)]}'
{"specs/rocky/mark-backend-pool-sold-out.rst":[{"author":{"_account_id":11904,"name":"Sean McGinnis","email":"sean.mcginnis@gmail.com","username":"SeanM"},"change_message_id":"2744a7220f72a5e4f9d7c134390c9a29cb5df9cd","unresolved":false,"context_lines":[{"line_number":17,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":18,"context_line":""},{"line_number":19,"context_line":"This is more related to administrative operation, when most of one specific pool\u0027s"},{"line_number":20,"context_line":"capacity has been consumed, administrator need the functionality to mark the backend"},{"line_number":21,"context_line":"pool sold out, that is to say, when the pool is marked sold out, it will become"},{"line_number":22,"context_line":"unavailable for **new users or customers**, but the pool\u0027s existing users still can"},{"line_number":23,"context_line":"perform any capacity consumed operations on their resources (volume, snapshot)."}],"source_content_type":"text/x-rst","patch_set":1,"id":"bf659307_d5703ddf","line":20,"updated":"2018-03-28 14:46:36.000000000","message":"If you do an update, please reformat lines to stay within 80 characters in case we want to enable doc8 checking on these.","commit_id":"1ecceffb6f5156c15706e3cb06552eaf700b23af"},{"author":{"_account_id":2243,"name":"John Griffith","email":"john.griffith8@gmail.com","username":"john-griffith"},"change_message_id":"23d7f986c78967290e5f09dd7aa1b882db4f047e","unresolved":false,"context_lines":[{"line_number":68,"context_line":"7. retype_volume"},{"line_number":69,"context_line":"8. manage_existing_snapshot"},{"line_number":70,"context_line":""},{"line_number":71,"context_line":"Alternatives"},{"line_number":72,"context_line":"------------"},{"line_number":73,"context_line":""},{"line_number":74,"context_line":"One of the alternative is to fully support this in cinder and in-tree plugins,"}],"source_content_type":"text/x-rst","patch_set":1,"id":"bf659307_ae7e9a68","line":71,"range":{"start_line":71,"start_character":0,"end_line":71,"end_character":12},"updated":"2018-03-28 15:42:24.000000000","message":"Another alternative to consider is that we could provide thresholds to drivers and automate this entirely.\n\nWhat I mean is introduce things like \"capacity_threshold\" variables to the config file, and when a backend reaches said threshold it\u0027s no longer valid to schedule to it.\n\nI have some issues with things like snapshots in your proposal, mostly because IMO there is no global answer here, backends that use qcow, or deduplication of blocks etc do not and should not care about things like snapshots in terms of capacity.\n\nThis spec from what I\u0027m gathering sort of brute forces every backend to use hard coded thick provisioning semantics regardless of their implementation.  Maybe I\u0027m incorrect on that, but it sort of seems that way.\n\nYou state yourself in the spec that this is a \"complicated feature\" I think it\u0027s worth considering a more flexible and simpler implementation to achieve the end result.","commit_id":"1ecceffb6f5156c15706e3cb06552eaf700b23af"}],"specs/rocky/support-filter-backend-on-operation-type.rst":[{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"685ed8383778870a8ac23b148240334eb4b75566","unresolved":false,"context_lines":[{"line_number":15,"context_line":"Problem description"},{"line_number":16,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":17,"context_line":""},{"line_number":18,"context_line":"Now our filter and weigher plugins\u0027 algorithm is all designed based on"},{"line_number":19,"context_line":"two factors, backend information and user input. But there could be"},{"line_number":20,"context_line":"another factor that may also affect the scheduler result, operation"},{"line_number":21,"context_line":"type. For instance, if administrators want to disable some of the"}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_f2725e7c","line":18,"range":{"start_line":18,"start_character":46,"end_line":18,"end_character":48},"updated":"2018-04-09 11:28:32.000000000","message":"nit: are","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"c693a130c7b9d8e1f8a1443d4f27f16b5d40eef6","unresolved":false,"context_lines":[{"line_number":15,"context_line":"Problem description"},{"line_number":16,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":17,"context_line":""},{"line_number":18,"context_line":"Now our filter and weigher plugins\u0027 algorithm is all designed based