)]}'
{"specs/liberty/volume-encryption-metadata.rst":[{"author":{"_account_id":6804,"name":"bruce-benjamin","email":"bruce.benjamin@jhuapl.edu","username":"bruce-benjamin"},"change_message_id":"c7a356630386f203573cd03ac330f9dc2ca79583","unresolved":false,"context_lines":[{"line_number":27,"context_line":"to user needs. An example would be to increase the key length of an encryption"},{"line_number":28,"context_line":"key for a given encryption scheme. Volumes rely on volume type information to"},{"line_number":29,"context_line":"obtain the parameters needed to decrypt their contents. If a volume type were"},{"line_number":30,"context_line":"to be change, the volumes previous to that change would effectively receive"},{"line_number":31,"context_line":"the wrong parameters as a result they cannot be decrypted."},{"line_number":32,"context_line":""},{"line_number":33,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7a016987_dce7bcaa","line":30,"updated":"2015-05-27 20:36:14.000000000","message":"s/change/changed/","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"65005ce6846c3c410eb2d3072c7d20872740db48","unresolved":false,"context_lines":[{"line_number":27,"context_line":"to user needs. An example would be to increase the key length of an encryption"},{"line_number":28,"context_line":"key for a given encryption scheme. Volumes rely on volume type information to"},{"line_number":29,"context_line":"obtain the parameters needed to decrypt their contents. If a volume type were"},{"line_number":30,"context_line":"to be change, the volumes previous to that change would effectively receive"},{"line_number":31,"context_line":"the wrong parameters as a result they cannot be decrypted."},{"line_number":32,"context_line":""},{"line_number":33,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"3a50d1a3_36a9efdf","line":30,"in_reply_to":"7a016987_dce7bcaa","updated":"2015-07-29 13:23:18.000000000","message":"Tweaked wording.","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":12924,"name":"Patrick East","email":"east.patrick@gmail.com","username":"patrick.east"},"change_message_id":"a6ff13359fc501351c5ce3a7a4e84acb5902ab24","unresolved":false,"context_lines":[{"line_number":35,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":36,"context_line":""},{"line_number":37,"context_line":"This proposed change copies the encryption metadata from an encrypted volume"},{"line_number":38,"context_line":"type when a volume is created. The encryption metadata is then stored with"},{"line_number":39,"context_line":"snapshots, clones, backups, etc., but there is no need to retain the original"},{"line_number":40,"context_line":"volume\u0027s type."},{"line_number":41,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7a016987_2aab508b","line":38,"updated":"2015-05-14 21:54:23.000000000","message":"Just to make sure I am understanding correctly the plan is to duplicate the encryption info for every volume?\n\nMaybe an alternative approach to this (and there is a chance I\u0027m missing something, so please ignore if i\u0027m way off the mark). What If the volume_type had a reference to an encryption entry, and any volumes created from that type pick up a reference to the encryption entry too. Then any time the encryption data is changed you just update the reference on the type and new volumes pick up the new updated reference. That would at least allow for modifying the data since volumes go straight to their encryption entry skipping the volume type (similar to what you are proposing but no new table and no duplicating the data).\n\nDoes this need to handle deleting types too? Its mentioned in the problem description, but I\u0027m not sure that the proposed solution handles that (again, maybe I\u0027m missing something). My understanding is that the volume always needs to keep its type for more than just encryption info. If that is the case maybe we could do even softer deletes than we do now, so allow for a \u0027delete\u0027 of a type that is in use, but it doesn\u0027t really remove it unless/until all volumes that reference it are deleted. Maybe we put a flag on the type thats like a \u0027pending_delete\u0027 to indicate it can\u0027t be used for new volumes but is still reference by old ones.","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":170,"name":"Mike Perez","email":"thingee@gmail.com","username":"thingee"},"change_message_id":"e2b3ae1f52293918f013b3854d1d76d8e1693b33","unresolved":false,"context_lines":[{"line_number":37,"context_line":"This proposed change copies the encryption metadata from an encrypted volume"},{"line_number":38,"context_line":"type when a volume is created. The encryption metadata is then stored with"},{"line_number":39,"context_line":"snapshots, clones, backups, etc., but there is no need to retain the original"},{"line_number":40,"context_line":"volume\u0027s type."