)]}'
{"doc/source/placement.rst":[{"author":{"_account_id":7,"name":"Jay Pipes","email":"jaypipes@gmail.com","username":"jaypipes"},"change_message_id":"245ee3206315854720b7202ec8fbab0bde5e8984","unresolved":false,"context_lines":[{"line_number":23,"context_line":"inventories and usages, along with different classes of resources. For example,"},{"line_number":24,"context_line":"a resource provider can be a compute node, a shared storage pool, or an IP"},{"line_number":25,"context_line":"allocation pool. The placement service tracks the inventory and usage of each"},{"line_number":26,"context_line":"provider. For example, an instance created on a compute node will be a consumer"},{"line_number":27,"context_line":"of resources such as RAM and CPU from a compute node resource provider, DISK"},{"line_number":28,"context_line":"from an external shared storage pool resource provider and IP addresses from"},{"line_number":29,"context_line":"an external IP pool resource provider."}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a629dbe_99c72a63","line":26,"updated":"2016-11-11 20:42:16.000000000","message":"s/will/might/\n\nThe way the above reads is that all instances will consume RAM and CPU from a compute node resource provider, DISK from a shared storage pool resource provider and IPv4 addresses from an IP allocation pool provider. That might be a common case especially in the future but won\u0027t always be true.","commit_id":"d4ce7891f63b9d7cd6f32e6a8d13af525bdcd03e"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"2db3cd1cc69649330c5ffdddc17a9b3afaf9f138","unresolved":false,"context_lines":[{"line_number":23,"context_line":"inventories and usages, along with different classes of resources. For example,"},{"line_number":24,"context_line":"a resource provider can be a compute node, a shared storage pool, or an IP"},{"line_number":25,"context_line":"allocation pool. The placement service tracks the inventory and usage of each"},{"line_number":26,"context_line":"provider. For example, an instance created on a compute node will be a consumer"},{"line_number":27,"context_line":"of resources such as RAM and CPU from a compute node resource provider, DISK"},{"line_number":28,"context_line":"from an external shared storage pool resource provider and IP addresses from"},{"line_number":29,"context_line":"an external IP pool resource provider."}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a629dbe_3d8c49d9","line":26,"in_reply_to":"9a629dbe_99c72a63","updated":"2016-11-11 21:14:16.000000000","message":"I had some wording in here at some point like \u0027potentially DISK from an external shared storage pool resource provider and potentially IP addresses from an external IP pool resource provider.\u0027 I could switch back to that. Changing \u0027will\u0027 to \u0027might\u0027 up there could also be confusing if you\u0027re reading it as the instance might consume CPU/RAM from the compute node but maybe something else, and I don\u0027t think we have something else for that. Either way I\u0027m fine updating this to be a bit more clear.","commit_id":"d4ce7891f63b9d7cd6f32e6a8d13af525bdcd03e"},{"author":{"_account_id":7,"name":"Jay Pipes","email":"jaypipes@gmail.com","username":"jaypipes"},"change_message_id":"245ee3206315854720b7202ec8fbab0bde5e8984","unresolved":false,"context_lines":[{"line_number":57,"context_line":"The placement-api service must be deployed at some point after you have"},{"line_number":58,"context_line":"upgraded to the 14.0.0 Newton release but before you can upgrade to the 15.0.0"},{"line_number":59,"context_line":"Ocata release. This is so that the resource tracker in the nova-compute service"},{"line_number":60,"context_line":"can populate resource provider (compute node) inventory and allocation"},{"line_number":61,"context_line":"information which will be used by the nova-scheduler service in Ocata."},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"Steps"}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a629dbe_99ac8a8d","line":60,"updated":"2016-11-11 20:42:16.000000000","message":"s/resource provider (compute node)//\n\nI think it\u0027s easier to read by just saying \"populate inventory and allocation information\"...","commit_id":"d4ce7891f63b9d7cd6f32e6a8d13af525bdcd03e"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"2db3cd1cc69649330c5ffdddc17a9b3afaf9f138","unresolved":false,"context_lines":[{"line_number":57,"context_line":"The placement-api service must be deployed at some point after you have"},{"line_number":58,"context_line":"upgraded to the 14.0.0 Newton release but before you can upgrade to the 15.0.0"},{"line_number":59,"context_line":"Ocata release. This is so that the resource tracker in the nova-compute service"},{"line_number":60,"context_line":"can populate resource provider (compute node) inventory and allocation"},{"line_number":61,"context_line":"information which will be used by the nova-scheduler service in Ocata."