on"},{"line_number":19,"context_line":"two factors, backend information and user input. But there could be"},{"line_number":20,"context_line":"another factor that may also affect the scheduler result, operation"},{"line_number":21,"context_line":"type. For instance, if administrators want to disable some of the"}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_5511b403","line":18,"range":{"start_line":18,"start_character":46,"end_line":18,"end_character":48},"in_reply_to":"bf659307_f2725e7c","updated":"2018-04-09 12:06:15.000000000","message":"Done","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"685ed8383778870a8ac23b148240334eb4b75566","unresolved":false,"context_lines":[{"line_number":26,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":27,"context_line":""},{"line_number":28,"context_line":"We have discussed the concept of **sold out** during Rocky PTG `[1]`_ and"},{"line_number":29,"context_line":"that word is more related to administrative operation, when most of one"},{"line_number":30,"context_line":"specific pool\u0027s capacity has been consumed, administrator need the"},{"line_number":31,"context_line":"functionality to mark the backend pool sold out, that is to say, when the"},{"line_number":32,"context_line":"pool is marked sold out, it will become unavailable for allocating new"}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_d2671ab7","line":29,"range":{"start_line":29,"start_character":68,"end_line":29,"end_character":71},"updated":"2018-04-09 11:28:32.000000000","message":"nit: an","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"c693a130c7b9d8e1f8a1443d4f27f16b5d40eef6","unresolved":false,"context_lines":[{"line_number":26,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":27,"context_line":""},{"line_number":28,"context_line":"We have discussed the concept of **sold out** during Rocky PTG `[1]`_ and"},{"line_number":29,"context_line":"that word is more related to administrative operation, when most of one"},{"line_number":30,"context_line":"specific pool\u0027s capacity has been consumed, administrator need the"},{"line_number":31,"context_line":"functionality to mark the backend pool sold out, that is to say, when the"},{"line_number":32,"context_line":"pool is marked sold out, it will become unavailable for allocating new"}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_15f33ce9","line":29,"range":{"start_line":29,"start_character":68,"end_line":29,"end_character":71},"in_reply_to":"bf659307_d2671ab7","updated":"2018-04-09 12:06:15.000000000","message":"Done","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"685ed8383778870a8ac23b148240334eb4b75566","unresolved":false,"context_lines":[{"line_number":27,"context_line":""},{"line_number":28,"context_line":"We have discussed the concept of **sold out** during Rocky PTG `[1]`_ and"},{"line_number":29,"context_line":"that word is more related to administrative operation, when most of one"},{"line_number":30,"context_line":"specific pool\u0027s capacity has been consumed, administrator need the"},{"line_number":31,"context_line":"functionality to mark the backend pool sold out, that is to say, when the"},{"line_number":32,"context_line":"pool is marked sold out, it will become unavailable for allocating new"},{"line_number":33,"context_line":"volumes, but end users still can perform any operations on their existing"}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_920762ca","line":30,"range":{"start_line":30,"start_character":44,"end_line":30,"end_character":57},"updated":"2018-04-09 11:28:32.000000000","message":"nit: administrators","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"c693a130c7b9d8e1f8a1443d4f27f16b5d40eef6","unresolved":false,"context_lines":[{"line_number":27,"context_line":""},{"line_number":28,"context_line":"We have discussed the concept of **sold out** during Rocky PTG `[1]`_ and"},{"line_number":29,"context_line":"that word is more related to administrative operation, when most of one"},{"line_number":30,"context_line":"specific pool\u0027s capacity has been consumed, administrator need the"},{"line_number":31,"context_line":"functionality to mark the backend pool sold out, that is to say, when the"},{"line_number":32,"context_line":"pool is marked sold out, it will become unavailable for allocating new"},{"line_number":33,"context_line":"volumes, but end users still can perform any operations on their