},{"line_number":41,"context_line":""},{"line_number":42,"context_line":"Alternatives"},{"line_number":43,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7ad8896b_f7dc321a","line":40,"updated":"2015-05-11 17:33:19.000000000","message":"Can you list the use cases for this under the new use case section?\n\nhttps://github.com/openstack/cinder-specs/blob/master/specs/template.rst#use-cases","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":6804,"name":"bruce-benjamin","email":"bruce.benjamin@jhuapl.edu","username":"bruce-benjamin"},"change_message_id":"c7a356630386f203573cd03ac330f9dc2ca79583","unresolved":false,"context_lines":[{"line_number":43,"context_line":"------------"},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"The current alternative would be to leave the system as is, which precludes"},{"line_number":46,"context_line":"modifying or deleting an encrypted volume type once it has been used."},{"line_number":47,"context_line":""},{"line_number":48,"context_line":"Data model impact"},{"line_number":49,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7a016987_dc59fc8f","line":46,"updated":"2015-05-27 20:36:14.000000000","message":"Consider appending the following:  \"The risk is that If the encrypted volume type is accidentally modified or deleted, all inactive encrypted volumes using that type would be rendered unreadable and useless.\"","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"65005ce6846c3c410eb2d3072c7d20872740db48","unresolved":false,"context_lines":[{"line_number":43,"context_line":"------------"},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"The current alternative would be to leave the system as is, which precludes"},{"line_number":46,"context_line":"modifying or deleting an encrypted volume type once it has been used."},{"line_number":47,"context_line":""},{"line_number":48,"context_line":"Data model impact"},{"line_number":49,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3a50d1a3_96fd633c","line":46,"in_reply_to":"7a016987_dc59fc8f","updated":"2015-07-29 13:23:18.000000000","message":"Done","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":170,"name":"Mike Perez","email":"thingee@gmail.com","username":"thingee"},"change_message_id":"e2b3ae1f52293918f013b3854d1d76d8e1693b33","unresolved":false,"context_lines":[{"line_number":48,"context_line":"Data model impact"},{"line_number":49,"context_line":"-----------------"},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"A new cinder database table ( \"volume_encryption_metadata\" ) is proposed for"},{"line_number":52,"context_line":"storing each volume\u0027s corresponding encryption metadata."},{"line_number":53,"context_line":""},{"line_number":54,"context_line":"The initial migration will rely on copying existing volume encryption metadata"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7ad8896b_f778b2fa","line":51,"updated":"2015-05-11 17:33:19.000000000","message":"Why can\u0027t you use the existing volume metadata table?","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"65005ce6846c3c410eb2d3072c7d20872740db48","unresolved":false,"context_lines":[{"line_number":48,"context_line":"Data model impact"},{"line_number":49,"context_line":"-----------------"},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"A new cinder database table ( \"volume_encryption_metadata\" ) is proposed for"},{"line_number":52,"context_line":"storing each volume\u0027s corresponding encryption metadata."},{"line_number":53,"context_line":""},{"line_number":54,"context_line":"The initial migration will rely on copying existing volume encryption metadata"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3a50d1a3_eccdfb85","line":51,"in_reply_to":"7a016987_7c615da8","updated":"2015-07-29 13:23:18.000000000","message":"Done","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"220f9d94cd6871e5dd7ab554c8743fbc9608e0c7","unresolved":false,"context_lines":[{"line_number":48,"context_line":"Data model impact"},{"line_number":49,"context_line":"-----------------"},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"A new cinder database table ( \"volume_encryption_metadata\" ) is proposed for"},{"line_number":52,"context_line":"storing each volume\u0027s corresponding encryption metadata."