)]}'
{"specs/rocky/approved/granular-resource-requests.rst":[{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"1310ffbaf92550e4d0753f37568229de97586012","unresolved":false,"context_lines":[{"line_number":251,"context_line":"* The semantic for the (single) un-numbered grouping is unchanged.  That is, it"},{"line_number":252,"context_line":"  may still return results from different RPs in the same tree (or, when"},{"line_number":253,"context_line":"  \"shared\" is fully implemented, the same aggregate)."},{"line_number":254,"context_line":"* However, a numbered group will always return results from the *same* RP."},{"line_number":255,"context_line":"  This is to satisfy `Requirement 2`_."},{"line_number":256,"context_line":"* With ``group_policy\u003dnone``, separate groups (numbered or un-numbered) may"},{"line_number":257,"context_line":"  return results from different RPs *or* the same RP (assuming isolation is not"},{"line_number":258,"context_line":"  otherwise forced e.g. via traits or inventory/usage constraints)."}],"source_content_type":"text/x-rst","patch_set":1,"id":"9f6a8fd7_d98061c2","line":255,"range":{"start_line":254,"start_character":2,"end_line":255,"end_character":38},"updated":"2018-04-20 14:32:48.000000000","message":"Is this still true with group_policy\u003dnone?","commit_id":"cfbfa180824c94d44d7389445c3af86de9b28d37"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"3b15ecb78feb1741cce6250287d761b3c911513d","unresolved":false,"context_lines":[{"line_number":251,"context_line":"* The semantic for the (single) un-numbered grouping is unchanged.  That is, it"},{"line_number":252,"context_line":"  may still return results from different RPs in the same tree (or, when"},{"line_number":253,"context_line":"  \"shared\" is fully implemented, the same aggregate)."},{"line_number":254,"context_line":"* However, a numbered group will always return results from the *same* RP."},{"line_number":255,"context_line":"  This is to satisfy `Requirement 2`_."},{"line_number":256,"context_line":"* With ``group_policy\u003dnone``, separate groups (numbered or un-numbered) may"},{"line_number":257,"context_line":"  return results from different RPs *or* the same RP (assuming isolation is not"},{"line_number":258,"context_line":"  otherwise forced e.g. via traits or inventory/usage constraints)."}],"source_content_type":"text/x-rst","patch_set":1,"id":"9f6a8fd7_93572259","line":255,"range":{"start_line":254,"start_character":2,"end_line":255,"end_character":38},"in_reply_to":"9f6a8fd7_8d2943e7","updated":"2018-04-20 18:17:20.000000000","message":"OK, I thought so, but without the \"within\" confirmation / wording in this sentence I wanted to make sure.","commit_id":"cfbfa180824c94d44d7389445c3af86de9b28d37"},{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"bdeee0264256ca5a3b1841912e8090af97f9b274","unresolved":false,"context_lines":[{"line_number":251,"context_line":"* The semantic for the (single) un-numbered grouping is unchanged.  That is, it"},{"line_number":252,"context_line":"  may still return results from different RPs in the same tree (or, when"},{"line_number":253,"context_line":"  \"shared\" is fully implemented, the same aggregate)."},{"line_number":254,"context_line":"* However, a numbered group will always return results from the *same* RP."},{"line_number":255,"context_line":"  This is to satisfy `Requirement 2`_."},{"line_number":256,"context_line":"* With ``group_policy\u003dnone``, separate groups (numbered or un-numbered) may"},{"line_number":257,"context_line":"  return results from different RPs *or* the same RP (assuming isolation is not"},{"line_number":258,"context_line":"  otherwise forced e.g. via traits or inventory/usage constraints)."}],"source_content_type":"text/x-rst","patch_set":1,"id":"9f6a8fd7_8d2943e7","line":255,"range":{"start_line":254,"start_character":2,"end_line":255,"end_character":38},"in_reply_to":"9f6a8fd7_d98061c2","updated":"2018-04-20 16:40:07.000000000","message":"Yes.  