)]}'
{"specs/train/approved/support-placement-api.rst":[{"author":{"_account_id":13111,"name":"Hidekazu Nakamura","email":"hidekazuna@gmail.com","username":"hid-nakamura"},"change_message_id":"f49db2c2ab5dfb5e49d7d9fe186e98a5eb68ac37","unresolved":false,"context_lines":[{"line_number":19,"context_line":"within the nova repository and extracted to the placement repository"},{"line_number":20,"context_line":"in the Stein release."},{"line_number":21,"context_line":"Now Watcher gets data directly from Nova for building data model"},{"line_number":22,"context_line":"and strategies processing. Watcher can get more data from Placement"},{"line_number":23,"context_line":"such as cpu allocation ratio and improve the processing of strategy."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"Use Cases"},{"line_number":26,"context_line":"----------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_ad1c6a6a","line":23,"range":{"start_line":22,"start_character":26,"end_line":23,"end_character":68},"updated":"2019-05-09 03:01:17.000000000","message":"Problem is not clear for me. Is that current compute data model does not have enough information the problem?\nI think now that nova delegate resource usage information to the Placement API so that we simply should use Placement API for building data model. Moreover, we can get more information from Placement API such as cpu allocation ratio and even allocation candidates so that we would improve current strategies.\nI am not sure this spec or future specs include substituting nova API with placement API for building data model.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":21692,"name":"licanwei","email":"li.canwei2@zte.com.cn","username":"licanwei"},"change_message_id":"3c59a4eb1def66717e153fb9c7fea304a3c08878","unresolved":false,"context_lines":[{"line_number":19,"context_line":"within the nova repository and extracted to the placement repository"},{"line_number":20,"context_line":"in the Stein release."},{"line_number":21,"context_line":"Now Watcher gets data directly from Nova for building data model"},{"line_number":22,"context_line":"and strategies processing. Watcher can get more data from Placement"},{"line_number":23,"context_line":"such as cpu allocation ratio and improve the processing of strategy."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"Use Cases"},{"line_number":26,"context_line":"----------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_d5a0d644","line":23,"range":{"start_line":22,"start_character":26,"end_line":23,"end_character":68},"in_reply_to":"dfbec78f_20b497d0","updated":"2019-05-16 03:00:25.000000000","message":"OK, I\u0027ll update","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":29911,"name":"Dantali0n","email":"info@dantalion.nl","username":"Dantali0n"},"change_message_id":"68933ad2c42ba788f912a08f05a6ecd6e6bd470a","unresolved":false,"context_lines":[{"line_number":19,"context_line":"within the nova repository and extracted to the placement repository"},{"line_number":20,"context_line":"in the Stein release."},{"line_number":21,"context_line":"Now Watcher gets data directly from Nova for building data model"},{"line_number":22,"context_line":"and strategies processing. Watcher can get more data from Placement"},{"line_number":23,"context_line":"such as cpu allocation ratio and improve the processing of strategy."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"Use Cases"},{"line_number":26,"context_line":"----------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_20b497d0","line":23,"range":{"start_line":22,"start_character":26,"end_line":23,"end_character":68},"in_reply_to":"dfbec78f_8d188575","updated":"2019-05-15 15:47:24.000000000","message":"The problem description is slightly unclear. Maybe add something like \u0027The knowledge available to strategies to make informed decisions is currently limited due to the placement api being unavailable.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":21692,"name":"licanwei","email":"li.canwei2@zte.com.cn","username":"licanwei"},"change_message_id":"bfd0c3e53fa8f6b36b8ec4358670e35a8dc4083b","unresolved":false,"context_lines":[{"line_number":19,"context_line":"within the nova repository and extracted to the placement repository"},{"line_number":20,"context_line":"in the Stein release."},{"line_number":21,"context_line":"Now Watcher gets data directly from Nova for building data model"},{"line_number":22,"context_line":"and strategies processing. Watcher can get more data from Placement"},{"line_number":23,"context_line":"such as cpu allocation ratio and improve the processing of strategy."