},{"line_number":53,"context_line":""},{"line_number":54,"context_line":"The initial migration will rely on copying existing volume encryption metadata"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7a016987_7c615da8","line":51,"in_reply_to":"7ad8896b_f778b2fa","updated":"2015-05-22 22:09:06.000000000","message":"I think that\u0027s possible, and it would be more elegant from the database perspective.","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":170,"name":"Mike Perez","email":"thingee@gmail.com","username":"thingee"},"change_message_id":"e2b3ae1f52293918f013b3854d1d76d8e1693b33","unresolved":false,"context_lines":[{"line_number":119,"context_line":"The volume encryption tests (VolumeTestCase) will be expanded for this new"},{"line_number":120,"context_line":"feature. Volume creation and deletion will be tested for encrypted volumes."},{"line_number":121,"context_line":"Unit tests and tempest tests will be expanded to verify this feature is"},{"line_number":122,"context_line":"properly working."},{"line_number":123,"context_line":""},{"line_number":124,"context_line":""},{"line_number":125,"context_line":"Documentation Impact"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7ad8896b_97872ed7","line":122,"updated":"2015-05-11 17:33:19.000000000","message":"Would like tempest tests as well.","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"220f9d94cd6871e5dd7ab554c8743fbc9608e0c7","unresolved":false,"context_lines":[{"line_number":119,"context_line":"The volume encryption tests (VolumeTestCase) will be expanded for this new"},{"line_number":120,"context_line":"feature. Volume creation and deletion will be tested for encrypted volumes."},{"line_number":121,"context_line":"Unit tests and tempest tests will be expanded to verify this feature is"},{"line_number":122,"context_line":"properly working."},{"line_number":123,"context_line":""},{"line_number":124,"context_line":""},{"line_number":125,"context_line":"Documentation Impact"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7a016987_fce0ad2a","line":122,"in_reply_to":"7ad8896b_97872ed7","updated":"2015-05-22 22:09:06.000000000","message":"The existing Tempest tests should provide appropriate coverage for this change. See https://review.openstack.org/#/c/58018/","commit_id":"16e5cb001f71290630c33f40b1f145cfd403d114"},{"author":{"_account_id":7111,"name":"hadi esiely","email":"hadi.esiely-barrera@jhuapl.edu","username":"HEsiely"},"change_message_id":"dc004c6e4980d27cba527c0a02f4257c5e521b43","unresolved":false,"context_lines":[{"line_number":30,"context_line":"to be changed, existing volumes would effectively receive the wrong parameters,"},{"line_number":31,"context_line":"and as a result, they cannot be decrypted."},{"line_number":32,"context_line":""},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"Proposed change"},{"line_number":35,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":36,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"3a50d1a3_2242a60f","line":33,"updated":"2015-07-29 14:13:28.000000000","message":"We are missing the usecases section here.","commit_id":"30e1315bc3a6de1133805a574b16e212a750a54f"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"3ebed549e984f2735f71d31133c7d9ad4d6900be","unresolved":false,"context_lines":[{"line_number":30,"context_line":"to be changed, existing volumes would effectively receive the wrong parameters,"},{"line_number":31,"context_line":"and as a result, they cannot be decrypted."},{"line_number":32,"context_line":""},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"Proposed change"},{"line_number":35,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":36,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"1a4dcd0f_674f02be","line":33,"in_reply_to":"3a50d1a3_2242a60f","updated":"2015-08-04 14:44:43.000000000","message":"Working on writing them...","commit_id":"30e1315bc3a6de1133805a574b16e212a750a54f"},{"author":{"_account_id":17444,"name":"Jay.Shen","email":"shenhong09@huawei.com","username":"Jay.Shen"},"change_message_id":"2f2c259da4317777285bbb686b8ca9ae45dab826","unresolved":false,"context_lines":[{"line_number":37,"context_line":"This proposed change associates encryption metadata with each encrypted volume"},{"line_number":38,"context_line":"in addition to the existing association with the encrypted volume type. The"},{"line_number":39,"context_line":"encryption metadata is also associated with snapshots, clones, backups, etc."