f621f24_41379cb1","updated":"2020-11-03 15:23:30.000000000","message":"I don\u0027t think we need to run the firewall config in advance and we can leave what we have in place in THT. \n\nAFAICT cephadm doesn\u0027t configure the firewall [1] so TripleO won\u0027t remove any firewall configuration done by the ceph tool. Provided the deployed hosts don\u0027t have a deny-all firewall settings by default everything should just work. Later when firewall is configured by \u0027openstack overcloud deploy\u0027 it will open the ports Ceph would need for the roles deployed.\n\n[1] http://paste.openstack.org/show/799656/","commit_id":"f8641f5c0f8204214cb79b7a115417b9fc5fe052"},{"author":{"_account_id":24245,"name":"Harald Jensås","email":"hjensas@redhat.com","username":"harald.jensas"},"change_message_id":"41b10b266940bcc36f5b0e2f7db68773d4d5adaf","unresolved":false,"context_lines":[{"line_number":87,"context_line":""},{"line_number":88,"context_line":"1. Deploy the hardware with metalsmith [4]_"},{"line_number":89,"context_line":"2. Configure networking (including storage networks)"},{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_"},{"line_number":94,"context_line":"and a subsequent network-related feature which moves port management"}],"source_content_type":"text/x-rst","patch_set":32,"id":"1f621f24_19c5145c","line":91,"range":{"start_line":90,"start_character":0,"end_line":91,"end_character":53},"in_reply_to":"1f621f24_41379cb1","updated":"2020-11-03 21:31:15.000000000","message":"What does the \"tripleo_firewall\" role do with the rules added as Francesco suggests above when the role is \"executed\" again while deploying the overcloud? Unless those rules are defined in THT and thus in the config-download inventory I assume \"tripleo_firewall\" will remove them?","commit_id":"f8641f5c0f8204214cb79b7a115417b9fc5fe052"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"cb90ebcf16f7c163932ad8a445f77486c061bf64","unresolved":false,"context_lines":[{"line_number":87,"context_line":""},{"line_number":88,"context_line":"1. Deploy the hardware with metalsmith [4]_"},{"line_number":89,"context_line":"2. Configure networking (including storage networks)"},{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_"},{"line_number":94,"context_line":"and a subsequent network-related feature which moves port management"}],"source_content_type":"text/x-rst","patch_set":32,"id":"1f621f24_a8d123fe","line":91,"range":{"start_line":90,"start_character":0,"end_line":91,"end_character":53},"in_reply_to":"1f621f24_f98818c7","updated":"2020-11-04 13:16:40.000000000","message":"My mistake for bringing up iptables. cephadm doesn\u0027t set iptables. It can set firewalld rules but we will run it with `cephadm --skip-firewalld` so it doesn\u0027t.\n\nSo nothing will break when tripleo_firewall is run as there will be no firewall rules (iptables or firewalld) before `opentack overcloud deploy` is run.\n\nI\u0027ll update the spec on line 327 to state this.","commit_id":"f8641f5c0f8204214cb79b7a115417b9fc5fe052"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b010e00abbfecca48fe2bca84a6d2fff6ef29174","unresolved":false,"context_lines":[{"line_number":44,"context_line":"Modify tripleo-ansible, tripleo-heat-templates, and"},{"line_number":45,"context_line":"python-tripleoclient in support of the following goals:"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"- Provide Ansible roles which deploy Ceph by calling cephadm and Ceph"},{"line_number":48,"context_line":"  orchestrator"},{"line_number":49,"context_line":"- Focus on the day1 problem for Ceph RBD, RGW, CephFS, and Dashboard"},{"line_number":50,"context_line":"  deployment by leveraging `cephadm bootstrap --apply-spec` [7]_"},{"line_number":51,"context_line":"- By default, day2 Ceph operations should be done directly with Ceph"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_2e9cfbbe","line":48,"range":{"start_line":47,"start_character":2,"end_line":48,"end_character":14},"updated":"2020-11-04 15:06:46.000000000","message":"this seems at odds though with the \u0027ultimate goal\u0027 stated below where deploying ceph is \u0027outside\u0027 of the tripleo deployment in its own step 3. \n\nmy confusion is that you mention modifying tripleoclient/tripleo-heat-templates etc which implies that ceph is \u0027inside\u0027 the tripleo deployment?","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"acedd099b67cb8d771073bb254f6ffaf4644e7e0","unresolved":false,"context_lines":[{"line_number":44,"context_line":"Modify tripleo-ansible, tripleo-heat-templates, and"},{"line_number":45,"context_line":"python-tripleoclient in support of the following goals:"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"- Provide Ansible roles which deploy Ceph by calling cephadm and Ceph"},{"line_number":48,"context_line":"  orchestrator"},{"line_number":49,"context_line":"- Focus on the day1 problem for Ceph RBD, RGW, CephFS, and Dashboard"},{"line_number":50,"context_line":"  deployment by leveraging `cephadm bootstrap --apply-spec` [7]_"},{"line_number":51,"context_line":"- By default, day2 Ceph operations should be done directly with Ceph"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_feb8ecc8","line":48,"range":{"start_line":47,"start_character":2,"end_line":48,"end_character":14},"in_reply_to":"1f621f24_28450e18","updated":"2020-11-04 23:32:31.000000000","message":"\u003e This is an issue with transitioning the ownership of services away\n \u003e from tripleo.  If tripleo owns the firewall, it owns the whole\n \u003e thing and assumes it\u0027s managed in such a way.  In order for the\n \u003e firewalls to be properly owned, you\u0027d still have to tell tripleo\n \u003e which ports need to be open on which roles in order for the rules\n \u003e to properly be applied on the correct hosts. We do have an extra\n \u003e rules variable that could be used but would need to be merged with\n \u003e any customer\u0027s values.  It reduces the overall usefullness of\n \u003e letting ceph deploy bits. You could create dummy ceph services that\n \u003e can be used to handle the firewall/haproxy integration and not\n \u003e actually do any \"deployment\" actions.\n\nI think line 327 covers this. Is there something missing there regarding the problem pointed out above?","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"acedd099b67cb8d771073bb254f6ffaf4644e7e0","unresolved":false,"context_lines":[{"line_number":44,"context_line":"Modify tripleo-ansible, tripleo-heat-templates, and"},{"line_number":45,"context_line":"python-tripleoclient in support of the following goals:"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"- Provide Ansible roles which deploy Ceph by calling cephadm and Ceph"},{"line_number":48,"context_line":"  orchestrator"},{"line_number":49,"context_line":"- Focus on the day1 problem for Ceph RBD, RGW, CephFS, and Dashboard"},{"line_number":50,"context_line":"  deployment by leveraging `cephadm bootstrap --apply-spec` [7]_"},{"line_number":51,"context_line":"- By default, day2 Ceph operations should be done directly with Ceph"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_5e680020","line":48,"range":{"start_line":47,"start_character":2,"end_line":48,"end_character":14},"in_reply_to":"1f621f24_2e9cfbbe","updated":"2020-11-04 23:32:31.000000000","message":"\u003e this seems at odds though with the \u0027ultimate goal\u0027 stated below\n \u003e where deploying ceph is \u0027outside\u0027 of the tripleo deployment in its\n \u003e own step 3.\n \u003e \n \u003e my confusion is that you mention modifying tripleoclient/tripleo-heat-templates\n \u003e etc which implies that ceph is \u0027inside\u0027 the tripleo deployment?\n\nInstead of saying outside vs inside of TripleO I should be saying before or during overcloud deployment. When this spec is implemented, Ceph will not be configured when \u0027openstack overcloud deploy\u0027 but the Ceph deployment remains a part of TripleO. The change is that the Ceph deployment is decoupled from the overcloud deployment. By decoupling we reduce complexity. It fits the same pattern as PXE\u0027ing the servers outside of the overcloud deployment by using metalsmith.