existing"}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_d5fce4b6","line":30,"range":{"start_line":30,"start_character":44,"end_line":30,"end_character":57},"in_reply_to":"bf659307_920762ca","updated":"2018-04-09 12:06:15.000000000","message":"Done","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"685ed8383778870a8ac23b148240334eb4b75566","unresolved":false,"context_lines":[{"line_number":28,"context_line":"We have discussed the concept of **sold out** during Rocky PTG `[1]`_ and"},{"line_number":29,"context_line":"that word is more related to administrative operation, when most of one"},{"line_number":30,"context_line":"specific pool\u0027s capacity has been consumed, administrator need the"},{"line_number":31,"context_line":"functionality to mark the backend pool sold out, that is to say, when the"},{"line_number":32,"context_line":"pool is marked sold out, it will become unavailable for allocating new"},{"line_number":33,"context_line":"volumes, but end users still can perform any operations on their existing"},{"line_number":34,"context_line":"resources (extend volume, create snapshot)."},{"line_number":35,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_320e96e5","line":32,"range":{"start_line":31,"start_character":65,"end_line":32,"end_character":24},"updated":"2018-04-09 11:28:32.000000000","message":"nit: this is unnecessary, we can remove it without reducing the comprehension of the sentence.","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"c693a130c7b9d8e1f8a1443d4f27f16b5d40eef6","unresolved":false,"context_lines":[{"line_number":28,"context_line":"We have discussed the concept of **sold out** during Rocky PTG `[1]`_ and"},{"line_number":29,"context_line":"that word is more related to administrative operation, when most of one"},{"line_number":30,"context_line":"specific pool\u0027s capacity has been consumed, administrator need the"},{"line_number":31,"context_line":"functionality to mark the backend pool sold out, that is to say, when the"},{"line_number":32,"context_line":"pool is marked sold out, it will become unavailable for allocating new"},{"line_number":33,"context_line":"volumes, but end users still can perform any operations on their existing"},{"line_number":34,"context_line":"resources (extend volume, create snapshot)."},{"line_number":35,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_b51a9014","line":32,"range":{"start_line":31,"start_character":65,"end_line":32,"end_character":24},"in_reply_to":"bf659307_320e96e5","updated":"2018-04-09 12:06:15.000000000","message":"Done","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"685ed8383778870a8ac23b148240334eb4b75566","unresolved":false,"context_lines":[{"line_number":33,"context_line":"volumes, but end users still can perform any operations on their existing"},{"line_number":34,"context_line":"resources (extend volume, create snapshot)."},{"line_number":35,"context_line":""},{"line_number":36,"context_line":"Adding this feature in cinder will help our cloud vendors to write their"},{"line_number":37,"context_line":"own plugins which can filter backend based on the operation type as well"},{"line_number":38,"context_line":"as backend information."},{"line_number":39,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_9230c2a0","line":36,"range":{"start_line":36,"start_character":40,"end_line":36,"end_character":43},"updated":"2018-04-09 11:28:32.000000000","message":"nit: remove \"our\"","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"c693a130c7b9d8e1f8a1443d4f27f16b5d40eef6","unresolved":false,"context_lines":[{"line_number":33,"context_line":"volumes, but end users still can perform any operations on their existing"},{"line_number":34,"context_line":"resources (extend volume, create snapshot)."},{"line_number":35,"context_line":""},{"line_number":36,"context_line":"Adding this feature in cinder will help our cloud vendors to write their"},{"line_number":37,"context_line":"own plugins which can filter backend based on the operation type as well"},{"line_number":38,"context_line":"as backend information."