},{"line_number":40,"context_line":"that require the encrypton metadata for decryption (i.e., to use the snapshot,"},{"line_number":41,"context_line":"clone, backup, etc.)."},{"line_number":42,"context_line":""},{"line_number":43,"context_line":"When implemented, this change avoids a number of existing special cases in the"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1a4dcd0f_7387e40f","line":40,"updated":"2015-08-04 06:07:32.000000000","message":"S/encrypton/encryption","commit_id":"30e1315bc3a6de1133805a574b16e212a750a54f"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"3ebed549e984f2735f71d31133c7d9ad4d6900be","unresolved":false,"context_lines":[{"line_number":37,"context_line":"This proposed change associates encryption metadata with each encrypted volume"},{"line_number":38,"context_line":"in addition to the existing association with the encrypted volume type. The"},{"line_number":39,"context_line":"encryption metadata is also associated with snapshots, clones, backups, etc."},{"line_number":40,"context_line":"that require the encrypton metadata for decryption (i.e., to use the snapshot,"},{"line_number":41,"context_line":"clone, backup, etc.)."},{"line_number":42,"context_line":""},{"line_number":43,"context_line":"When implemented, this change avoids a number of existing special cases in the"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1a4dcd0f_07107ea7","line":40,"in_reply_to":"1a4dcd0f_7387e40f","updated":"2015-08-04 14:44:43.000000000","message":"Done","commit_id":"30e1315bc3a6de1133805a574b16e212a750a54f"},{"author":{"_account_id":6804,"name":"bruce-benjamin","email":"bruce.benjamin@jhuapl.edu","username":"bruce-benjamin"},"change_message_id":"e35a10dd7c18918312fcf0030c9b7844843ed0d3","unresolved":false,"context_lines":[{"line_number":80,"context_line":"other relations such as snapshots and backups, or the volume_id attribute of"},{"line_number":81,"context_line":"these relations may be used to retrieve the original volume and encryption"},{"line_number":82,"context_line":"metadata (assuming that the volume still exists in the database)."},{"line_number":83,"context_line":""},{"line_number":84,"context_line":"REST API impact"},{"line_number":85,"context_line":"---------------"},{"line_number":86,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"3a50d1a3_b7159640","line":83,"updated":"2015-07-29 13:57:30.000000000","message":"I suggest the following changes to clarify the paragraph above:\ns/will be/will also be/\ns/attribute is/attribute is also/","commit_id":"30e1315bc3a6de1133805a574b16e212a750a54f"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"3ebed549e984f2735f71d31133c7d9ad4d6900be","unresolved":false,"context_lines":[{"line_number":80,"context_line":"other relations such as snapshots and backups, or the volume_id attribute of"},{"line_number":81,"context_line":"these relations may be used to retrieve the original volume and encryption"},{"line_number":82,"context_line":"metadata (assuming that the volume still exists in the database)."},{"line_number":83,"context_line":""},{"line_number":84,"context_line":"REST API impact"},{"line_number":85,"context_line":"---------------"},{"line_number":86,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"1a4dcd0f_47644648","line":83,"in_reply_to":"3a50d1a3_b7159640","updated":"2015-08-04 14:44:43.000000000","message":"Done","commit_id":"30e1315bc3a6de1133805a574b16e212a750a54f"},{"author":{"_account_id":13868,"name":"Danny Wilson","email":"daniel.wilson@purestorage.com","username":"dannywilson"},"change_message_id":"a908a2a581f41924da09262b45de786c6dd41788","unresolved":false,"context_lines":[{"line_number":47,"context_line":"    once a volume has been created from that type because deleting the"},{"line_number":48,"context_line":"    encrypted volume type would prevent the encrypted volume from being"},{"line_number":49,"context_line":"    decrypted. This change will allow encrypted volume types to be deleted just"},{"line_number":50,"context_line":"    like ordinary volume types."