)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":13157,"name":"Zhao Chao","email":"zhaochao1984@gmail.com","username":"zhaochao"},"change_message_id":"4a4b04a8575137571206cd95fbc234ba5e7807db","unresolved":false,"context_lines":[{"line_number":6,"context_line":""},{"line_number":7,"context_line":"Support restore backup to volumes simultaneously"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"This spec proposes supporting restore backup to"},{"line_number":10,"context_line":"volumes simultaneously."},{"line_number":11,"context_line":""},{"line_number":12,"context_line":"Change-Id: I7467e9526f9be54ae79d6a3306d4a08bf2ec6cc4"},{"line_number":13,"context_line":"Partial-Implements: bp support-simultaneously-restore-backups"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":9,"id":"9f91af0f_0836a753","line":10,"range":{"start_line":9,"start_character":30,"end_line":10,"end_character":7},"updated":"2018-01-05 03:19:51.000000000","message":"This should be \"restore one backup to different volumes\" just like the summary line of the bp, otherwise it\u0027s ambiguous, as we are able to restore different backups to different volumes simultaneously.","commit_id":"69fd60187272e2ed14a17864048f53740043e7c0"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"f1b5f6d39662e09e1b366de6556cd3e5ca763c34","unresolved":false,"context_lines":[{"line_number":6,"context_line":""},{"line_number":7,"context_line":"Support restore backup to volumes simultaneously"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"This spec proposes supporting restore backup to"},{"line_number":10,"context_line":"volumes simultaneously."},{"line_number":11,"context_line":""},{"line_number":12,"context_line":"Change-Id: I7467e9526f9be54ae79d6a3306d4a08bf2ec6cc4"},{"line_number":13,"context_line":"Partial-Implements: bp support-simultaneously-restore-backups"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":9,"id":"5f93b717_26ddee41","line":10,"range":{"start_line":9,"start_character":30,"end_line":10,"end_character":7},"in_reply_to":"9f91af0f_0836a753","updated":"2018-01-27 02:40:19.000000000","message":"Done","commit_id":"69fd60187272e2ed14a17864048f53740043e7c0"}],"specs/queens/support-restore-backup-simultaneously.rst":[{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"6e0b8bc410b2feae92cdc948a8edb6ed12293c34","unresolved":false,"context_lines":[{"line_number":37,"context_line":"being restoring to. To support multiple requests, the mentionable changes"},{"line_number":38,"context_line":"are below:"},{"line_number":39,"context_line":""},{"line_number":40,"context_line":"1. New foreign key ``restore_backup_id`` will be added in volume table ("},{"line_number":41,"context_line":"   back reference ``volumes`` in backup table). Once cinder starts or ends"},{"line_number":42,"context_line":"   restoring action, this attribute will be set or unset."},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"2. When service starts, the backup cleanup operation will use backup\u0027s"},{"line_number":45,"context_line":"   ``volumes`` attribute to find out unfinished tasks, update the"}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_c746d250","line":42,"range":{"start_line":40,"start_character":0,"end_line":42,"end_character":57},"updated":"2017-12-01 15:06:18.000000000","message":"It is probably easier to make this a separate table - we\u0027ve plenty enough in the volumes table already, and for most operations the new field is not needed, so the cost of the extra joins is low. It also means there\u0027s no expensive \u0027alter table\u0027 statement of the (potentially very large) volumes table during migration, which always helps migrations go smoother.","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b52e17e37b22143f4c02e657bb4f6495b972b4ff","unresolved":false,"context_lines":[{"line_number":37,"context_line":"being restoring to. To support multiple requests, the mentionable changes"},{"line_number":38,"context_line":"are below:"},{"line_number":39,"context_line":""},{"line_number":40,"context_line":"1. New foreign key ``restore_backup_id`` will be added in volume table ("},{"line_number":41,"context_line":"   back reference ``volumes`` in backup table). Once cinder starts or ends"},{"line_number":42,"context_line":"   restoring action, this attribute will be set or unset."},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"2. When service starts, the backup cleanup operation will use backup\u0027s"},{"line_number":45,"context_line":"   ``volumes`` attribute to find out unfinished tasks, update the"}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_7d5f3b41","line":42,"range":{"start_line":40,"start_character":0,"end_line":42,"end_character":57},"in_reply_to":"df87a7cf_c746d250","updated":"2017-12-04 03:02:38.000000000","message":"Thanks for your suggestion, will propose to add the relationship in a new table in next patch.","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"6e0b8bc410b2feae92cdc948a8edb6ed12293c34","unresolved":false,"context_lines":[{"line_number":41,"context_line":"   back reference ``volumes`` in backup table). Once cinder starts or ends"},{"line_number":42,"context_line":"   restoring action, this attribute will be set or unset."