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"Use Cases"},{"line_number":26,"context_line":"----------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_8d188575","line":23,"range":{"start_line":22,"start_character":26,"end_line":23,"end_character":68},"in_reply_to":"dfbec78f_ad1c6a6a","updated":"2019-05-09 05:57:52.000000000","message":"Watcher can\u0027t get cpu or ram allocation ratio, they are config option in nova.conf.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"fcd7991909ad0370bbd61d03413796c75e5ee9c6","unresolved":false,"context_lines":[{"line_number":31,"context_line":"Proposed change"},{"line_number":32,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"In this spec we just add Placement helper to Watcher."},{"line_number":35,"context_line":"We plan to improve the data model and strategies in"},{"line_number":36,"context_line":"the future specs."},{"line_number":37,"context_line":"The functions in the helper as below:"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_a62b44e9","line":34,"updated":"2019-05-07 18:49:44.000000000","message":"How do you plan on doing this? There is no python-placementclient like there is a python-*client for all of the other services that watcher uses. Nova just uses keystoneauth1 requests for working with placement:\n\nhttps://github.com/openstack/nova/blob/master/nova/scheduler/client/report.py\n\nI suppose you could do similar, or better yet use the openstacksdk. Nova is talking about moving away from the legacy python clients to the sdk [1].\n\n[1] http://lists.openstack.org/pipermail/openstack-discuss/2019-May/005810.html","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":21692,"name":"licanwei","email":"li.canwei2@zte.com.cn","username":"licanwei"},"change_message_id":"5a00d009faeba728b606bfd83b92ce48c5d81089","unresolved":false,"context_lines":[{"line_number":31,"context_line":"Proposed change"},{"line_number":32,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"In this spec we just add Placement helper to Watcher."},{"line_number":35,"context_line":"We plan to improve the data model and strategies in"},{"line_number":36,"context_line":"the future specs."},{"line_number":37,"context_line":"The functions in the helper as below:"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_3f3e71da","line":34,"in_reply_to":"dfbec78f_a62b44e9","updated":"2019-05-08 09:31:14.000000000","message":"Yes, I\u0027ll do similar like Nova.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"fcd7991909ad0370bbd61d03413796c75e5ee9c6","unresolved":false,"context_lines":[{"line_number":36,"context_line":"the future specs."},{"line_number":37,"context_line":"The functions in the helper as below:"},{"line_number":38,"context_line":""},{"line_number":39,"context_line":"* List resource providers"},{"line_number":40,"context_line":"* List resource classes"},{"line_number":41,"context_line":"* List resource provider inventories"},{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_6699cc6c","line":41,"range":{"start_line":39,"start_character":0,"end_line":41,"end_character":36},"updated":"2019-05-07 18:49:44.000000000","message":"For the majority of cases what you\u0027ll care about with this is just the compute node providers and their available inventory of VCPU, MEMORY_MB and DISK_GB. There are also provider aggregates:\n\nhttps://developer.openstack.org/api-ref/placement/#resource-provider-aggregates\n\nWhich should be mirrored from nova\u0027s host aggregates to placement as of Rocky:\n\nhttps://specs.openstack.org/openstack/nova-specs/specs/rocky/implemented/placement-mirror-host-aggregates.html\n\nThose aggregates are just a grouping of compute node providers though, they don\u0027t have metadata like a nova host aggregate. I\u0027m not sure if they\u0027d be useful to watcher at this point.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":21692,"name":"licanwei","email":"li.canwei2@zte.com.cn","username":"licanwei"},"change_message_id":"5a00d009faeba728b606bfd83b92ce48c5d81089","unresolved":false,"context_lines":[{"line_number":36,"context_line":"the future specs."