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"Modify encryption metadata for encrypted volume type"},{"line_number":53,"context_line":"    As explained above, the encryption metadata associated with a volume type"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1a4dcd0f_53f19a69","line":50,"updated":"2015-08-13 22:46:16.000000000","message":"How do you handle deleting the \"encryption\" database entries if it has no references?  I am thinking along these lines:\n\n1. Encrypted volume type is created with reference to encryption table.\n\n2. Volume is created with that vol type.\n\n3. Vol type is deleted but entry in database is still there so volume can access it.\n\n4. Volume is deleted.\n\nIs the entry in encryption database now just taking up space?  Is there a way to delete it?\n\nIt may be my lack of understanding of encryption and if that is the case please detail the steps here on how it is deleted.","commit_id":"1fc06ce14951a798e1e6b0a7d46a8d756e56d69f"},{"author":{"_account_id":13868,"name":"Danny Wilson","email":"daniel.wilson@purestorage.com","username":"dannywilson"},"change_message_id":"8f55a0b6361ed5ff74e0e7d0cf54665ea2f7423d","unresolved":false,"context_lines":[{"line_number":47,"context_line":"    once a volume has been created from that type because deleting the"},{"line_number":48,"context_line":"    encrypted volume type would prevent the encrypted volume from being"},{"line_number":49,"context_line":"    decrypted. This change will allow encrypted volume types to be deleted just"},{"line_number":50,"context_line":"    like ordinary volume types."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"Modify encryption metadata for encrypted volume type"},{"line_number":53,"context_line":"    As explained above, the encryption metadata associated with a volume type"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1a4dcd0f_44c6894a","line":50,"in_reply_to":"1a4dcd0f_11f1ae95","updated":"2015-08-14 17:26:30.000000000","message":"Can you add those notes into the spec please so we keep track of them?\n\nMaybe in the deployer section so there is some documentation about how to do this?\n\nThanks.","commit_id":"1fc06ce14951a798e1e6b0a7d46a8d756e56d69f"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"7b251adb60e0e9a59683ce746da1d4bd2d08eda1","unresolved":false,"context_lines":[{"line_number":47,"context_line":"    once a volume has been created from that type because deleting the"},{"line_number":48,"context_line":"    encrypted volume type would prevent the encrypted volume from being"},{"line_number":49,"context_line":"    decrypted. This change will allow encrypted volume types to be deleted just"},{"line_number":50,"context_line":"    like ordinary volume types."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"Modify encryption metadata for encrypted volume type"},{"line_number":53,"context_line":"    As explained above, the encryption metadata associated with a volume type"}],"source_content_type":"text/x-rst","patch_set":4,"id":"fa1b9901_fcce73b3","line":50,"in_reply_to":"1a4dcd0f_44c6894a","updated":"2015-08-19 20:16:02.000000000","message":"Sure. I\u0027ll add them and post them as part of the next major revision.","commit_id":"1fc06ce14951a798e1e6b0a7d46a8d756e56d69f"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"04e4beb97a6fdc52b681a8a3c52b8096d52a703a","unresolved":false,"context_lines":[{"line_number":47,"context_line":"    once a volume has been created from that type because deleting the"},{"line_number":48,"context_line":"    encrypted volume type would prevent the encrypted volume from being"},{"line_number":49,"context_line":"    decrypted. This change will allow encrypted volume types to be deleted just"},{"line_number":50,"context_line":"    like ordinary volume types."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"Modify encryption metadata for encrypted volume type"},{"line_number":53,"context_line":"    As explained above, the encryption metadata associated with a volume type"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1a4dcd0f_11f1ae95","line":50,"in_reply_to":"1a4dcd0f_53f19a69","updated":"2015-08-14 13:06:46.000000000","message":"The number of \"orphaned\" entries should be minimal. Under the current implementation, only administrators can define new encrypted volume types so there\u0027s little danger of (e.g.) resource exhaustion due to users repeatedly creating and deleting encrypted volume types.