},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"2. When service starts, the backup cleanup operation will use backup\u0027s"},{"line_number":45,"context_line":"   ``volumes`` attribute to find out unfinished tasks, update the"},{"line_number":46,"context_line":"   corresponding volume\u0027s status and unset the ``restore_backup_id``"},{"line_number":47,"context_line":"   attribute."},{"line_number":48,"context_line":""},{"line_number":49,"context_line":"3. ``restore_volume_id`` will not be used in the new process and will be"},{"line_number":50,"context_line":"   dropped in the next cycle."}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_87644abb","line":47,"range":{"start_line":44,"start_character":3,"end_line":47,"end_character":13},"updated":"2017-12-01 15:06:18.000000000","message":"There may (and in any sensible system, will) be several cinder-backup processes running, potentially spread across several nodes.\n\nHow does the newly started process differentiate between a backup that is still in-progress by another process, and one that is in need of cleanup?","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b52e17e37b22143f4c02e657bb4f6495b972b4ff","unresolved":false,"context_lines":[{"line_number":41,"context_line":"   back reference ``volumes`` in backup table). Once cinder starts or ends"},{"line_number":42,"context_line":"   restoring action, this attribute will be set or unset."},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"2. When service starts, the backup cleanup operation will use backup\u0027s"},{"line_number":45,"context_line":"   ``volumes`` attribute to find out unfinished tasks, update the"},{"line_number":46,"context_line":"   corresponding volume\u0027s status and unset the ``restore_backup_id``"},{"line_number":47,"context_line":"   attribute."},{"line_number":48,"context_line":""},{"line_number":49,"context_line":"3. ``restore_volume_id`` will not be used in the new process and will be"},{"line_number":50,"context_line":"   dropped in the next cycle."}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_8012d457","line":47,"range":{"start_line":44,"start_character":3,"end_line":47,"end_character":13},"in_reply_to":"df87a7cf_87644abb","updated":"2017-12-04 03:02:38.000000000","message":"The clean up operation is proceed during host initialization and the host will not accept any new requests during this period. Your concern will not happen as long as the service only fetches the backups on that host and api denies the request when the corresponding service is not ready.","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"6e0b8bc410b2feae92cdc948a8edb6ed12293c34","unresolved":false,"context_lines":[{"line_number":46,"context_line":"   corresponding volume\u0027s status and unset the ``restore_backup_id``"},{"line_number":47,"context_line":"   attribute."},{"line_number":48,"context_line":""},{"line_number":49,"context_line":"3. ``restore_volume_id`` will not be used in the new process and will be"},{"line_number":50,"context_line":"   dropped in the next cycle."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"4. When new restoring request comes, only the backup that in the"},{"line_number":53,"context_line":"   \u0027available\u0027 status can be used as before, and the difference is now we"}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_c717b2ed","line":50,"range":{"start_line":49,"start_character":0,"end_line":50,"end_character":29},"updated":"2017-12-01 15:06:18.000000000","message":"It will have to be checked in the code still though, since during an upgrade you can potentially have both old and new cinder-backup processes working at the same time, and they mustn\u0027t step on each other, so it might take two releases to remove the field - one to check it in case an old c-bak process is using it, then one to remove it.","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b52e17e37b22143f4c02e657bb4f6495b972b4ff","unresolved":false,"context_lines":[{"line_number":46,"context_line":"   corresponding volume\u0027s status and unset the ``restore_backup_id``"},{"line_number":47,"context_line":"   attribute."},{"line_number":48,"context_line":""},{"line_number":49,"context_line":"3. ``restore_volume_id`` will not be used in the new process and will be"},{"line_number":50,"context_line":"   dropped in the next cycle."