\n\nIn the next patchset lines 98-117 will make this more clear. Thanks for the suggestion.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"1fd13194101b55e09781b293bedb750d66c720f7","unresolved":false,"context_lines":[{"line_number":44,"context_line":"Modify tripleo-ansible, tripleo-heat-templates, and"},{"line_number":45,"context_line":"python-tripleoclient in support of the following goals:"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"- Provide Ansible roles which deploy Ceph by calling cephadm and Ceph"},{"line_number":48,"context_line":"  orchestrator"},{"line_number":49,"context_line":"- Focus on the day1 problem for Ceph RBD, RGW, CephFS, and Dashboard"},{"line_number":50,"context_line":"  deployment by leveraging `cephadm bootstrap --apply-spec` [7]_"},{"line_number":51,"context_line":"- By default, day2 Ceph operations should be done directly with Ceph"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_f992cb9d","line":48,"range":{"start_line":47,"start_character":2,"end_line":48,"end_character":14},"in_reply_to":"1f621f24_2e9cfbbe","updated":"2020-11-04 15:59:16.000000000","message":"The fact is that some TripleO configurations (firewalls, Haproxy, TLS and cert generation) rely on the overcloud deployment flow.\nThis means that you can split the \"provisioning\" aspect of Ceph (when cephadm is triggered), but during the overcloud deployment some tht are still needed to provide values and info to ensure a fully supported integration and avoid breaking existing features","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":14985,"name":"Alex Schultz","email":"aschultz@next-development.com","username":"mwhahaha"},"change_message_id":"74de75c38e2b35ae1071bf5bdca687706a821832","unresolved":false,"context_lines":[{"line_number":44,"context_line":"Modify tripleo-ansible, tripleo-heat-templates, and"},{"line_number":45,"context_line":"python-tripleoclient in support of the following goals:"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"- Provide Ansible roles which deploy Ceph by calling cephadm and Ceph"},{"line_number":48,"context_line":"  orchestrator"},{"line_number":49,"context_line":"- Focus on the day1 problem for Ceph RBD, RGW, CephFS, and Dashboard"},{"line_number":50,"context_line":"  deployment by leveraging `cephadm bootstrap --apply-spec` [7]_"},{"line_number":51,"context_line":"- By default, day2 Ceph operations should be done directly with Ceph"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_28450e18","line":48,"range":{"start_line":47,"start_character":2,"end_line":48,"end_character":14},"in_reply_to":"1f621f24_f992cb9d","updated":"2020-11-04 18:22:10.000000000","message":"This is an issue with transitioning the ownership of services away from tripleo.  If tripleo owns the firewall, it owns the whole thing and assumes it\u0027s managed in such a way.  In order for the firewalls to be properly owned, you\u0027d still have to tell tripleo which ports need to be open on which roles in order for the rules to properly be applied on the correct hosts. We do have an extra rules variable that could be used but would need to be merged with any customer\u0027s values.  It reduces the overall usefullness of letting ceph deploy bits. You could create dummy ceph services that can be used to handle the firewall/haproxy integration and not actually do any \"deployment\" actions.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"4fde3624efcfdcb056261ece50439fdbe4fcc9de","unresolved":false,"context_lines":[{"line_number":52,"context_line":"  orchestrator or Ceph Dashboard and not by running `openstack"},{"line_number":53,"context_line":"  overcloud deploy`"},{"line_number":54,"context_line":"- TripleO stack updates do not trigger the new Ansible roles."},{"line_number":55,"context_line":"- Provide an opinionated Ceph installation based on parameters from"},{"line_number":56,"context_line":"  TripleO (including hardware details from Ironic)"},{"line_number":57,"context_line":"- Configure cephx keyrings and pools for OpenStack on a deployed Ceph"},{"line_number":58,"context_line":"  cluster"},{"line_number":59,"context_line":"- Support collocation (hyperconvergence) of OpenStack/Ceph containers"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_5e09202f","line":56,"range":{"start_line":55,"start_character":0,"end_line":56,"end_character":50},"updated":"2020-11-04 23:33:34.000000000","message":"What about adding new nodes to Ceph cluster?\nDo you still dependent on TripleO and Ironic to get physical nodes into opinionated ready state, including, base platform, before any ansible scripts that use Ceph-adm are run?\nThe same 2 step process defined below but now part of day 2 operation.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"08a177ca12cc7937d55372b3fdeeaaebf0fc0a68","unresolved":false,"context_lines":[{"line_number":52,"context_line":"  orchestrator or Ceph Dashboard and not by running `openstack"},{"line_number":53,"context_line":"  overcloud deploy`"},{"line_number":54,"context_line":"- TripleO stack updates do not trigger the new Ansible roles."},{"line_number":55,"context_line":"- Provide an opinionated Ceph installation based on parameters from"},{"line_number":56,"context_line":"  TripleO (including hardware details from Ironic)"},{"line_number":57,"context_line":"- Configure cephx keyrings and pools for OpenStack on a deployed Ceph"},{"line_number":58,"context_line":"  cluster"},{"line_number":59,"context_line":"- Support collocation (hyperconvergence) of OpenStack/Ceph containers"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_fa9bcfc5","line":56,"range":{"start_line":55,"start_character":0,"end_line":56,"end_character":50},"in_reply_to":"1f621f24_5e09202f","updated":"2020-11-05 13:11:16.000000000","message":"Yes, the process of scaling up/down is tied to TripleO.\nIf step-2 process:\n* before any ansible run we should provision and provide the new resources using metalsmith and then we should be able to place them within the Ceph cluster applying the related Ceph Spec.\n\nIf step-4 process:\n* same story, metalsmith should provide the new hw, the network should be configured and then we can add the new node using the Ceph spec applied to cephadm.\n\nThe difference between the 2-step and 4-step approach is that in the first case the process of scaling-up the ceph cluster happens using the `overcloud deploy command`, instead in the 4-step approach the Ceph spec can be applied (and passed) using the provisioning command.\nHeat will not be the single source of truth for Ceph, and the status of the cluster is maintained by cephadm, for this reason after provisioning the new hw and setting up the network, cephadm should be triggered to add the new resources to the Ceph cluster.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"8cd3e1336cfab3452ab256f7c9414136deb2e7dd","unresolved":false,"context_lines":[{"line_number":52,"context_line":"  orchestrator or Ceph Dashboard and not by running `openstack"},{"line_number":53,"context_line":"  overcloud deploy`"},{"line_number":54,"context_line":"- TripleO stack updates do not trigger the new Ansible roles."},{"line_number":55,"context_line":"- Provide an opinionated Ceph installation based on parameters from"},{"line_number":56,"context_line":"  TripleO (including hardware details from Ironic)"},{"line_number":57,"context_line":"- Configure cephx keyrings and pools for OpenStack on a deployed Ceph"},{"line_number":58,"context_line":"  cluster"},{"line_number":59,"context_line":"- Support collocation (hyperconvergence) of OpenStack/Ceph containers"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_da3a5396","line":56,"range":{"start_line":55,"start_character":0,"end_line":56,"end_character":50},"in_reply_to":"1f621f24_fa9bcfc5","updated":"2020-11-05 13:24:43.000000000","message":"Yes. Scale up is covered on line 515 and 538 of the latest patchset.