\n\nIf there\u0027s concern, one could easily scan to see if any volumes, snapshots, volume types, etc. are still in using the encryption entry and remove it if they are not. (Note: I\u0027d need to verify that this solution would work with backups -- i.e., the backup stores sufficient metadata to recreate the encryption entry if necessary.)","commit_id":"1fc06ce14951a798e1e6b0a7d46a8d756e56d69f"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"06137fe45c6d84515a5072dde801d6f9543c7029","unresolved":false,"context_lines":[{"line_number":187,"context_line":"Documentation Impact"},{"line_number":188,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":189,"context_line":""},{"line_number":190,"context_line":"None"},{"line_number":191,"context_line":""},{"line_number":192,"context_line":""},{"line_number":193,"context_line":"References"}],"source_content_type":"text/x-rst","patch_set":4,"id":"fa1b9901_3868fbb8","line":190,"updated":"2015-08-19 19:55:41.000000000","message":"I disagree that there is no documentation impact.  If the existing limitation has already been documented that will need to be removed and/or it seems like we need to document how this now can be used.","commit_id":"1fc06ce14951a798e1e6b0a7d46a8d756e56d69f"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"ca418ccc02b0cc0b058eb6f0925badaa4d800510","unresolved":false,"context_lines":[{"line_number":187,"context_line":"Documentation Impact"},{"line_number":188,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":189,"context_line":""},{"line_number":190,"context_line":"None"},{"line_number":191,"context_line":""},{"line_number":192,"context_line":""},{"line_number":193,"context_line":"References"}],"source_content_type":"text/x-rst","patch_set":4,"id":"9a1a9d01_b232d25c","line":190,"in_reply_to":"fa1b9901_3868fbb8","updated":"2015-09-29 15:30:42.000000000","message":"Done","commit_id":"1fc06ce14951a798e1e6b0a7d46a8d756e56d69f"}],"specs/mitaka/volume-encryption-metadata.rst":[{"author":{"_account_id":15961,"name":"lisali","email":"xiaoyan.li@intel.com","username":"lisali"},"change_message_id":"8dfd8f2b58c6e6b333235c77471a5d45462409f6","unresolved":false,"context_lines":[{"line_number":47,"context_line":"    once a volume has been created from that type because deleting the"},{"line_number":48,"context_line":"    encrypted volume type would prevent the encrypted volume from being"},{"line_number":49,"context_line":"    decrypted. This change will allow encrypted volume types to be deleted just"},{"line_number":50,"context_line":"    like ordinary volume types."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"Modify encryption metadata for encrypted volume type"},{"line_number":53,"context_line":"    As explained above, the encryption metadata associated with a volume type"}],"source_content_type":"text/x-rst","patch_set":6,"id":"9a8ffd7b_d90cba44","line":50,"updated":"2015-11-25 02:49:07.000000000","message":"Currently if a volume type is in-use, it can\u0027t be deleted:\n\nERROR: Target volume type is still in use.","commit_id":"9f8347c0859fa2bdac13d8837f96196284d4c2bc"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"b0e74ec66d7919e09d26111bb127953f974511a3","unresolved":false,"context_lines":[{"line_number":47,"context_line":"    once a volume has been created from that type because deleting the"},{"line_number":48,"context_line":"    encrypted volume type would prevent the encrypted volume from being"},{"line_number":49,"context_line":"    decrypted. This change will allow encrypted volume types to be deleted just"},{"line_number":50,"context_line":"    like ordinary volume types."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"Modify encryption metadata for encrypted volume type"},{"line_number":53,"context_line":"    As explained above, the encryption metadata associated with a volume type"}],"source_content_type":"text/x-rst","patch_set":6,"id":"9a8ffd7b_2f8fe9cd","line":50,"in_reply_to":"9a8ffd7b_d90cba44","updated":"2015-11-30 13:33:19.000000000","message":"Yes, but an encrypted volume type cannot be deleted even if it isn\u0027t currently in use. We want to remove that restriction.","commit_id":"9f8347c0859fa2bdac13d8837f96196284d4c2bc"},{"author":{"_account_id":15961,"name":"lisali","email":"xiaoyan.li@intel.com","username":"lisali"},"change_message_id":"8dfd8f2b58c6e6b333235c77471a5d45462409f6","unresolved":false,"context_lines":[{"line_number":60,"context_line":"    with the new settings. For example, an organization may currently mandate"},{"line_number":61,"context_line":"    128-bit encryption but eventually increase the key size to 256 bits. With"},{"line_number":62,"context_line":"    the proposed code change, this modification can be handled transparently to"},{"line_number":63,"context_line":"    end users: they continue to use the same volume type but all new encrypted"},{"line_number":64,"context_line":"    volumes use the larger key size."