},{"line_number":37,"context_line":"The functions in the helper as below:"},{"line_number":38,"context_line":""},{"line_number":39,"context_line":"* List resource providers"},{"line_number":40,"context_line":"* List resource classes"},{"line_number":41,"context_line":"* List resource provider inventories"},{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_df0dd53e","line":41,"range":{"start_line":39,"start_character":0,"end_line":41,"end_character":36},"in_reply_to":"dfbec78f_6699cc6c","updated":"2019-05-08 09:31:14.000000000","message":"Watcher uses nova host aggregate, maybe need to consider provider aggregate if they are similar, but I need to study how to use them.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"e5fa1bb366e01c493ce93425589016f666335795","unresolved":false,"context_lines":[{"line_number":36,"context_line":"the future specs."},{"line_number":37,"context_line":"The functions in the helper as below:"},{"line_number":38,"context_line":""},{"line_number":39,"context_line":"* List resource providers"},{"line_number":40,"context_line":"* List resource classes"},{"line_number":41,"context_line":"* List resource provider inventories"},{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_f992ba56","line":41,"range":{"start_line":39,"start_character":0,"end_line":41,"end_character":36},"in_reply_to":"dfbec78f_df0dd53e","updated":"2019-05-15 18:43:56.000000000","message":"A resource provider aggregate in placement is just a grouping of resource providers, which may or may not be nova compute nodes. The compute node providers uuid should map back to a hypervisor (compute node) in nova\u0027s hypervisors API. I\u0027m not sure how helpful that will be for you though since like I said you can\u0027t know if the provider is a compute node or something else, like a provider that represents shared storage:\n\nhttps://review.opendev.org/#/c/650188/","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"fcd7991909ad0370bbd61d03413796c75e5ee9c6","unresolved":false,"context_lines":[{"line_number":39,"context_line":"* List resource providers"},{"line_number":40,"context_line":"* List resource classes"},{"line_number":41,"context_line":"* List resource provider inventories"},{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"},{"line_number":45,"context_line":"* List allocation candidates"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_66b22ce9","line":42,"updated":"2019-05-07 18:49:44.000000000","message":"Traits are just qualitative tags essentially on a provider, which can be used to determine where a resource could be scheduled once its inventory constraints are satisfied. Again, I\u0027m not sure how much watcher cares about that kind of information right now. Here is a recent nova spec on how traits are being used during scheduling:\n\nhttps://specs.openstack.org/openstack/nova-specs/specs/train/approved/request-filter-image-types.html","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":21692,"name":"licanwei","email":"li.canwei2@zte.com.cn","username":"licanwei"},"change_message_id":"5a00d009faeba728b606bfd83b92ce48c5d81089","unresolved":false,"context_lines":[{"line_number":39,"context_line":"* List resource providers"},{"line_number":40,"context_line":"* List resource classes"},{"line_number":41,"context_line":"* List resource provider inventories"},{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"},{"line_number":45,"context_line":"* List allocation candidates"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_020eb231","line":42,"in_reply_to":"dfbec78f_66b22ce9","updated":"2019-05-08 09:31:14.000000000","message":"I\u0027m not sure how to use them, but I think it should be useful.\nThis spec just add placement to Watcher, we need to do more work to use placement in strategies.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"fcd7991909ad0370bbd61d03413796c75e5ee9c6","unresolved":false,"context_lines":[{"line_number":40,"context_line":"* List resource classes"},{"line_number":41,"context_line":"* List resource provider inventories"},{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"},{"line_number":45,"context_line":"* List allocation candidates"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_e6c5bc3d","line":44,"range":{"start_line":43,"start_character":0,"end_line":44,"end_character":31},"updated":"2019-05-07 18:49:44.000000000","message":"These are all about determining usage of resource provider inventory for a given provider or a specific consuming project. Maybe useful for determining where you could migrate a server.