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b010e00abbfecca48fe2bca84a6d2fff6ef29174","unresolved":false,"context_lines":[{"line_number":87,"context_line":""},{"line_number":88,"context_line":"1. Deploy the hardware with metalsmith [4]_"},{"line_number":89,"context_line":"2. Configure networking (including storage networks)"},{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_6e7f1367","line":90,"updated":"2020-11-04 15:06:46.000000000","message":"... \"with the roles and interface provided by tripleo-ansible/python-tripleoclient\" \n\njust suggestion but please be a bit more explicit ... by what is ceph deployed here.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"acedd099b67cb8d771073bb254f6ffaf4644e7e0","unresolved":false,"context_lines":[{"line_number":87,"context_line":""},{"line_number":88,"context_line":"1. Deploy the hardware with metalsmith [4]_"},{"line_number":89,"context_line":"2. Configure networking (including storage networks)"},{"line_number":90,"context_line":"3. Deploy Ceph"},{"line_number":91,"context_line":"4. Deploy OpenStack with `openstack overcloud deploy`"},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Item 2 above depends on the spec for network data v2 format [11]_"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_febe2c63","line":90,"in_reply_to":"1f621f24_6e7f1367","updated":"2020-11-04 23:32:31.000000000","message":"\u003e ... \"with the roles and interface provided by tripleo-ansible/python-tripleoclient\"\n \u003e \n \u003e just suggestion but please be a bit more explicit ... by what is\n \u003e ceph deployed here.\n\nI added \"with the roles...\" and also changed the wording on why we deploy Ceph at step 3 to line 107 of the next patchset.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b010e00abbfecca48fe2bca84a6d2fff6ef29174","unresolved":false,"context_lines":[{"line_number":102,"context_line":"external_deploy_steps_tasks as it currently does in Victoria via the"},{"line_number":103,"context_line":"2-step-process."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"Additional details on how the 2-step-process and 4-step-process work"},{"line_number":106,"context_line":"are covered in the implementation section. Regardless of which process"},{"line_number":107,"context_line":"is followed, the desired end state of the Ceph cluster to be deployed"},{"line_number":108,"context_line":"will be defined in a new set of yaml files described in more detail in"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_cee2a72f","line":105,"range":{"start_line":105,"start_character":29,"end_line":105,"end_character":61},"updated":"2020-11-04 15:06:46.000000000","message":"please consider changing all references to these below... \n\nsomething like something \"the inside tripleo process\" \u0026 \"outside tripleo process\" is easier for the humans to process imo","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"acedd099b67cb8d771073bb254f6ffaf4644e7e0","unresolved":false,"context_lines":[{"line_number":102,"context_line":"external_deploy_steps_tasks as it currently does in Victoria via the"},{"line_number":103,"context_line":"2-step-process."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"Additional details on how the 2-step-process and 4-step-process work"},{"line_number":106,"context_line":"are covered in the implementation section. Regardless of which process"},{"line_number":107,"context_line":"is followed, the desired end state of the Ceph cluster to be deployed"},{"line_number":108,"context_line":"will be defined in a new set of yaml files described in more detail in"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_3eb9c479","line":105,"range":{"start_line":105,"start_character":29,"end_line":105,"end_character":61},"in_reply_to":"1f621f24_cee2a72f","updated":"2020-11-04 23:32:31.000000000","message":"OK, the next patchset will draw that distinction on if Ceph is deployed _before_ or _during_ the overcloud deployment.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"4fde3624efcfdcb056261ece50439fdbe4fcc9de","unresolved":false,"context_lines":[{"line_number":113,"context_line":""},{"line_number":114,"context_line":"We could ask deployers to do this:"},{"line_number":115,"context_line":""},{"line_number":116,"context_line":"- Deploy hardware and configure networking"},{"line_number":117,"context_line":"- Use cephadm and orchestrator directly to configure that hardware"},{"line_number":118,"context_line":"  with Ceph and create OpenStack pools accessible by CephX clients"},{"line_number":119,"context_line":"- Use TripleO to configure OpenStack"},{"line_number":120,"context_line":""},{"line_number":121,"context_line":"We have completed a POC of the above using Ussuri and config-download"},{"line_number":122,"context_line":"tags to only run certain steps but would prefer to offer an option to"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_c136f5e6","line":119,"range":{"start_line":116,"start_character":0,"end_line":119,"end_character":36},"updated":"2020-11-04 23:33:34.000000000","message":"what if anything will orchestrate these steps?\nOr that is left to user to handle on their own?","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"8cd3e1336cfab3452ab256f7c9414136deb2e7dd","unresolved":false,"context_lines":[{"line_number":113,"context_line":""},{"line_number":114,"context_line":"We could ask deployers to do this:"},{"line_number":115,"context_line":""},{"line_number":116,"context_line":"- Deploy hardware and configure networking"},{"line_number":117,"context_line":"- Use cephadm and orchestrator directly to configure that hardware"},{"line_number":118,"context_line":"  with Ceph and create OpenStack pools accessible by CephX clients"},{"line_number":119,"context_line":"- Use TripleO to configure OpenStack"},{"line_number":120,"context_line":""},{"line_number":121,"context_line":"We have completed a POC of the above using Ussuri and config-download"},{"line_number":122,"context_line":"tags to only run certain steps but would prefer to offer an option to"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_3acd47b8","line":119,"range":{"start_line":116,"start_character":0,"end_line":119,"end_character":36},"in_reply_to":"1f621f24_c136f5e6","updated":"2020-11-05 13:24:43.000000000","message":"In this alternatives section nothing would orchestrate those steps. The user would need to run run cephadm to bootstap a one-node cluster and then use the orchestrator to scale up. That is an alternative if we do nothing. However, I am proposing in this spec that TripleO deploy Ceph.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"4fde3624efcfdcb056261ece50439fdbe4fcc9de","unresolved":false,"context_lines":[{"line_number":192,"context_line":"by default, and the future upgrade path will be provided by these tools [9]_,"},{"line_number":193,"context_line":"which will be able to run, stop and resume all the Ceph upgrade phases."},{"line_number":194,"context_line":"At that point day2 ceph operations will need to be carried out"},{"line_number":195,"context_line":"direclty with ceph orchestrator. Thus, it will no longer be necessary"},{"line_number":196,"context_line":"to include the `tripleo-heat-templates/environments/ceph-ansible/*`"},{"line_number":197,"context_line":"files in the `openstack overcloud deploy` command with the exception"},{"line_number":198,"context_line":"of the Ceph client configuration as described in [13]_, which will"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_21a9a9de","line":195,"range":{"start_line":195,"start_character":0,"end_line":195,"end_character":8},"updated":"2020-11-04 23:33:34.000000000","message":"typo","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"8cd3e1336cfab3452ab256f7c9414136deb2e7dd","unresolved":false,"context_lines":[{"line_number":192,"context_line":"by default, and the future upgrade path will be provided by these tools [9]_,"},{"line_number":193,"context_line":"which will be able to run, stop and resume all the Ceph upgrade phases."