},{"line_number":39,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_75d2b836","line":36,"range":{"start_line":36,"start_character":40,"end_line":36,"end_character":43},"in_reply_to":"bf659307_9230c2a0","updated":"2018-04-09 12:06:15.000000000","message":"Done","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"685ed8383778870a8ac23b148240334eb4b75566","unresolved":false,"context_lines":[{"line_number":84,"context_line":"Alternatives"},{"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":"Data model impact"},{"line_number":90,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"bf659307_9208c291","line":87,"range":{"start_line":87,"start_character":0,"end_line":87,"end_character":4},"updated":"2018-04-09 11:28:32.000000000","message":"nit: There is an alternative, as I mentioned in patch 2.","commit_id":"5018ca2b7e062b31acd9d386422c05f79e9c0beb"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"454e1de93f333e38abbd37dce20cc64e49abc1a0","unresolved":false,"context_lines":[{"line_number":33,"context_line":"can perform any operations on their existing resources (extend volume,"},{"line_number":34,"context_line":"create snapshot)."},{"line_number":35,"context_line":""},{"line_number":36,"context_line":"Adding this feature in cinder will help cloud vendors to write their"},{"line_number":37,"context_line":"own plugins which can filter backend based on the operation type as well"},{"line_number":38,"context_line":"as backend information."},{"line_number":39,"context_line":""}],"source_content_type":"text/x-rst","patch_set":4,"id":"bf659307_52020d0b","line":36,"range":{"start_line":36,"start_character":54,"end_line":36,"end_character":56},"updated":"2018-04-10 09:09:16.000000000","message":"nit: remove \"to\"","commit_id":"5cf7c6fa803aae5b4e012f0cfa3fb796d0c44141"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"454e1de93f333e38abbd37dce20cc64e49abc1a0","unresolved":false,"context_lines":[{"line_number":84,"context_line":"Alternatives"},{"line_number":85,"context_line":"------------"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"The alternative is to implement a custom scheduler driver and configure it"},{"line_number":88,"context_line":"in cinder scheduler and then use either a custom filter or the JSON filter."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"Data model impact"}],"source_content_type":"text/x-rst","patch_set":4,"id":"bf659307_120c9539","line":87,"range":{"start_line":87,"start_character":51,"end_line":87,"end_character":57},"updated":"2018-04-10 09:09:16.000000000","message":"nit: manager","commit_id":"5cf7c6fa803aae5b4e012f0cfa3fb796d0c44141"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"454e1de93f333e38abbd37dce20cc64e49abc1a0","unresolved":false,"context_lines":[{"line_number":85,"context_line":"------------"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"The alternative is to implement a custom scheduler driver and configure it"},{"line_number":88,"context_line":"in cinder scheduler and then use either a custom filter or the JSON filter."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"Data model impact"},{"line_number":91,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":4,"id":"bf659307_f221e1a0","line":88,"range":{"start_line":88,"start_character":3,"end_line":88,"end_character":19},"updated":"2018-04-10 09:09:16.000000000","message":"nit: Cinder\u0027s ``scheduler_manager``","commit_id":"5cf7c6fa803aae5b4e012f0cfa3fb796d0c44141"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"153aea468f882cd49afa8e23f8880b3da33155a6","unresolved":false,"context_lines":[{"line_number":85,"context_line":"------------"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"The alternative is to implement a custom scheduler driver and configure it"},{"line_number":88,"context_line":"in cinder scheduler and then use either a custom filter or the JSON filter."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"Data model impact"},{"line_number":91,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":4,"id":"bf659307_28c7ec14","line":88,"range":{"start_line":88,"start_character":3,"end_line":88,"end_character":19},"in_reply_to":"bf659307_f221e1a0","updated":"2018-04-11 09:59:29.000000000","message":"Done","commit_id":"5cf7c6fa803aae5b4e012f0cfa3fb796d0c44141"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"45cc5e15b06d533aa9a669db4ca6329cf643f307","unresolved":false,"context_lines":[{"line_number":85,"context_line":"------------"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"The alternative is to implement a custom scheduler manager and configure it"},{"line_number":88,"context_line":"in ``cinder_scheduler`` and then use either a custom filter or the JSON"},{"line_number":89,"context_line":"filter."