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"4. When new restoring request comes, only the backup that in the"},{"line_number":53,"context_line":"   \u0027available\u0027 status can be used as before, and the difference is now we"}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_e074009c","line":50,"range":{"start_line":49,"start_character":0,"end_line":50,"end_character":29},"in_reply_to":"df87a7cf_c717b2ed","updated":"2017-12-04 03:02:38.000000000","message":"If this feature can be merged in this cycle, the ``restore_volume_id`` will be kept and can be dropped until next cycle finished, and this attribute will still be used for clean up operations.","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"6e0b8bc410b2feae92cdc948a8edb6ed12293c34","unresolved":false,"context_lines":[{"line_number":49,"context_line":"3. ``restore_volume_id`` will not be used in the new process and will be"},{"line_number":50,"context_line":"   dropped in the next cycle."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"4. When new restoring request comes, only the backup that in the"},{"line_number":53,"context_line":"   \u0027available\u0027 status can be used as before, and the difference is now we"},{"line_number":54,"context_line":"   will not update the backup with \u0027restoring\u0027 status."},{"line_number":55,"context_line":""},{"line_number":56,"context_line":"5. When delete request comes, cinder will both check the backup\u0027s status"},{"line_number":57,"context_line":"   and whether its ``volumes`` list is empty."}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_07911a53","line":54,"range":{"start_line":52,"start_character":0,"end_line":54,"end_character":54},"updated":"2017-12-01 15:06:18.000000000","message":"That status change is user visible and is used by some scripts to see if the restore has finished yet.\n\nIt also internally serves as a lock against the backup being deleted while a restore is still happening.\n\nI\u0027d suggest that the best new behaviour we can manage is to keep the status as \u0027restoring\u0027 while any restore operations are ongoing.","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b52e17e37b22143f4c02e657bb4f6495b972b4ff","unresolved":false,"context_lines":[{"line_number":49,"context_line":"3. ``restore_volume_id`` will not be used in the new process and will be"},{"line_number":50,"context_line":"   dropped in the next cycle."},{"line_number":51,"context_line":""},{"line_number":52,"context_line":"4. When new restoring request comes, only the backup that in the"},{"line_number":53,"context_line":"   \u0027available\u0027 status can be used as before, and the difference is now we"},{"line_number":54,"context_line":"   will not update the backup with \u0027restoring\u0027 status."},{"line_number":55,"context_line":""},{"line_number":56,"context_line":"5. When delete request comes, cinder will both check the backup\u0027s status"},{"line_number":57,"context_line":"   and whether its ``volumes`` list is empty."}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_e0a960df","line":54,"range":{"start_line":52,"start_character":0,"end_line":54,"end_character":54},"in_reply_to":"df87a7cf_07911a53","updated":"2017-12-04 03:02:38.000000000","message":"That\u0027s is true. We even have internal codes which depends on backup\u0027s status, but we still have the chance to know whether the backup operation finished, as we aslo have status change in volume objects. and what\u0027s more we still could not use backup\u0027s \u0027restoring\u0027 status after we support multi restore actions. because the backup\u0027s restoring status is not depend on a single restoring action anymore.","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"6e0b8bc410b2feae92cdc948a8edb6ed12293c34","unresolved":false,"context_lines":[{"line_number":66,"context_line":"------------"},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"Keep ``restoring`` status when performing restore action is one of"},{"line_number":69,"context_line":"the alternative solutions. And the drawback of this is it will complicate"},{"line_number":70,"context_line":"the status synchronization logic and we have to fix it manually once it"},{"line_number":71,"context_line":"fails into incorrect status."},{"line_number":72,"context_line":""},{"line_number":73,"context_line":""},{"line_number":74,"context_line":"Data model impact"}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_c7a81279","line":71,"range":{"start_line":69,"start_character":28,"end_line":71,"end_character":28},"updated":"2017-12-01 15:06:18.000000000","message":"Not sure I follow why this is particularly complicated?","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"b52e17e37b22143f4c02e657bb4f6495b972b4ff","unresolved":false,"context_lines":[{"line_number":66,"context_line":"------------"},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"Keep ``restoring`` status when performing restore action is one of"},{"line_number":69,"context_line":"the alternative solutions. And the drawback of this is it will complicate"},{"line_number":70,"context_line":"the status synchronization logic and we have to fix it manually once it"},{"line_number":71,"context_line":"fails into incorrect status."