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":21692,"name":"licanwei","email":"li.canwei2@zte.com.cn","username":"licanwei"},"change_message_id":"5a00d009faeba728b606bfd83b92ce48c5d81089","unresolved":false,"context_lines":[{"line_number":40,"context_line":"* List resource classes"},{"line_number":41,"context_line":"* List resource provider inventories"},{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"},{"line_number":45,"context_line":"* List allocation candidates"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_6228ce85","line":44,"range":{"start_line":43,"start_character":0,"end_line":44,"end_character":31},"in_reply_to":"dfbec78f_e6c5bc3d","updated":"2019-05-08 09:31:14.000000000","message":"Thanks","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"fcd7991909ad0370bbd61d03413796c75e5ee9c6","unresolved":false,"context_lines":[{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"},{"line_number":45,"context_line":"* List allocation candidates"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"Alternatives"},{"line_number":48,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_26d194fc","line":45,"updated":"2019-05-07 18:49:44.000000000","message":"This is maybe the most important for watcher to be able to make queries to placement, like nova-scheduler, to determine where a server could be migrated and then migrate attempt to migrate it there via the compute API, but kind of like a dry run before asking nova to migrate the server.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"50e25630e931d0b95ddcb1153e694bf11481cb8e","unresolved":false,"context_lines":[{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"},{"line_number":45,"context_line":"* List allocation candidates"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"Alternatives"},{"line_number":48,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_b971a246","line":45,"in_reply_to":"dfbec78f_19830e06","updated":"2019-05-15 18:45:09.000000000","message":"Some more details on that here:\n\nhttps://docs.openstack.org/nova/latest/admin/configuration/schedulers.html#allocation-ratios","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":21692,"name":"licanwei","email":"li.canwei2@zte.com.cn","username":"licanwei"},"change_message_id":"5a00d009faeba728b606bfd83b92ce48c5d81089","unresolved":false,"context_lines":[{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"},{"line_number":45,"context_line":"* List allocation candidates"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"Alternatives"},{"line_number":48,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_c2309aeb","line":45,"in_reply_to":"dfbec78f_26d194fc","updated":"2019-05-08 09:31:14.000000000","message":"Yes, we wish Watcher can get more suitable node for VM migration through placement.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"e5fa1bb366e01c493ce93425589016f666335795","unresolved":false,"context_lines":[{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"},{"line_number":45,"context_line":"* List allocation candidates"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"Alternatives"},{"line_number":48,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_19830e06","line":45,"in_reply_to":"dfbec78f_6095af32","updated":"2019-05-15 18:43:56.000000000","message":"Yes I think so. The resource provider inventory has an allocation ratio per resource class:\n\nhttps://developer.openstack.org/api-ref/placement/?expanded\u003dlist-resource-provider-inventories-detail#list-resource-provider-inventories\n\nSo for example the cpu_allocation_ratio for a nova compute node is reported to the matching VCPU resource class inventory for the provider that represents that node.\n\nThis is different from the os-hypervisors API which does not take allocation ratio into account when reporting total capacity:\n\nId8532d5a9480e4f71a9390e3b98886c85e27d003","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":29911,"name":"Dantali0n","email":"info@dantalion.nl","username":"Dantali0n"},"change_message_id":"68933ad2c42ba788f912a08f05a6ecd6e6bd470a","unresolved":false,"context_lines":[{"line_number":42,"context_line":"* List resource provider traits"},{"line_number":43,"context_line":"* List resource provider allocations"},{"line_number":44,"context_line":"* List resource provider usages"},{"line_number":45,"context_line":"* List allocation candidates"},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"Alternatives"},{"line_number":48,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_6095af32","line":45,"in_reply_to":"dfbec78f_c2309aeb","updated":"2019-05-15 15:47:24.000000000","message":"Could the placement API be used to determine a valid location to move an instance to that respects the desired over commit ratio?