)]}'
{"specs/backlog/approved/quota-future.rst":[{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"e499f9ba6ddd2f116a2a238744480163052a5503","unresolved":false,"context_lines":[{"line_number":181,"context_line":"additional resources, or ideally part of the claim process."},{"line_number":182,"context_line":""},{"line_number":183,"context_line":"Scope"},{"line_number":184,"context_line":"-----"},{"line_number":185,"context_line":""},{"line_number":186,"context_line":"Lets consider the quota checking process defined above. The scope is relevant"},{"line_number":187,"context_line":"for these steps:"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1a430d35_16929e57","line":184,"updated":"2017-02-06 15:40:20.000000000","message":"This was a bit messed up, got a slightly clearer version in the follow up version.","commit_id":"ecf76d9437f10d470378c9316276cda643397d5c"},{"author":{"_account_id":10343,"name":"Jim Rollenhagen","email":"jim@jimrollenhagen.com","username":"jimrollenhagen"},"change_message_id":"2a433419c2dac47b96719e0d4b60ddece9dfeb0b","unresolved":false,"context_lines":[{"line_number":34,"context_line":""},{"line_number":35,"context_line":"However, the above model only allows a parent project to delegate a fixed"},{"line_number":36,"context_line":"portion of the quota to a child project. It doesn\u0027t allow the over allocation"},{"line_number":37,"context_line":"of quota between sub projects."},{"line_number":38,"context_line":""},{"line_number":39,"context_line":"Overall, we need a unified approach to quotas that can be applied across all"},{"line_number":40,"context_line":"relevant OpenStack projects, as required. That is not to be confused with"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a430d35_c201026c","line":37,"updated":"2017-02-07 15:28:50.000000000","message":"Maybe I\u0027m just out of the loop, but I don\u0027t get this paragraph. Could you expand on it, and maybe talk about why we need over allocation (seems like a thing we wouldn\u0027t want, to me)","commit_id":"c6d6a4c689aa4e83b51b598b24c72b08d1f58b53"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"b799a863536e6166664b52fef4f9c05e6471f985","unresolved":false,"context_lines":[{"line_number":34,"context_line":""},{"line_number":35,"context_line":"However, the above model only allows a parent project to delegate a fixed"},{"line_number":36,"context_line":"portion of the quota to a child project. It doesn\u0027t allow the over allocation"},{"line_number":37,"context_line":"of quota between sub projects."},{"line_number":38,"context_line":""},{"line_number":39,"context_line":"Overall, we need a unified approach to quotas that can be applied across all"},{"line_number":40,"context_line":"relevant OpenStack projects, as required. That is not to be confused with"}],"source_content_type":"text/x-rst","patch_set":3,"id":"ba2be162_8bb5f1d1","line":37,"in_reply_to":"1a430d35_c201026c","updated":"2017-02-28 13:45:07.000000000","message":"I should add more in the use cases on that one. Basically, you want to max the utilisation of your cloud, and most folks don\u0027t use all the quota they are given, its just a protection against one group being overly greedy.","commit_id":"c6d6a4c689aa4e83b51b598b24c72b08d1f58b53"},{"author":{"_account_id":10343,"name":"Jim Rollenhagen","email":"jim@jimrollenhagen.com","username":"jimrollenhagen"},"change_message_id":"2a433419c2dac47b96719e0d4b60ddece9dfeb0b","unresolved":false,"context_lines":[{"line_number":124,"context_line":"https://specs.openstack.org/openstack/nova-specs/specs/backlog/approved/flavor-class.html"},{"line_number":125,"context_line":""},{"line_number":126,"context_line":"With the new ironic custom resource classes in the Placement API, we are"},{"line_number":127,"context_line":"better served by looking at moving to restricting "},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"Quota templates"},{"line_number":130,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a430d35_b67b578b","line":127,"updated":"2017-02-07 15:28:50.000000000","message":"This sentence isn\u0027t finished, but I\u0027m hoping you\u0027re going to say something about restricting tenants to N machines of each flavor or whatever :)","commit_id":"c6d6a4c689aa4e83b51b598b24c72b08d1f58b53"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"b799a863536e6166664b52fef4f9c05e6471f985","unresolved":false,"context_lines":[{"line_number":124,"context_line":"https://specs.openstack.org/openstack/nova-specs/specs/backlog/approved/flavor-class.html"},{"line_number":125,"context_line":""},{"line_number":126,"context_line":"With the new ironic custom resource classes in the Placement API, we are"},{"line_number":127,"context_line":"better served by looking at moving to restricting "},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"Quota templates"},{"line_number":130,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"da36d5c6_4bb0fe43","line":127,"in_reply_to":"1a430d35_b67b578b","updated":"2017-02-28 13:45:07.000000000","message":"Opps, yeah, that.","commit_id":"c6d6a4c689aa4e83b51b598b24c72b08d1f58b53"},{"author":{"_account_id":10343,"name":"Jim Rollenhagen","email":"jim@jimrollenhagen.com","username":"jimrollenhagen"},"change_message_id":"f33683a3b97872f669fe95d7be1571e5dd86a8e7","unresolved":false,"context_lines":[{"line_number":258,"context_line":"To achieve this, lets consider the concept of a Resource Class Limit. It is"},{"line_number":259,"context_line":"the maximum quantity allowed for the given scope."},{"line_number":260,"context_line":""},{"line_number":261,"context_line":"TODO - what about qualitative things, should we limit the counts?"},{"line_number":262,"context_line":""},{"line_number":263,"context_line":"Quota Class"},{"line_number":264,"context_line":"-----------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a430d35_85653f1f","line":261,"updated":"2017-02-07 16:51:10.000000000","message":"I guess it depends on the thing... e.g. \"can do uefi\" we should not, \"has direct connection to internet backbone\" maybe? I\u0027d almost leave this out for now unless you think the outcome might be different with it here.","commit_id":"c6d6a4c689aa4e83b51b598b24c72b08d1f58b53"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"b799a863536e6166664b52fef4f9c05e6471f985","unresolved":false,"context_lines":[{"line_number":258,"context_line":"To achieve this, lets consider the concept of a Resource Class Limit. It is"},{"line_number":259,"context_line":"the maximum quantity allowed for the given scope."},{"line_number":260,"context_line":""},{"line_number":261,"context_line":"TODO - what about qualitative things, should we limit the counts?"},{"line_number":262,"context_line":""},{"line_number":263,"context_line":"Quota Class"},{"line_number":264,"context_line":"-----------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"ba2be162_eb13cd0b","line":261,"in_reply_to":"1a430d35_85653f1f","updated":"2017-02-28 13:45:07.000000000","message":"yeah, probably too premature here.","commit_id":"c6d6a4c689aa4e83b51b598b24c72b08d1f58b53"}]}