},{"line_number":194,"context_line":"At that point day2 ceph operations will need to be carried out"},{"line_number":195,"context_line":"direclty with ceph orchestrator. Thus, it will no longer be necessary"},{"line_number":196,"context_line":"to include the `tripleo-heat-templates/environments/ceph-ansible/*`"},{"line_number":197,"context_line":"files in the `openstack overcloud deploy` command with the exception"},{"line_number":198,"context_line":"of the Ceph client configuration as described in [13]_, which will"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_fad2af5b","line":195,"range":{"start_line":195,"start_character":0,"end_line":195,"end_character":8},"in_reply_to":"1f621f24_21a9a9de","updated":"2020-11-05 13:24:43.000000000","message":"Done","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"4fde3624efcfdcb056261ece50439fdbe4fcc9de","unresolved":false,"context_lines":[{"line_number":230,"context_line":"with TripleO at all:"},{"line_number":231,"context_line":""},{"line_number":232,"context_line":"1. Monitor replacement"},{"line_number":233,"context_line":"2. OSD replacement (if a hardware change is necessary then Ironic"},{"line_number":234,"context_line":"   might be involved)"},{"line_number":235,"context_line":""},{"line_number":236,"context_line":".. note::"},{"line_number":237,"context_line":""}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_01858d5a","line":234,"range":{"start_line":233,"start_character":0,"end_line":234,"end_character":21},"updated":"2020-11-04 23:33:34.000000000","message":"please, include sections for ceph cluster scaling up and down. It will require Ironic \u0026 TripleO interactions so they know which nodes are in use by Ceph and which no longer needed by Ceph.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"8e352d277285d5fc40e4e287ea027c8770674af9","unresolved":false,"context_lines":[{"line_number":230,"context_line":"with TripleO at all:"},{"line_number":231,"context_line":""},{"line_number":232,"context_line":"1. Monitor replacement"},{"line_number":233,"context_line":"2. OSD replacement (if a hardware change is necessary then Ironic"},{"line_number":234,"context_line":"   might be involved)"},{"line_number":235,"context_line":""},{"line_number":236,"context_line":".. note::"},{"line_number":237,"context_line":""}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_9a50dbd6","line":234,"range":{"start_line":233,"start_character":0,"end_line":234,"end_character":21},"in_reply_to":"1f621f24_01858d5a","updated":"2020-11-05 13:24:44.000000000","message":"An interaction w/ Ironic/TripleO is required when a node is added/removed, but in case of OSDs replacement we\u0027re in what is considered day2 operations that don\u0027t necessary requires TripleO interactions anymore. We don\u0027t have a \"well-defined\" list of disks in tripleo-heat-templates because this info is managed by cephadm and this component (with the Ceph Orchestrator) is supposed to keep the status of the cluster.\nSo I think this section is ok (in line 228 you can find what I just said here)","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"8cd3e1336cfab3452ab256f7c9414136deb2e7dd","unresolved":false,"context_lines":[{"line_number":230,"context_line":"with TripleO at all:"},{"line_number":231,"context_line":""},{"line_number":232,"context_line":"1. Monitor replacement"},{"line_number":233,"context_line":"2. OSD replacement (if a hardware change is necessary then Ironic"},{"line_number":234,"context_line":"   might be involved)"},{"line_number":235,"context_line":""},{"line_number":236,"context_line":".. note::"},{"line_number":237,"context_line":""}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_3a426711","line":234,"range":{"start_line":233,"start_character":0,"end_line":234,"end_character":21},"in_reply_to":"1f621f24_01858d5a","updated":"2020-11-05 13:24:43.000000000","message":"Yes, it\u0027s described on lines 538 and 515 of patch set 35.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"6d229beca3e0e10b5983f6bd7c00b01edf663016","unresolved":false,"context_lines":[{"line_number":230,"context_line":"with TripleO at all:"},{"line_number":231,"context_line":""},{"line_number":232,"context_line":"1. Monitor replacement"},{"line_number":233,"context_line":"2. OSD replacement (if a hardware change is necessary then Ironic"},{"line_number":234,"context_line":"   might be involved)"},{"line_number":235,"context_line":""},{"line_number":236,"context_line":".. note::"},{"line_number":237,"context_line":""}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_3a6ee700","line":234,"range":{"start_line":233,"start_character":0,"end_line":234,"end_character":21},"in_reply_to":"1f621f24_7ae7bf13","updated":"2020-11-05 14:08:06.000000000","message":"+1 thanks","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"fdbda0db83afa73a07de9713f334736c77a3bb7b","unresolved":false,"context_lines":[{"line_number":230,"context_line":"with TripleO at all:"},{"line_number":231,"context_line":""},{"line_number":232,"context_line":"1. Monitor replacement"},{"line_number":233,"context_line":"2. OSD replacement (if a hardware change is necessary then Ironic"},{"line_number":234,"context_line":"   might be involved)"},{"line_number":235,"context_line":""},{"line_number":236,"context_line":".. note::"},{"line_number":237,"context_line":""}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_7ae7bf13","line":234,"range":{"start_line":233,"start_character":0,"end_line":234,"end_character":21},"in_reply_to":"1f621f24_9a50dbd6","updated":"2020-11-05 13:38:49.000000000","message":"As per line 538 of PS 35:\n\n- Node scale up operations for day2 Ceph should be done by running\n  `openstack overcloud node provision` and then `openstack overcloud\n  ceph ...`. If that operation includes adding hyperconvernged node\n  with both Ceph and OpenStack services then the third step will be to\n  run `openstack overcloud deploy`.\n\nSo you:\n\n1. add the node with metalsmith (openstack overcloud node provision)\n2. configure networking (openstack overcloud network ...)\n3. add the node to the ceph cluster (openstack overcloud ceph provision)\n4. optionally: if HCI, then add compute containers to the same node \n\nFor step 3, the following will get run unless you override it with a disk list:\n\n ceph orch apply osd --all-available-devices\n\nSo the new node is added and any unused clean disk becomes a new OSD.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"4fde3624efcfdcb056261ece50439fdbe4fcc9de","unresolved":false,"context_lines":[{"line_number":251,"context_line":"won\u0027t be delayed by Ceph operations. Ceph client configuration will"},{"line_number":252,"context_line":"take less time thought his is covered in the spec [10]_."},{"line_number":253,"context_line":""},{"line_number":254,"context_line":"Other Deployer Impact"},{"line_number":255,"context_line":"---------------------"},{"line_number":256,"context_line":""},{"line_number":257,"context_line":"Like ceph-ansible, cephadm is distributed as an RPM and can be installed"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_e1ecd12a","line":254,"range":{"start_line":254,"start_character":0,"end_line":254,"end_character":21},"updated":"2020-11-04 23:33:34.000000000","message":"What about deployment of Ceph Dashboard?\nIf outside the scope of deployment and this proposal let\u0027s state so.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"8cd3e1336cfab3452ab256f7c9414136deb2e7dd","unresolved":false,"context_lines":[{"line_number":251,"context_line":"won\u0027t be delayed by Ceph operations. Ceph client configuration will"},{"line_number":252,"context_line":"take less time thought his is covered in the spec [10]_."},{"line_number":253,"context_line":""},{"line_number":254,"context_line":"Other Deployer Impact"},{"line_number":255,"context_line":"---------------------"},{"line_number":256,"context_line":""},{"line_number":257,"context_line":"Like ceph-ansible, cephadm is distributed as an RPM and can be installed"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_7a4b3f2a","line":254,"range":{"start_line":254,"start_character":0,"end_line":254,"end_character":21},"in_reply_to":"1f621f24_7a985fbf","updated":"2020-11-05 13:24:43.000000000","message":"+1 patchset 35 has a dashboard section on line 603.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"577c497eabd22f91c06891a2a20d80264c6947f5","unresolved":false,"context_lines":[{"line_number":251,"context_line":"won\u0027t be delayed by Ceph operations. Ceph client configuration will"},{"line_number":252,"context_line":"take less time thought his is covered in the spec [10]_."