},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"Data model impact"}],"source_content_type":"text/x-rst","patch_set":5,"id":"bf659307_b9957034","line":88,"range":{"start_line":88,"start_character":5,"end_line":88,"end_character":21},"updated":"2018-04-11 09:34:23.000000000","message":"-1: This configuration option doesn\u0027t exist, like I mention in patch 4, it\u0027s ``scheduler_manager``.","commit_id":"c1a062c9b0d3401ad852384d32cbfa76c0926b0c"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"153aea468f882cd49afa8e23f8880b3da33155a6","unresolved":false,"context_lines":[{"line_number":85,"context_line":"------------"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"The alternative is to implement a custom scheduler manager and configure it"},{"line_number":88,"context_line":"in ``cinder_scheduler`` and then use either a custom filter or the JSON"},{"line_number":89,"context_line":"filter."},{"line_number":90,"context_line":""},{"line_number":91,"context_line":"Data model impact"}],"source_content_type":"text/x-rst","patch_set":5,"id":"bf659307_148a2169","line":88,"range":{"start_line":88,"start_character":5,"end_line":88,"end_character":21},"in_reply_to":"bf659307_b9957034","updated":"2018-04-11 09:59:29.000000000","message":"Oops!","commit_id":"c1a062c9b0d3401ad852384d32cbfa76c0926b0c"}],"specs/rocky/support-filter-backend-on-requesttype.rst":[{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"ba1c3f17c73384c328d2ae7f546a97a6c781fade","unresolved":false,"context_lines":[{"line_number":12,"context_line":""},{"line_number":13,"context_line":"This blueprint proposes to support filter backend based on request type."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"Problem description"},{"line_number":16,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":17,"context_line":""},{"line_number":18,"context_line":"We have discussed the concept of **sold out** during Rocky PTG `[1]`_ and"}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_ba2a06d8","line":15,"updated":"2018-04-02 15:17:58.000000000","message":"-1: Most of the problem description seems to be the Use case, the problem description should mostly talk about filters that depend on the action as much as the backend information and how we are not currently passing that information.","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b2f905a50e6636330f746e64008ca95c2e237af5","unresolved":false,"context_lines":[{"line_number":12,"context_line":""},{"line_number":13,"context_line":"This blueprint proposes to support filter backend based on request type."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"Problem description"},{"line_number":16,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":17,"context_line":""},{"line_number":18,"context_line":"We have discussed the concept of **sold out** during Rocky PTG `[1]`_ and"}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_0966442b","line":15,"in_reply_to":"bf659307_ba2a06d8","updated":"2018-04-03 11:18:07.000000000","message":"Thanks","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"ba1c3f17c73384c328d2ae7f546a97a6c781fade","unresolved":false,"context_lines":[{"line_number":20,"context_line":"specific pool\u0027s capacity has been consumed, administrator need the"},{"line_number":21,"context_line":"functionality to mark the backend pool sold out, that is to say, when the"},{"line_number":22,"context_line":"pool is marked sold out, it will become unavailable for"},{"line_number":23,"context_line":"**new users or customers**, but the pool\u0027s existing users still can perform"},{"line_number":24,"context_line":"any capacity consuming operations on their resources (extend volume,"},{"line_number":25,"context_line":"create snapshot)."},{"line_number":26,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_3abb56c5","line":23,"range":{"start_line":23,"start_character":0,"end_line":23,"end_character":27},"updated":"2018-04-02 15:17:58.000000000","message":"-1: As far as I remember it\u0027s not about becoming unavailable for new users, but not allowing new volumes","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b2f905a50e6636330f746e64008ca95c2e237af5","unresolved":false,"context_lines":[{"line_number":20,"context_line":"specific pool\u0027s capacity has been consumed, administrator need the"},{"line_number":21,"context_line":"functionality to mark the backend pool sold out, that is to say, when the"},{"line_number":22,"context_line":"pool is marked sold out, it will become unavailable for"},{"line_number":23,"context_line":"**new users or customers**, but the pool\u0027s existing users still can perform"},{"line_number":24,"context_line":"any capacity consuming operations on their resources (extend volume,"},{"line_number":25,"context_line":"create snapshot)."