},{"line_number":65,"context_line":""},{"line_number":66,"context_line":"Proposed change"}],"source_content_type":"text/x-rst","patch_set":6,"id":"9a8ffd7b_7d20199c","line":63,"updated":"2015-11-25 02:49:07.000000000","message":"Why not create a new volume type, and flag the old volume type deprecated in description?","commit_id":"9f8347c0859fa2bdac13d8837f96196284d4c2bc"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"b0e74ec66d7919e09d26111bb127953f974511a3","unresolved":false,"context_lines":[{"line_number":60,"context_line":"    with the new settings. For example, an organization may currently mandate"},{"line_number":61,"context_line":"    128-bit encryption but eventually increase the key size to 256 bits. With"},{"line_number":62,"context_line":"    the proposed code change, this modification can be handled transparently to"},{"line_number":63,"context_line":"    end users: they continue to use the same volume type but all new encrypted"},{"line_number":64,"context_line":"    volumes use the larger key size."},{"line_number":65,"context_line":""},{"line_number":66,"context_line":"Proposed change"}],"source_content_type":"text/x-rst","patch_set":6,"id":"9a8ffd7b_12bd6a02","line":63,"in_reply_to":"9a8ffd7b_7d20199c","updated":"2015-11-30 13:33:19.000000000","message":"That is certainly an alternative (which we should have acknowledged in the \"Alternatives\" section below). The \"deprecated\" volume types could still not be deleted (see above), which admits the possibility of a proliferation of deprecated volume types.","commit_id":"9f8347c0859fa2bdac13d8837f96196284d4c2bc"}],"specs/newton/volume-encryption-metadata.rst":[{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"2b86216737cd50c86a8d78137b62d36d3da942a6","unresolved":false,"context_lines":[{"line_number":111,"context_line":"the \"id\" attribute of the \"encryption\" relation. Similar attributes may be"},{"line_number":112,"context_line":"added to other relations such as snapshots and backups, or the volume_id"},{"line_number":113,"context_line":"attribute of these relations may be used to retrieve the original volume and"},{"line_number":114,"context_line":"encryption metadata (assuming that the volume still exists in the database)."},{"line_number":115,"context_line":""},{"line_number":116,"context_line":"REST API impact"},{"line_number":117,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"1af94dfe_dd05ccc2","line":114,"range":{"start_line":114,"start_character":0,"end_line":114,"end_character":76},"updated":"2016-03-16 18:12:14.000000000","message":"How do we add this to a live system with potentially tens of thousands of volume records?","commit_id":"1669ed57301ed746632948156bc9e73237534b17"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"fc6c1bbdabad9f9023b0697448c9bd5b13a21fcc","unresolved":false,"context_lines":[{"line_number":111,"context_line":"the \"id\" attribute of the \"encryption\" relation. Similar attributes may be"},{"line_number":112,"context_line":"added to other relations such as snapshots and backups, or the volume_id"},{"line_number":113,"context_line":"attribute of these relations may be used to retrieve the original volume and"},{"line_number":114,"context_line":"encryption metadata (assuming that the volume still exists in the database)."},{"line_number":115,"context_line":""},{"line_number":116,"context_line":"REST API impact"},{"line_number":117,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"fa0719c6_b3472176","line":114,"range":{"start_line":114,"start_character":0,"end_line":114,"end_character":76},"in_reply_to":"1af94dfe_dd05ccc2","updated":"2016-03-23 14:27:44.000000000","message":"Schema migration -- see https://review.openstack.org/#/c/213835/1/cinder/db/sqlalchemy/migrate_repo/versions/053_per_volume_encryption.py\n\nPerformance concerns could be a deal-breaker. Different databases have different policies on schema changes, and clearly holding a write-lock for the entire table would not work in a production system.","commit_id":"1669ed57301ed746632948156bc9e73237534b17"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"0918406b527a16e93984f006bfb512c1d44ac875","unresolved":false,"context_lines":[{"line_number":111,"context_line":"the \"id\" attribute of the \"encryption\" relation. Similar attributes may be"},{"line_number":112,"context_line":"added to other relations such as snapshots and backups, or the volume_id"},{"line_number":113,"context_line":"attribute of these relations may be used to retrieve the original volume and"},{"line_number":114,"context_line":"encryption metadata (assuming that the volume still exists in the database)."