},{"line_number":253,"context_line":""},{"line_number":254,"context_line":"Other Deployer Impact"},{"line_number":255,"context_line":"---------------------"},{"line_number":256,"context_line":""},{"line_number":257,"context_line":"Like ceph-ansible, cephadm is distributed as an RPM and can be installed"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_7a985fbf","line":254,"range":{"start_line":254,"start_character":0,"end_line":254,"end_character":21},"in_reply_to":"1f621f24_e1ecd12a","updated":"2020-11-05 13:14:15.000000000","message":"The Ceph Dashboard is not outside this proposal.\nI\u0027m going to update the spec describing how the Ceph Dashboard will be deployed.\nThanks for pointing that out.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b010e00abbfecca48fe2bca84a6d2fff6ef29174","unresolved":false,"context_lines":[{"line_number":298,"context_line":""},{"line_number":299,"context_line":"The current Heat parameters which describe the end state of the Ceph"},{"line_number":300,"context_line":"cluster will be deprecated and a new format will be introduced. How"},{"line_number":301,"context_line":"this format is passed to both the 2-step and 4-step processes,"},{"line_number":302,"context_line":"described in the Integration Points section of the beginning of this"},{"line_number":303,"context_line":"spec, will be covered in more detail in this section."},{"line_number":304,"context_line":""}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_4ed05751","line":301,"range":{"start_line":301,"start_character":34,"end_line":301,"end_character":40},"updated":"2020-11-04 15:06:46.000000000","message":"\u0027inside\u0027 or \u0027outside\u0027 the tripleo openstack deployment","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"acedd099b67cb8d771073bb254f6ffaf4644e7e0","unresolved":false,"context_lines":[{"line_number":298,"context_line":""},{"line_number":299,"context_line":"The current Heat parameters which describe the end state of the Ceph"},{"line_number":300,"context_line":"cluster will be deprecated and a new format will be introduced. How"},{"line_number":301,"context_line":"this format is passed to both the 2-step and 4-step processes,"},{"line_number":302,"context_line":"described in the Integration Points section of the beginning of this"},{"line_number":303,"context_line":"spec, will be covered in more detail in this section."},{"line_number":304,"context_line":""}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_e13a1196","line":301,"range":{"start_line":301,"start_character":34,"end_line":301,"end_character":40},"in_reply_to":"1f621f24_4ed05751","updated":"2020-11-04 23:32:31.000000000","message":"Done","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b010e00abbfecca48fe2bca84a6d2fff6ef29174","unresolved":false,"context_lines":[{"line_number":339,"context_line":"Ceph End State Definition YAML Input"},{"line_number":340,"context_line":"------------------------------------"},{"line_number":341,"context_line":""},{"line_number":342,"context_line":"Regardless of if the 2-step or 4-step process is followed, a new"},{"line_number":343,"context_line":"playbook which deploys Ceph using cephadm will be created and it"},{"line_number":344,"context_line":"will accept the following files as input:"},{"line_number":345,"context_line":""}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_0ed65f50","line":342,"range":{"start_line":342,"start_character":22,"end_line":342,"end_character":26},"updated":"2020-11-04 15:06:46.000000000","message":"\u0027inside or with tripleo\u0027 or \u0027outside or before tripleo\u0027 process is followed (just suggestion easier to follow for me anyway )","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"acedd099b67cb8d771073bb254f6ffaf4644e7e0","unresolved":false,"context_lines":[{"line_number":339,"context_line":"Ceph End State Definition YAML Input"},{"line_number":340,"context_line":"------------------------------------"},{"line_number":341,"context_line":""},{"line_number":342,"context_line":"Regardless of if the 2-step or 4-step process is followed, a new"},{"line_number":343,"context_line":"playbook which deploys Ceph using cephadm will be created and it"},{"line_number":344,"context_line":"will accept the following files as input:"},{"line_number":345,"context_line":""}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_0138ed9d","line":342,"range":{"start_line":342,"start_character":22,"end_line":342,"end_character":26},"in_reply_to":"1f621f24_0ed65f50","updated":"2020-11-04 23:32:31.000000000","message":"Done","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b010e00abbfecca48fe2bca84a6d2fff6ef29174","unresolved":false,"context_lines":[{"line_number":503,"context_line":"  ContainerCephDaemonImage: undercloud.ctlplane.mydomain.tld:8787/ceph-ci/daemon:v4.0.13-stable-4.0-nautilus-centos-7-x86_64"},{"line_number":504,"context_line":""},{"line_number":505,"context_line":""},{"line_number":506,"context_line":"Properties of the 2-step-process"},{"line_number":507,"context_line":"--------------------------------"},{"line_number":508,"context_line":""},{"line_number":509,"context_line":"- The external_deploy_steps_tasks playbook will execute the new"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_4e5e97e6","line":506,"range":{"start_line":506,"start_character":20,"end_line":506,"end_character":24},"updated":"2020-11-04 15:06:46.000000000","message":"so to be clear as outlined in the schedule below, for wallaby we will support the current way of deploying Ceph. That is  \u0027deploy inside the tripleo openstack deployment\u0027. At the same time you will develop the \u0027deploy before/outside tripleo openstack with tripleo-ceph\u0027/4 step process. If it is stable we can adopt it at the end of W possibly.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"acedd099b67cb8d771073bb254f6ffaf4644e7e0","unresolved":false,"context_lines":[{"line_number":503,"context_line":"  ContainerCephDaemonImage: undercloud.ctlplane.mydomain.tld:8787/ceph-ci/daemon:v4.0.13-stable-4.0-nautilus-centos-7-x86_64"},{"line_number":504,"context_line":""},{"line_number":505,"context_line":""},{"line_number":506,"context_line":"Properties of the 2-step-process"},{"line_number":507,"context_line":"--------------------------------"},{"line_number":508,"context_line":""},{"line_number":509,"context_line":"- The external_deploy_steps_tasks playbook will execute the new"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_215ec9e5","line":506,"range":{"start_line":506,"start_character":20,"end_line":506,"end_character":24},"in_reply_to":"1f621f24_4e5e97e6","updated":"2020-11-04 23:32:31.000000000","message":"\u003e so to be clear as outlined in the schedule below, for wallaby we\n \u003e will support the current way of deploying Ceph. That is  \u0027deploy\n \u003e inside the tripleo openstack deployment\u0027. At the same time you will\n \u003e develop the \u0027deploy before/outside tripleo openstack with\n \u003e tripleo-ceph\u0027/4 step process. If it is stable we can adopt it at\n \u003e the end of W possibly.\n\nYes, exactly.