},{"line_number":26,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_e96e70fe","line":23,"range":{"start_line":23,"start_character":0,"end_line":23,"end_character":27},"in_reply_to":"bf659307_3abb56c5","updated":"2018-04-03 11:18:07.000000000","message":"Correct!","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"ba1c3f17c73384c328d2ae7f546a97a6c781fade","unresolved":false,"context_lines":[{"line_number":34,"context_line":"Use Cases"},{"line_number":35,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":36,"context_line":""},{"line_number":37,"context_line":"Scheduler plugins can filter backend based on request type."},{"line_number":38,"context_line":""},{"line_number":39,"context_line":"Proposed change"},{"line_number":40,"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":2,"id":"bf659307_da3c9a30","line":37,"updated":"2018-04-02 15:17:58.000000000","message":"-1: This is missing the Use Case for requiring the ``request_type``, which should be the sold out example.","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b2f905a50e6636330f746e64008ca95c2e237af5","unresolved":false,"context_lines":[{"line_number":34,"context_line":"Use Cases"},{"line_number":35,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":36,"context_line":""},{"line_number":37,"context_line":"Scheduler plugins can filter backend based on request type."},{"line_number":38,"context_line":""},{"line_number":39,"context_line":"Proposed change"},{"line_number":40,"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":2,"id":"bf659307_497f5caa","line":37,"in_reply_to":"bf659307_da3c9a30","updated":"2018-04-03 11:18:07.000000000","message":"done","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"ba1c3f17c73384c328d2ae7f546a97a6c781fade","unresolved":false,"context_lines":[{"line_number":39,"context_line":"Proposed change"},{"line_number":40,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":41,"context_line":""},{"line_number":42,"context_line":"New attribute ``request_type`` will be added into ``RequestSpec`` object"},{"line_number":43,"context_line":"and plugins can filter backend based on backend states as well as"},{"line_number":44,"context_line":"request type:"},{"line_number":45,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_ba8be6c0","line":42,"range":{"start_line":42,"start_character":16,"end_line":42,"end_character":28},"updated":"2018-04-02 15:17:58.000000000","message":"?: Should we call these actions or operations instead?","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"ba1c3f17c73384c328d2ae7f546a97a6c781fade","unresolved":false,"context_lines":[{"line_number":69,"context_line":"6. migrate_volume"},{"line_number":70,"context_line":"7. retype_volume"},{"line_number":71,"context_line":"8. manage_existing_snapshot"},{"line_number":72,"context_line":""},{"line_number":73,"context_line":"Alternatives"},{"line_number":74,"context_line":"------------"},{"line_number":75,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_7a06ae22","line":72,"updated":"2018-04-02 15:17:58.000000000","message":"I think we should have an automatic mechanism to inject the request_type, we can do this with a class decorator that checks via introspection the class\u0027 methods and looks for parameter filter_properties or group_filter_properties and filter_properties_list, if they are found it returns a decorated method that will automatically inject the ``request_type`` based on the manager\u0027s method name, and if not returns the unmodified method.\n\nWith this we can ensure that as long as we keep calling the parameter filter_properties we won\u0027t miss adding the field to the dictionary.","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b2f905a50e6636330f746e64008ca95c2e237af5","unresolved":false,"context_lines":[{"line_number":69,"context_line":"6. migrate_volume"},{"line_number":70,"context_line":"7. retype_volume"},{"line_number":71,"context_line":"8. manage_existing_snapshot"},{"line_number":72,"context_line":""},{"line_number":73,"context_line":"Alternatives"},{"line_number":74,"context_line":"------------"},{"line_number":75,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_983938d8","line":72,"in_reply_to":"bf659307_7a06ae22","updated":"2018-04-03 11:18:07.000000000","message":"I am wondering if we can just add *the* decorator to those methods rather than the whole class. It\u0027s true that if we add this in the class it would be