\n\nIf so it would greatly benefit many existing strategies such as workload_balance which currently only considers hosts with an over commit of less then 1 as a valid migration candidate.","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"fcd7991909ad0370bbd61d03413796c75e5ee9c6","unresolved":false,"context_lines":[{"line_number":127,"context_line":"References"},{"line_number":128,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":129,"context_line":""},{"line_number":130,"context_line":"None"},{"line_number":131,"context_line":""},{"line_number":132,"context_line":""},{"line_number":133,"context_line":"History"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_463a8834","line":130,"updated":"2019-05-07 18:49:44.000000000","message":"Would probably be useful to refer to at least the placement API reference:\n\nhttps://developer.openstack.org/api-ref/placement/\n\nand/or docs:\n\nhttps://docs.openstack.org/placement/latest/","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":21692,"name":"licanwei","email":"li.canwei2@zte.com.cn","username":"licanwei"},"change_message_id":"5a00d009faeba728b606bfd83b92ce48c5d81089","unresolved":false,"context_lines":[{"line_number":127,"context_line":"References"},{"line_number":128,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":129,"context_line":""},{"line_number":130,"context_line":"None"},{"line_number":131,"context_line":""},{"line_number":132,"context_line":""},{"line_number":133,"context_line":"History"}],"source_content_type":"text/x-rst","patch_set":2,"id":"dfbec78f_22573605","line":130,"in_reply_to":"dfbec78f_463a8834","updated":"2019-05-08 09:31:14.000000000","message":"thanks, will update","commit_id":"3eef72fd59b54e82ebf807be41e4d1200c41bd0d"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"d7b2df6bc31f440445de5a0971ccfc1bf25c636d","unresolved":false,"context_lines":[{"line_number":84,"context_line":"Other deployer impact"},{"line_number":85,"context_line":"---------------------"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"None"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"Developer impact"},{"line_number":90,"context_line":"----------------"}],"source_content_type":"text/x-rst","patch_set":4,"id":"bfb3d3c7_ec0ee78b","line":87,"updated":"2019-05-22 19:23:26.000000000","message":"I assume there will be a new [placement_client] section added to watcher.conf for configuring placement, right? And that section would have api_version, endpoint_type and region_name options.","commit_id":"87db97710dfeb25303cebfa667bc841d27a7f3a3"},{"author":{"_account_id":29911,"name":"Dantali0n","email":"info@dantalion.nl","username":"Dantali0n"},"change_message_id":"ec64dc4826b2248c45774e1e2e6491bd214abf7a","unresolved":false,"context_lines":[{"line_number":84,"context_line":"Other deployer impact"},{"line_number":85,"context_line":"---------------------"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"None"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"Developer impact"},{"line_number":90,"context_line":"----------------"}],"source_content_type":"text/x-rst","patch_set":4,"id":"bfb3d3c7_f67309b0","line":87,"in_reply_to":"bfb3d3c7_ec0ee78b","updated":"2019-05-23 06:26:29.000000000","message":"isen\u0027t placement just part of nova client? And can\u0027t the same configuration for nova be used for placement?","commit_id":"87db97710dfeb25303cebfa667bc841d27a7f3a3"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"3a9b401135ce2e48d36fd70cfab77a0e4b4565d7","unresolved":false,"context_lines":[{"line_number":84,"context_line":"Other deployer impact"},{"line_number":85,"context_line":"---------------------"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"None"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"Developer impact"},{"line_number":90,"context_line":"----------------"}],"source_content_type":"text/x-rst","patch_set":4,"id":"bfb3d3c7_6df97fc6","line":87,"in_reply_to":"bfb3d3c7_f67309b0","updated":"2019-05-23 16:28:04.000000000","message":"As mentioned in the review comments placement is a separate service from nova. It started in the nova git repo but was always a separate service and was extracted from the nova repo in stein and the placement-in-nova code was dropped from the nova git repo in train. There is no python-placementclient like there is a python-novaclient. Services like nova, neutron cyborg and blazar that are using placement are writing their own internal client code using keystoneauth1 and/or openstacksdk.","commit_id":"87db97710dfeb25303cebfa667bc841d27a7f3a3"}]}