},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"Steps"}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a629dbe_5dae5d2e","line":60,"in_reply_to":"9a629dbe_99ac8a8d","updated":"2016-11-11 21:14:16.000000000","message":"I qualified with compute node here because that\u0027s what the resource tracker is checking specific when the nova RT is the client. I didn\u0027t want it to sound like the nova-compute RT is going to be tracking external shared storage or IP pool inventories and allocations.","commit_id":"d4ce7891f63b9d7cd6f32e6a8d13af525bdcd03e"},{"author":{"_account_id":11564,"name":"Chris Dent","email":"cdent@anticdent.org","username":"chdent"},"change_message_id":"fa1f3d899836ff30dfb11957386f66e69d4a39f6","unresolved":false,"context_lines":[{"line_number":24,"context_line":"a resource provider can be a compute node, a shared storage pool, or an IP"},{"line_number":25,"context_line":"allocation pool. The placement service tracks the inventory and usage of each"},{"line_number":26,"context_line":"provider. For example, an instance created on a compute node may be a consumer"},{"line_number":27,"context_line":"of resources such as RAM and CPU from a compute node resource provider, DISK"},{"line_number":28,"context_line":"from an external shared storage pool resource provider and IP addresses from"},{"line_number":29,"context_line":"an external IP pool resource provider."},{"line_number":30,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a629dbe_5c48f287","line":27,"updated":"2016-11-14 14:13:39.000000000","message":"not that it really matters but the use of all caps DISK here has nice parallelism with RAM and CPU but isn\u0027t quite right.","commit_id":"310d66c9a68adbdb239a8a88713499fc039d39ba"},{"author":{"_account_id":11564,"name":"Chris Dent","email":"cdent@anticdent.org","username":"chdent"},"change_message_id":"fa1f3d899836ff30dfb11957386f66e69d4a39f6","unresolved":false,"context_lines":[{"line_number":30,"context_line":""},{"line_number":31,"context_line":"The types of resources consumed are tracked as **classes**. The service"},{"line_number":32,"context_line":"provides a set of standard resource classes but provides the ability to define"},{"line_number":33,"context_line":"custom resource classes as needed."},{"line_number":34,"context_line":""},{"line_number":35,"context_line":"References"},{"line_number":36,"context_line":"~~~~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a629dbe_bc2e6ea3","line":33,"updated":"2016-11-14 14:13:39.000000000","message":"make any sense to say \"per deployment\"\n\nPerhaps in a later tuneup.","commit_id":"310d66c9a68adbdb239a8a88713499fc039d39ba"},{"author":{"_account_id":11564,"name":"Chris Dent","email":"cdent@anticdent.org","username":"chdent"},"change_message_id":"fa1f3d899836ff30dfb11957386f66e69d4a39f6","unresolved":false,"context_lines":[{"line_number":96,"context_line":"resource tracker in the nova-compute node will use to look up the endpoint."},{"line_number":97,"context_line":""},{"line_number":98,"context_line":"Devstack sets up the port used in the endpoint URL as 8778 which is"},{"line_number":99,"context_line":"intentionally different from the 8774 port used with the compute endpoint."},{"line_number":100,"context_line":""},{"line_number":101,"context_line":"**4. Configure and restart nova-compute services**"},{"line_number":102,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a629dbe_97cd0560","line":99,"updated":"2016-11-14 14:13:39.000000000","message":"This isn\u0027t actually what devstack does, despite bits of the code making it appear that way :( . The endpoint that gets registered in devstack is on port 80 with a /placement prefix:\n\n    $PLACEMENT_SERVICE_PROTOCOL://$PLACEMENT_SERVICE_HOST/placement\n\nThat\u0027s because it tries to follow the \"all services under one host and port\" principle that was a popular idea around devstack for a while.\n\nThat doesn\u0027t have to mean anything here, really, just wanted to point it out lest there be confusion.","commit_id":"310d66c9a68adbdb239a8a88713499fc039d39ba"}]}