smarter, but the parameter might change in the new method. And inspecting the whole methods in class in order to find out these eight (for now) methods seems not a good deal.","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"685ed8383778870a8ac23b148240334eb4b75566","unresolved":false,"context_lines":[{"line_number":69,"context_line":"6. migrate_volume"},{"line_number":70,"context_line":"7. retype_volume"},{"line_number":71,"context_line":"8. manage_existing_snapshot"},{"line_number":72,"context_line":""},{"line_number":73,"context_line":"Alternatives"},{"line_number":74,"context_line":"------------"},{"line_number":75,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_12cb524a","line":72,"in_reply_to":"bf659307_983938d8","updated":"2018-04-09 11:28:32.000000000","message":"The class will only be instantiated once, since it\u0027s the manager, and therefore the code that checks the methods will only run once, so you won\u0027t get any real performance improvement doing it on a per method case (except on the service start), but I am OK doing it either way.","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"ba1c3f17c73384c328d2ae7f546a97a6c781fade","unresolved":false,"context_lines":[{"line_number":73,"context_line":"Alternatives"},{"line_number":74,"context_line":"------------"},{"line_number":75,"context_line":""},{"line_number":76,"context_line":"One of the alternative is to fully support this in cinder and in-tree"},{"line_number":77,"context_line":"plugins, but the biggest issue is the sold out strategy could be different"},{"line_number":78,"context_line":"among different OpenStack vendors, They may need different inputs or"},{"line_number":79,"context_line":"backend states. Only have request type supported in the framework means"}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_7a4eaec1","line":76,"updated":"2018-04-02 15:17:58.000000000","message":"-1: This is not an alternative, the only alternative would be to implement a custom scheduler manager and configure it in cinder_scheduler and then use either a custom filter or the JSON filter.\n\nThis alternative is inferior to the proposed change, since with the proposed change operators could just use the JSON filter without having to write any Python code.","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b2f905a50e6636330f746e64008ca95c2e237af5","unresolved":false,"context_lines":[{"line_number":73,"context_line":"Alternatives"},{"line_number":74,"context_line":"------------"},{"line_number":75,"context_line":""},{"line_number":76,"context_line":"One of the alternative is to fully support this in cinder and in-tree"},{"line_number":77,"context_line":"plugins, but the biggest issue is the sold out strategy could be different"},{"line_number":78,"context_line":"among different OpenStack vendors, They may need different inputs or"},{"line_number":79,"context_line":"backend states. Only have request type supported in the framework means"}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_7d8bf6c1","line":76,"in_reply_to":"bf659307_7a4eaec1","updated":"2018-04-03 11:18:07.000000000","message":"but, even with this proposal, the JSON filter still needs an update in code to support filter backend based on operation type.","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"685ed8383778870a8ac23b148240334eb4b75566","unresolved":false,"context_lines":[{"line_number":73,"context_line":"Alternatives"},{"line_number":74,"context_line":"------------"},{"line_number":75,"context_line":""},{"line_number":76,"context_line":"One of the alternative is to fully support this in cinder and in-tree"},{"line_number":77,"context_line":"plugins, but the biggest issue is the sold out strategy could be different"},{"line_number":78,"context_line":"among different OpenStack vendors, They may need different inputs or"},{"line_number":79,"context_line":"backend states. Only have request type supported in the framework means"}],"source_content_type":"text/x-rst","patch_set":2,"id":"bf659307_d2c39a56","line":76,"in_reply_to":"bf659307_7d8bf6c1","updated":"2018-04-09 11:28:32.000000000","message":"It would need to pass it, but we wouldn\u0027t need to change the Cinder code, because that change would be done by the customer on its own Scheduler code (not modifying the cinder code but creating it\u0027s own module) and then configured in cinder.conf to be used.\n\nBut I believe that alternative to be inferior.","commit_id":"6141c8991de8bac5549e0c6adfb9600743b20c5a"}]}