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b010e00abbfecca48fe2bca84a6d2fff6ef29174","unresolved":false,"context_lines":[{"line_number":698,"context_line":".. [8] `cephadm new host \u003chttps://docs.ceph.com/docs/master/cephadm/install/#add-hosts-to-the-cluster\u003e`_"},{"line_number":699,"context_line":".. [9] `cephadm cluster upgrade \u003chttps://docs.ceph.com/docs/master/cephadm/upgrade\u003e`_"},{"line_number":700,"context_line":".. [10] `tripleo-ansible ceph-client role \u003chttps://review.opendev.org/#/c/757644\u003e`_"},{"line_number":701,"context_line":".. [11] `Spec for network data v2 format \u003chttps://review.opendev.org/#/c/752437\u003e`_"},{"line_number":702,"context_line":".. [12] `openstack overcloud network provision \u003chttps://review.opendev.org/751875\u003e`_"},{"line_number":703,"context_line":".. [13] `Spec for tripleo-ceph-client \u003chttps://review.opendev.org/#/c/757644\u003e`_"},{"line_number":704,"context_line":".. [14] `Network Data v2 - node ports spec \u003chttps://review.opendev.org/760536\u003e`_"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_ce69078c","line":701,"range":{"start_line":701,"start_character":2,"end_line":701,"end_character":3},"updated":"2020-11-04 15:06:46.000000000","message":"these look a bit weird in the build see https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_3a5/723108/34/check/openstack-tox-docs/3a525b5/docs/specs/wallaby/tripleo-ceph.html#id49","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"35fa546406b1671e9d8c8a1eb70017251bd89a1d","unresolved":false,"context_lines":[{"line_number":698,"context_line":".. [8] `cephadm new host \u003chttps://docs.ceph.com/docs/master/cephadm/install/#add-hosts-to-the-cluster\u003e`_"},{"line_number":699,"context_line":".. [9] `cephadm cluster upgrade \u003chttps://docs.ceph.com/docs/master/cephadm/upgrade\u003e`_"},{"line_number":700,"context_line":".. [10] `tripleo-ansible ceph-client role \u003chttps://review.opendev.org/#/c/757644\u003e`_"},{"line_number":701,"context_line":".. [11] `Spec for network data v2 format \u003chttps://review.opendev.org/#/c/752437\u003e`_"},{"line_number":702,"context_line":".. [12] `openstack overcloud network provision \u003chttps://review.opendev.org/751875\u003e`_"},{"line_number":703,"context_line":".. [13] `Spec for tripleo-ceph-client \u003chttps://review.opendev.org/#/c/757644\u003e`_"},{"line_number":704,"context_line":".. [14] `Network Data v2 - node ports spec \u003chttps://review.opendev.org/760536\u003e`_"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_73e24531","line":701,"range":{"start_line":701,"start_character":2,"end_line":701,"end_character":3},"in_reply_to":"1f621f24_be815492","updated":"2020-11-05 23:14:57.000000000","message":"I used Indirect Hyperlink Targets in place of the footnotes and  the next build won\u0027t look weird.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"acedd099b67cb8d771073bb254f6ffaf4644e7e0","unresolved":false,"context_lines":[{"line_number":698,"context_line":".. [8] `cephadm new host \u003chttps://docs.ceph.com/docs/master/cephadm/install/#add-hosts-to-the-cluster\u003e`_"},{"line_number":699,"context_line":".. [9] `cephadm cluster upgrade \u003chttps://docs.ceph.com/docs/master/cephadm/upgrade\u003e`_"},{"line_number":700,"context_line":".. [10] `tripleo-ansible ceph-client role \u003chttps://review.opendev.org/#/c/757644\u003e`_"},{"line_number":701,"context_line":".. [11] `Spec for network data v2 format \u003chttps://review.opendev.org/#/c/752437\u003e`_"},{"line_number":702,"context_line":".. [12] `openstack overcloud network provision \u003chttps://review.opendev.org/751875\u003e`_"},{"line_number":703,"context_line":".. [13] `Spec for tripleo-ceph-client \u003chttps://review.opendev.org/#/c/757644\u003e`_"},{"line_number":704,"context_line":".. [14] `Network Data v2 - node ports spec \u003chttps://review.opendev.org/760536\u003e`_"}],"source_content_type":"text/x-rst","patch_set":34,"id":"1f621f24_be815492","line":701,"range":{"start_line":701,"start_character":2,"end_line":701,"end_character":3},"in_reply_to":"1f621f24_ce69078c","updated":"2020-11-04 23:32:31.000000000","message":"In the render, when there\u0027s \u003e1 reference to a footnote, in place of the footnote number there\u0027s a count 1,2,3,etc for each back reference. Perhaps if I make the references match 1:1 then that won\u0027t happen. I\u0027ll try that in a future revision.","commit_id":"524eef2b837286a79fc129f947a832a75863672c"},{"author":{"_account_id":7160,"name":"arkady kanevsky","email":"akanevsk@redhat.com","username":"arkady"},"change_message_id":"34d5cb6974329ed4ebff91e0059b0580f3382f83","unresolved":false,"context_lines":[{"line_number":626,"context_line":"service as is while also providing an option to use the dashboard"},{"line_number":627,"context_line":"without HA."},{"line_number":628,"context_line":""},{"line_number":629,"context_line":"Provided that the work described in review 760536_ allows us "},{"line_number":630,"context_line":"to know the VIP in advance, then that should be sufficient to deploy"},{"line_number":631,"context_line":"the Ceph dashboard as we do today. The automation to stand up Ceph can"},{"line_number":632,"context_line":"be passed the VIP so that the Ceph dashboard will use it after"}],"source_content_type":"text/x-rst","patch_set":36,"id":"1f621f24_8e6418e6","line":629,"range":{"start_line":629,"start_character":60,"end_line":629,"end_character":61},"updated":"2020-11-06 01:21:02.000000000","message":"unprintable char at the end of the line.","commit_id":"15a88a74cb2adf19f0e8b5055027cad74c8da176"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"9045985b4e6a17ccf38f7dad9142030bf1f37425","unresolved":false,"context_lines":[{"line_number":626,"context_line":"service as is while also providing an option to use the dashboard"},{"line_number":627,"context_line":"without HA."},{"line_number":628,"context_line":""},{"line_number":629,"context_line":"Provided that the work described in review 760536_ allows us "},{"line_number":630,"context_line":"to know the VIP in advance, then that should be sufficient to deploy"},{"line_number":631,"context_line":"the Ceph dashboard as we do today. The automation to stand up Ceph can"},{"line_number":632,"context_line":"be passed the VIP so that the Ceph dashboard will use it after"}],"source_content_type":"text/x-rst","patch_set":36,"id":"1f621f24_010d5fa0","line":629,"range":{"start_line":629,"start_character":60,"end_line":629,"end_character":61},"in_reply_to":"1f621f24_8e6418e6","updated":"2020-11-09 19:33:16.000000000","message":"Done","commit_id":"15a88a74cb2adf19f0e8b5055027cad74c8da176"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"67c151e9813bec15abe07592cfa9c138c58a3b77","unresolved":false,"context_lines":[{"line_number":64,"context_line":"  - TripleO configuration updates must not break Ceph configuration"},{"line_number":65,"context_line":"- Provide Ceph integration but maximize orthogonality between"},{"line_number":66,"context_line":"  OpenStack and Ceph"},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"The implementation of the TripleO CephClient service during the W"},{"line_number":69,"context_line":"cycle is covered in a different spec in review 757644_. This work will"},{"line_number":70,"context_line":"be merged before the work described in this spec as it will be"}],"source_content_type":"text/x-rst","patch_set":38,"id":"1f621f24_6ae86de1","line":67,"updated":"2020-11-09 17:18:11.000000000","message":"+1 nice list","commit_id":"f84116a1a80220485ccb8182aa305705ff40e86c"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"67c151e9813bec15abe07592cfa9c138c58a3b77","unresolved":false,"context_lines":[{"line_number":321,"context_line":"one release we will have intree support both ceph-ansible_ and"},{"line_number":322,"context_line":"cephadm_."},{"line_number":323,"context_line":""},{"line_number":324,"context_line":"New THT Parameters"},{"line_number":325,"context_line":"------------------"},{"line_number":326,"context_line":""},{"line_number":327,"context_line":"Not all THT configuration for Ceph can be removed. The firewall is"}],"source_content_type":"text/x-rst","patch_set":38,"id":"1f621f24_aaaa856d","line":324,"updated":"2020-11-09 17:18:11.000000000","message":"this might be misleading, we don\u0027t introduce new parameter, rather try to reuse the existing but we do introduce new templates","commit_id":"f84116a1a80220485ccb8182aa305705ff40e86c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"9045985b4e6a17ccf38f7dad9142030bf1f37425","unresolved":false,"context_lines":[{"line_number":321,"context_line":"one release we will have intree support both ceph-ansible_ and"},{"line_number":322,"context_line":"cephadm_."