Resources *within* a numbered group are *always* from the same provider.  The group_policy affects how numbered groups interact with each other.","commit_id":"cfbfa180824c94d44d7389445c3af86de9b28d37"},{"author":{"_account_id":16929,"name":"Giridhar Jayavelu","email":"gjayavelu@vmware.com","username":"gjayavelu"},"change_message_id":"e1a132be80ca5ef0e8377eadb3489ce86ef03de1","unresolved":false,"context_lines":[{"line_number":344,"context_line":"    resources2:NET_EGRESS_BYTES_SEC:20000"},{"line_number":345,"context_line":"    trait2:CUSTOM_PHYSNET_NET2\u003drequired"},{"line_number":346,"context_line":"    trait2:HW_NIC_ACCEL_SSL\u003drequired"},{"line_number":347,"context_line":"    group_policy\u003disolate"},{"line_number":348,"context_line":""},{"line_number":349,"context_line":"Syntax In the Placement API"},{"line_number":350,"context_line":"~~~~~~~~~~~~~~~~~~~~~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":1,"id":"9f6a8fd7_c5a2cb4b","line":347,"updated":"2018-04-19 22:59:52.000000000","message":"this policy seems to apply globally for all grouped resources. \nIf there is a need for one resource class to follow affinity and another resource class to follow anti-affinity, it may not be possible to express in same flavor. Is my understanding correct? Say, I want numa affinity but different PFs.\n\nwould it make sense to use group policy per resource class?\nex: group_policy:RESOURCE_CLASS_A\u003disolate\ngroup_policy:RESOURCE_CLASS_B\u003dnone","commit_id":"cfbfa180824c94d44d7389445c3af86de9b28d37"},{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"bdeee0264256ca5a3b1841912e8090af97f9b274","unresolved":false,"context_lines":[{"line_number":344,"context_line":"    resources2:NET_EGRESS_BYTES_SEC:20000"},{"line_number":345,"context_line":"    trait2:CUSTOM_PHYSNET_NET2\u003drequired"},{"line_number":346,"context_line":"    trait2:HW_NIC_ACCEL_SSL\u003drequired"},{"line_number":347,"context_line":"    group_policy\u003disolate"},{"line_number":348,"context_line":""},{"line_number":349,"context_line":"Syntax In the Placement API"},{"line_number":350,"context_line":"~~~~~~~~~~~~~~~~~~~~~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":1,"id":"9f6a8fd7_6d1cefc0","line":347,"in_reply_to":"9f6a8fd7_7ca7c3a8","updated":"2018-04-20 16:40:07.000000000","message":"I threw a couple of suggestions in one of the ML responses.  Peanuts welcome there.","commit_id":"cfbfa180824c94d44d7389445c3af86de9b28d37"},{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"f0411c626a076e54af3e7146b081946d9e320dcf","unresolved":false,"context_lines":[{"line_number":344,"context_line":"    resources2:NET_EGRESS_BYTES_SEC:20000"},{"line_number":345,"context_line":"    trait2:CUSTOM_PHYSNET_NET2\u003drequired"},{"line_number":346,"context_line":"    trait2:HW_NIC_ACCEL_SSL\u003drequired"},{"line_number":347,"context_line":"    group_policy\u003disolate"},{"line_number":348,"context_line":""},{"line_number":349,"context_line":"Syntax In the Placement API"},{"line_number":350,"context_line":"~~~~~~~~~~~~~~~~~~~~~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":1,"id":"9f6a8fd7_c937384d","line":347,"in_reply_to":"9f6a8fd7_c5a2cb4b","updated":"2018-04-20 11:42:53.000000000","message":"\u003e this policy seems to apply globally for all grouped resources.\n\nYou are correct.\n\n \u003e If there is a need for one resource class to follow affinity and\n \u003e another resource class to follow anti-affinity, it may not be\n \u003e possible to express in same flavor. Is my understanding correct?\n \u003e Say, I want numa affinity but different PFs.\n \u003e \n \u003e would it make sense to use group policy per resource class?\n \u003e ex: group_policy:RESOURCE_CLASS_A\u003disolate\n \u003e group_policy:RESOURCE_CLASS_B\u003dnone\n\nThis and cases like it have been the source of the extensive and gruelling debate that led us to this compromise.  