},{"line_number":115,"context_line":""},{"line_number":116,"context_line":"REST API impact"},{"line_number":117,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"fa0719c6_4919b421","line":114,"range":{"start_line":114,"start_character":0,"end_line":114,"end_character":76},"in_reply_to":"fa0719c6_b3472176","updated":"2016-03-23 14:44:24.000000000","message":"As I\u0027ve just posted on that review, that migration is not usable on a live system. \n\nYou need to add the columns empty, and write the cinder code such that it can deal with that, and possibly fill the volumes in as they get touched. You then write a migration tool (maybe a cinder-manage sub-command?) that can copy the data into the volume records in small batches, such that all locks get dropped regularly - maybe a small sleep between batches too.\n\nBasically, you need to be design to be able to run the migration on a live system of e.g. 1 million volumes without locking the table while you update all the volumes. I think that\u0027s quite achievable using the method above - the null field handling (just fall back to reading the type) code can easily be hidden entirely in the cinder object layer, which is one of the purposes of having that layer","commit_id":"1669ed57301ed746632948156bc9e73237534b17"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"2b86216737cd50c86a8d78137b62d36d3da942a6","unresolved":false,"context_lines":[{"line_number":165,"context_line":""},{"line_number":166,"context_line":"If the above scenario remains a cause for concern, one could easily scan to see"},{"line_number":167,"context_line":"if any volumes, snapshots, volume types, etc. are still using the encryption"},{"line_number":168,"context_line":"metdata and remove it when there are no remaining references."},{"line_number":169,"context_line":""},{"line_number":170,"context_line":"Developer impact"},{"line_number":171,"context_line":"----------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"1af94dfe_bdd60025","line":168,"range":{"start_line":168,"start_character":0,"end_line":168,"end_character":61},"updated":"2016-03-16 18:12:14.000000000","message":"Add a command to cinder-manage to do this (or an admin rest API), rather than making everybody write their own version - history has shown it\u0027s easy to write a buggy version yourself","commit_id":"1669ed57301ed746632948156bc9e73237534b17"},{"author":{"_account_id":6802,"name":"Joel Coffman","email":"jmc7tp@gmail.com","username":"joel-coffman"},"change_message_id":"fc6c1bbdabad9f9023b0697448c9bd5b13a21fcc","unresolved":false,"context_lines":[{"line_number":165,"context_line":""},{"line_number":166,"context_line":"If the above scenario remains a cause for concern, one could easily scan to see"},{"line_number":167,"context_line":"if any volumes, snapshots, volume types, etc. are still using the encryption"},{"line_number":168,"context_line":"metdata and remove it when there are no remaining references."},{"line_number":169,"context_line":""},{"line_number":170,"context_line":"Developer impact"},{"line_number":171,"context_line":"----------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"fa0719c6_eedc0e75","line":168,"range":{"start_line":168,"start_character":0,"end_line":168,"end_character":61},"in_reply_to":"1af94dfe_bdd60025","updated":"2016-03-23 14:27:44.000000000","message":"Agree that it would be beneficial to have a standard version, but I personally think that such a capability shouldn\u0027t be required. In practice, there should be very few \"orphaned\" entries.\n\nThe prior paragraphs were added as requested in prior comments to document a potential issue and its resolution (see https://review.openstack.org/#/c/176050/4/specs/liberty/volume-encryption-metadata.rst).","commit_id":"1669ed57301ed746632948156bc9e73237534b17"}]}