},{"line_number":323,"context_line":""},{"line_number":324,"context_line":"New THT Parameters"},{"line_number":325,"context_line":"------------------"},{"line_number":326,"context_line":""},{"line_number":327,"context_line":"Not all THT configuration for Ceph can be removed. The firewall is"}],"source_content_type":"text/x-rst","patch_set":38,"id":"1f621f24_21917b48","line":324,"in_reply_to":"1f621f24_aaaa856d","updated":"2020-11-09 19:33:16.000000000","message":"OK, I\u0027ll change it to \"New THT Templates\" in the next PS.","commit_id":"f84116a1a80220485ccb8182aa305705ff40e86c"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"601becfcd462be1f71a843aa54882681e89865d0","unresolved":false,"context_lines":[{"line_number":358,"context_line":"described in the Ceph End State Definition YAML Input section and"},{"line_number":359,"context_line":"cephadm.yaml will only contain references to those files."},{"line_number":360,"context_line":""},{"line_number":361,"context_line":"The cephx keys and pools, created as described above, will result in"},{"line_number":362,"context_line":"output data which looks like the following::"},{"line_number":363,"context_line":""},{"line_number":364,"context_line":"  pools:"}],"source_content_type":"text/x-rst","patch_set":38,"id":"1f621f24_c7fa5273","line":361,"range":{"start_line":361,"start_character":4,"end_line":361,"end_character":24},"updated":"2020-11-10 08:55:32.000000000","message":"+1 thanks","commit_id":"f84116a1a80220485ccb8182aa305705ff40e86c"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"601becfcd462be1f71a843aa54882681e89865d0","unresolved":false,"context_lines":[{"line_number":603,"context_line":"undercloud. This registry contains Ceph and OpenStack containers"},{"line_number":604,"context_line":"and it may be populated before deployment or during deployment."},{"line_number":605,"context_line":"When deploying ceph before overcloud deployment it will need to be"},{"line_number":606,"context_line":"populated before deployment. The new integration described in this"},{"line_number":607,"context_line":"spec will direct cephadm_ to pull the Ceph containers from the same"},{"line_number":608,"context_line":"source identified by `ContainerCephDaemonImage`. For example::"},{"line_number":609,"context_line":""}],"source_content_type":"text/x-rst","patch_set":38,"id":"1f621f24_6737fe4e","line":606,"range":{"start_line":606,"start_character":10,"end_line":606,"end_character":27},"updated":"2020-11-10 08:55:32.000000000","message":"Guess we should run:\n\nopenstack tripleo container image prepare default \\\n  --output-env-file $HOME/containers-prepare-parameters.yaml\n\nand then trigger trigger [1]\n\n[1] https://github.com/openstack/tripleo-ansible/tree/master/tripleo_ansible/roles/tripleo_container_image_prepare","commit_id":"f84116a1a80220485ccb8182aa305705ff40e86c"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"67c151e9813bec15abe07592cfa9c138c58a3b77","unresolved":false,"context_lines":[{"line_number":607,"context_line":"spec will direct cephadm_ to pull the Ceph containers from the same"},{"line_number":608,"context_line":"source identified by `ContainerCephDaemonImage`. For example::"},{"line_number":609,"context_line":""},{"line_number":610,"context_line":"  ContainerCephDaemonImage: undercloud.ctlplane.mydomain.tld:8787/ceph-ci/daemon:v4.0.13-stable-4.0-nautilus-centos-7-x86_64"},{"line_number":611,"context_line":""},{"line_number":612,"context_line":"Network Requirements for Ceph to be deployed before the Overcloud"},{"line_number":613,"context_line":"-----------------------------------------------------------------"}],"source_content_type":"text/x-rst","patch_set":38,"id":"1f621f24_4acad14d","line":610,"updated":"2020-11-09 17:18:11.000000000","message":"this is what happens today for ceph-ansible as well; we\u0027d like to deprecate special ceph tags for container prepare but that\u0027s not part of this spec","commit_id":"f84116a1a80220485ccb8182aa305705ff40e86c"},{"author":{"_account_id":6796,"name":"Giulio Fidente","email":"gfidente@redhat.com","username":"gfidente"},"change_message_id":"67c151e9813bec15abe07592cfa9c138c58a3b77","unresolved":false,"context_lines":[{"line_number":635,"context_line":"Ceph services which need VIPs (Dashbard and RGW) need to know what the"},{"line_number":636,"context_line":"VIPs will be in advance but the VIPs do not need to be pingable before"},{"line_number":637,"context_line":"those Ceph services are deployed. Instead the VIPs are passed as input"},{"line_number":638,"context_line":"to cephadm. The VIPs for these services will not be managed by"},{"line_number":639,"context_line":"OpenStack pid1 services and will instead be configured by cephadm."},{"line_number":640,"context_line":""},{"line_number":641,"context_line":"Development work on for deploying Ceph before overcloud deployment"}],"source_content_type":"text/x-rst","patch_set":38,"id":"1f621f24_4acf7155","line":638,"updated":"2020-11-09 17:18:11.000000000","message":"I don\u0027t think we can say this; keepalived and pacemaker can\u0027t live together on the same node","commit_id":"f84116a1a80220485ccb8182aa305705ff40e86c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"9045985b4e6a17ccf38f7dad9142030bf1f37425","unresolved":false,"context_lines":[{"line_number":635,"context_line":"Ceph services which need VIPs (Dashbard and RGW) need to know what the"},{"line_number":636,"context_line":"VIPs will be in advance but the VIPs do not need to be pingable before"},{"line_number":637,"context_line":"those Ceph services are deployed. Instead the VIPs are passed as input"},{"line_number":638,"context_line":"to cephadm. The VIPs for these services will not be managed by"},{"line_number":639,"context_line":"OpenStack pid1 services and will instead be configured by cephadm."},{"line_number":640,"context_line":""},{"line_number":641,"context_line":"Development work on for deploying Ceph before overcloud deployment"}],"source_content_type":"text/x-rst","patch_set":38,"id":"1f621f24_81ca0f4b","line":638,"in_reply_to":"1f621f24_4acf7155","updated":"2020-11-09 19:33:16.000000000","message":"Good point. We should say something about the two not mixing in this spec. What though would we say? \n\nkeepalived is being configured by cephadm for HA RGW only. We could require that this service not be deployed on controllers anymore. Any reason this wouldn\u0027t work in theory?\n\nAre you saying we go back to letting TripleO configure these services w/ PCS in HA?","commit_id":"f84116a1a80220485ccb8182aa305705ff40e86c"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"219b2f01699710f6ec511ce460e185af77f31397","unresolved":false,"context_lines":[{"line_number":635,"context_line":"Ceph services which need VIPs (Dashbard and RGW) need to know what the"},{"line_number":636,"context_line":"VIPs will be in advance but the VIPs do not need to be pingable before"},{"line_number":637,"context_line":"those Ceph services are deployed. Instead the VIPs are passed as input"},{"line_number":638,"context_line":"to cephadm. The VIPs for these services will not be managed by"},{"line_number":639,"context_line":"OpenStack pid1 services and will instead be configured by cephadm."},{"line_number":640,"context_line":""},{"line_number":641,"context_line":"Development work on for deploying Ceph before overcloud deployment"}],"source_content_type":"text/x-rst","patch_set":38,"id":"1f621f24_a698efbf","line":638,"in_reply_to":"1f621f24_81ca0f4b","updated":"2020-11-10 19:03:02.000000000","message":"As per our conversation this might be the most practical way for now. I\u0027ve updated the spec accordingly.","commit_id":"f84116a1a80220485ccb8182aa305705ff40e86c"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b0bf520ec03d1df8f1cbaa2905504da49b8a2c95","unresolved":false,"context_lines":[{"line_number":325,"context_line":"-----------------"},{"line_number":326,"context_line":""},{"line_number":327,"context_line":"Not all THT configuration for Ceph can be removed. The firewall is"},{"line_number":328,"context_line":"still configured based on THT as descrbed in the next section and THT"},{"line_number":329,"context_line":"also controls which composable service is deployed and where. The"},{"line_number":330,"context_line":"following new files will be created in"},{"line_number":331,"context_line":"`tripleo-heat-templates/environments/`:"}],"source_content_type":"text/x-rst","patch_set":39,"id":"1f621f24_fd124d61","line":328,"updated":"2020-11-13 