If you\u0027ve got the time and patience, grab a cup of coffee and read through the ML thread and the other spec delta proposals linked...","commit_id":"cfbfa180824c94d44d7389445c3af86de9b28d37"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"1310ffbaf92550e4d0753f37568229de97586012","unresolved":false,"context_lines":[{"line_number":344,"context_line":"    resources2:NET_EGRESS_BYTES_SEC:20000"},{"line_number":345,"context_line":"    trait2:CUSTOM_PHYSNET_NET2\u003drequired"},{"line_number":346,"context_line":"    trait2:HW_NIC_ACCEL_SSL\u003drequired"},{"line_number":347,"context_line":"    group_policy\u003disolate"},{"line_number":348,"context_line":""},{"line_number":349,"context_line":"Syntax In the Placement API"},{"line_number":350,"context_line":"~~~~~~~~~~~~~~~~~~~~~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":1,"id":"9f6a8fd7_d9b5c17f","line":347,"in_reply_to":"9f6a8fd7_c937384d","updated":"2018-04-20 14:32:48.000000000","message":"(From the peanut gallery)\n\nI think if we get to the point that we need to define group policy per group, then we need to seriously consider changing the API to not be a GET with a highly-complex query parameter structure and instead change it to a POST with a request body to describe the detailed request.","commit_id":"cfbfa180824c94d44d7389445c3af86de9b28d37"},{"author":{"_account_id":7,"name":"Jay Pipes","email":"jaypipes@gmail.com","username":"jaypipes"},"change_message_id":"87357c51ac0877403cf7991baf7bdfcd774d4dd5","unresolved":false,"context_lines":[{"line_number":344,"context_line":"    resources2:NET_EGRESS_BYTES_SEC:20000"},{"line_number":345,"context_line":"    trait2:CUSTOM_PHYSNET_NET2\u003drequired"},{"line_number":346,"context_line":"    trait2:HW_NIC_ACCEL_SSL\u003drequired"},{"line_number":347,"context_line":"    group_policy\u003disolate"},{"line_number":348,"context_line":""},{"line_number":349,"context_line":"Syntax In the Placement API"},{"line_number":350,"context_line":"~~~~~~~~~~~~~~~~~~~~~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":1,"id":"9f6a8fd7_7ca7c3a8","line":347,"in_reply_to":"9f6a8fd7_d9b5c17f","updated":"2018-04-20 14:45:09.000000000","message":"Couldn\u0027t agree more, Matt.","commit_id":"cfbfa180824c94d44d7389445c3af86de9b28d37"},{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"f0411c626a076e54af3e7146b081946d9e320dcf","unresolved":false,"context_lines":[{"line_number":555,"context_line":".. _`Nested Resource Providers`: https://specs.openstack.org/openstack/nova-specs/specs/queens/approved/nested-resource-providers.html"},{"line_number":556,"context_line":".. _`Placement API`: https://developer.openstack.org/api-ref/placement/#list-allocation-candidates"},{"line_number":557,"context_line":".. _`Placement Devref`: https://docs.openstack.org/nova/latest/user/placement.html"},{"line_number":558,"context_line":".. _`separate_providers proposal`: https://review.openstack.org/#/c/561717/"},{"line_number":559,"context_line":".. _`can_split proposal`: https://review.openstack.org/#/c/560974/"},{"line_number":560,"context_line":".. _`mailing list thread`: http://lists.openstack.org/pipermail/openstack-dev/2018-April/129477.html"},{"line_number":561,"context_line":""},{"line_number":562,"context_line":"History"},{"line_number":563,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":1,"id":"9f6a8fd7_893dc02c","line":560,"range":{"start_line":558,"start_character":0,"end_line":560,"end_character":100},"updated":"2018-04-20 11:42:53.000000000","message":"...here.\n\nBut TL;DR, we decided on this \"global\" qparam toggle as an initial pass to get us moving forward, recognizing that it doesn\u0027t cover 100% of the use cases we\u0027ll eventually need.  But it covers a lot of them, and it\u0027s extensible so we can start to fill the gaps moving forward.","commit_id":"cfbfa180824c94d44d7389445c3af86de9b28d37"}]}