08:00:45.000000000","message":"i","commit_id":"778d562654e86ad54e28656db1fd0d35360a98dd"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b0bf520ec03d1df8f1cbaa2905504da49b8a2c95","unresolved":false,"context_lines":[{"line_number":390,"context_line":"--------"},{"line_number":391,"context_line":""},{"line_number":392,"context_line":"Today the firewall is not configured by ceph-ansible_ and it won\u0027t be"},{"line_number":393,"context_line":"configured by cephadm_ as its `--skip-firewalld` will be used. We"},{"line_number":394,"context_line":"expect the default overcloud to not have firewall rules until"},{"line_number":395,"context_line":"`openstack overcloud deploy` introduces them. The THT parameters"},{"line_number":396,"context_line":"described in the previous section will have the same firewall ports as"},{"line_number":397,"context_line":"the ones they will deprecate (`environments/ceph-ansible/*`) so that"},{"line_number":398,"context_line":"the appropriate ports per service and based on composable roles will"}],"source_content_type":"text/x-rst","patch_set":39,"id":"1f621f24_dd15d14a","line":395,"range":{"start_line":393,"start_character":63,"end_line":395,"end_character":27},"updated":"2020-11-13 08:00:45.000000000","message":"I get what you are saying but that sentence doesn\u0027t make sense... \u0027until the deploy introduces them\u0027... well there won\u0027t be firewall config or _anything_ else until we deploy the overcloud... I guess you mean more like \"The firewall will be configured during the tripleo overcloud deploy as was the case before\"","commit_id":"778d562654e86ad54e28656db1fd0d35360a98dd"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"b0bf520ec03d1df8f1cbaa2905504da49b8a2c95","unresolved":false,"context_lines":[{"line_number":423,"context_line":"Ceph End State Definition YAML Input"},{"line_number":424,"context_line":"------------------------------------"},{"line_number":425,"context_line":""},{"line_number":426,"context_line":"Regardless of if Ceph is deployed *before* or *during* overcloud"},{"line_number":427,"context_line":"deployment, a new playbook which deploys Ceph using cephadm_ will be"},{"line_number":428,"context_line":"created and it will accept the following files as input:"},{"line_number":429,"context_line":""}],"source_content_type":"text/x-rst","patch_set":39,"id":"1f621f24_3d104557","line":426,"range":{"start_line":426,"start_character":34,"end_line":426,"end_character":54},"updated":"2020-11-13 08:00:45.000000000","message":"thanks this makes it much easier to read :)","commit_id":"778d562654e86ad54e28656db1fd0d35360a98dd"},{"author":{"_account_id":25402,"name":"Francesco Pantano","email":"fpantano@redhat.com","username":"fmount"},"change_message_id":"d6c5a5cef02567beee36e3715af30a97f26ae211","unresolved":false,"context_lines":[{"line_number":664,"context_line":""},{"line_number":665,"context_line":"Why can\u0027t we fully deploy the HA RGW service before the overcloud?"},{"line_number":666,"context_line":"Though cephadm_ can deploy an HA RGW service without TripleO its"},{"line_number":667,"context_line":"implementation uses keepalived which cannot be collocated with"},{"line_number":668,"context_line":"pacemaker, which is required on controller nodes. Thus, during the"},{"line_number":669,"context_line":"W cycle we will keep using the RGW service with haproxy and revisit"},{"line_number":670,"context_line":"making it a separate deployment with collaboration with the PID1 team"},{"line_number":671,"context_line":"in a future cycle."}],"source_content_type":"text/x-rst","patch_set":39,"id":"1f621f24_dc54d300","line":668,"range":{"start_line":667,"start_character":15,"end_line":668,"end_character":49},"updated":"2020-11-11 09:42:45.000000000","message":"I\u0027m ok with this plan, at least for W. We can continue supporting the existing features (HA and TLS) by configuring the services at the overcloud deployment stage (using the existing interfaces in tht).\nAs per conversation w/ pid1 people, keepalived+haproxy can live in the overcloud (collocated with pacemaker), but still, we should continue providing HA for the dashboard using the legacy approach just because it\u0027s not yet supported by cephadm.\nWe can propose to mark the HA + TLS-e integration as deprecated when this feature (in Ceph) is ready (X cycle?) and move this config outside the overcloud deployment.","commit_id":"778d562654e86ad54e28656db1fd0d35360a98dd"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"44529fee0d8cb3b84ecbb23cd172c841d9ec78d9","unresolved":true,"context_lines":[{"line_number":33,"context_line":"consistent with the way the Ceph project is moving and decouple the"},{"line_number":34,"context_line":"complexity of day2 management of Ceph from TripleO."},{"line_number":35,"context_line":""},{"line_number":36,"context_line":".. _proposed-change:"},{"line_number":37,"context_line":""},{"line_number":38,"context_line":"Proposed Change"},{"line_number":39,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":41,"id":"19818f48_830cba1c","line":36,"range":{"start_line":36,"start_character":0,"end_line":36,"end_character":1},"updated":"2020-12-02 15:52:56.000000000","message":"this is causing the tox job to fail \n\nhttps://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_9e8/723108/41/check/openstack-tox-docs/9e87984/job-output.txt\n\n2020-12-01 16:28:12.139581 | ubuntu-focal | Warning, treated as error:\n2020-12-01 16:28:12.139593 | ubuntu-focal | /home/zuul/src/opendev.org/openstack/tripleo-specs/doc/source/specs/wallaby/tripleo-ceph-client.rst:36:duplicate label proposed-change, other instance in /home/zuul/src/opendev.org/openstack/tripleo-specs/doc/source/specs/wallaby/tripleo-ceph.rst\n2020-12-01 16:28:12.355572 | ubuntu-focal | ERROR: InvocationError for command /home/zuul/src/opendev.org/openstack/tripleo-specs/.tox/docs/bin/sphinx-build -W -b html doc/source doc/build/html (exited with code 2)\n2020-12-01 16:28:12.355646 | ubuntu-focal | docs finish: run-test  after 6.53 seconds\n2020-12-01 16:28:12.355984 | ubuntu-focal | docs start: run-test-post\n2020-12-01 16:28:12.356008 | ubuntu-focal | docs finish: run-test-post  after 0.00 seconds\n2020-12-01 16:28:12.356264 | ubuntu-focal | ___________________________________ summary ____________________________________\n2020-12-01 16:28:12.356290 | ubuntu-focal | ERROR:   docs: commands failed\n2020-12-01 16:28:12.356464 | ubuntu-focal | /home/zuul/.local/tox/lib/python3.8/site-packages/tox/config/__init__.py:642: UserWarning: conflicting basepython version (set 3.8, should be 3.6) for env \u0027py36\u0027;resolve conflict or set ignore_basepython_conflict\n2020-12-01 16:28:12.356485 | ubuntu-focal |   warnings.warn(\n2020-12-01 16:28:12.629170 | ubuntu-focal | ERROR\n2020-12-01 16:28:12.629497 | ubuntu-focal | {\n2020-12-01 16:28:12.629600 | ubuntu-focal |   \"delta\": \"0:00:07.934589\",\n2020-12-01 16:28:12.629693 | ubuntu-focal |   \"end\": \"2020-12-01 16:28:12.377262\",\n2020-12-01 16:28:12.629785 | ubuntu-focal |   \"msg\": \"non-zero return code\",\n2020-12-01 16:28:12.629874 | ubuntu-focal |   \"rc\": 1,\n2020-12-01 16:28:12.629964 | ubuntu-focal |   \"start\": \"2020-12-01 16:28:04.442673\"\n2020-12-01 16:28:12.630052 | ubuntu-focal | }","commit_id":"0f6a041d904483c2ae5a8a1d384b0c166c255f16"},{"author":{"_account_id":18002,"name":"John Fulton","email":"fulton@redhat.com","username":"fultonj"},"change_message_id":"e59fa7f1a40294545d192e397891d6827637bb59","unresolved":true,"context_lines":[{"line_number":33,"context_line":"consistent with the way the Ceph project is moving and decouple the"},{"line_number":34,"context_line":"complexity of day2 management of Ceph from TripleO."},{"line_number":35,"context_line":""},{"line_number":36,"context_line":".. _proposed-change:"},{"line_number":37,"context_line":""},{"line_number":38,"context_line":"Proposed Change"},{"line_number":39,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":41,"id":"fcddafe8_3c009768","line":36,"range":{"start_line":36,"start_character":0,"end_line":36,"end_character":1},"in_reply_to":"19818f48_830cba1c","updated":"2020-12-02 16:02:34.000000000","message":"Thanks Marios. I removed it in the most current patch set.","commit_id":"0f6a041d904483c2ae5a8a1d384b0c166c255f16"}]}