},{"line_number":72,"context_line":""},{"line_number":73,"context_line":""},{"line_number":74,"context_line":"Data model impact"}],"source_content_type":"text/x-rst","patch_set":3,"id":"df87a7cf_a0703874","line":71,"range":{"start_line":69,"start_character":28,"end_line":71,"end_character":28},"in_reply_to":"df87a7cf_c7a81279","updated":"2017-12-04 03:02:38.000000000","message":"oh, not that complicated, but only keep \u0027available\u0027 will make it much simplier and we don\u0027t need to add a status sync logic as we did such as quota sync. And ASAK. there can be no status change when object is used as a data resource (ready only), we don\u0027t have status change when creating volume from images or snapshots. AWS and Google don\u0027t have   \nstatus change when performing restoring or creating volume from backup(snapshot) either.","commit_id":"60a2d41a924c88b2d713aaa7240199dde1daf17f"}],"specs/rocky/support-restore-backup-simultaneously.rst":[{"author":{"_account_id":2243,"name":"John Griffith","email":"john.griffith8@gmail.com","username":"john-griffith"},"change_message_id":"8b6edbc4f5de0eb33cf51c2ac3d80e46106244b2","unresolved":false,"context_lines":[{"line_number":18,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"Now when cinder restoring backup, other restoring requests on this backup will"},{"line_number":21,"context_line":"not be allowed until the task finished. This restriction is both inefficiency"},{"line_number":22,"context_line":"and unnecessary. Also we support create volume from image, snapshot, volume"},{"line_number":23,"context_line":"and backup, and backup is the only source which can not support creating"},{"line_number":24,"context_line":"multiple volumes with same source at the same time."}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_a86a4d7d","line":21,"range":{"start_line":21,"start_character":65,"end_line":21,"end_character":77},"updated":"2017-12-22 16:27:41.000000000","message":"inefficient","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":2243,"name":"John Griffith","email":"john.griffith8@gmail.com","username":"john-griffith"},"change_message_id":"8b6edbc4f5de0eb33cf51c2ac3d80e46106244b2","unresolved":false,"context_lines":[{"line_number":21,"context_line":"not be allowed until the task finished. This restriction is both inefficiency"},{"line_number":22,"context_line":"and unnecessary. Also we support create volume from image, snapshot, volume"},{"line_number":23,"context_line":"and backup, and backup is the only source which can not support creating"},{"line_number":24,"context_line":"multiple volumes with same source at the same time."},{"line_number":25,"context_line":""},{"line_number":26,"context_line":"Use Cases"},{"line_number":27,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_08281932","line":24,"range":{"start_line":24,"start_character":0,"end_line":24,"end_character":50},"updated":"2017-12-22 16:27:41.000000000","message":"I believe there is some reasoning here (or was originally) that had to do with backup and restore in some cases being rather resource intensive.  That may not be a concern any longer but might be worth mentioning here and pointing out that it\u0027s a consideration.\n\nNote that in the case of images, without cache (and even with cache) it actually is problematic to try and run multiple create-from-image calls simultaneously with most backend devices.","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":2243,"name":"John Griffith","email":"john.griffith8@gmail.com","username":"john-griffith"},"change_message_id":"8b6edbc4f5de0eb33cf51c2ac3d80e46106244b2","unresolved":false,"context_lines":[{"line_number":28,"context_line":""},{"line_number":29,"context_line":"Add the ability to restore (and create) multiple volumes with one specific"},{"line_number":30,"context_line":"backup simultaneously."},{"line_number":31,"context_line":""},{"line_number":32,"context_line":"Proposed change"},{"line_number":33,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":34,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_08dd7941","line":31,"updated":"2017-12-22 16:27:41.000000000","message":"can you provide an end-user use case where this might be applied?  I\u0027m looking more for a \"why\" they would need/want this capability in their workflow.","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"1d2352b55ebfe87dd3a2101e65a585da81c22b22","unresolved":false,"context_lines":[{"line_number":28,"context_line":""},{"line_number":29,"context_line":"Add the ability to restore (and create) multiple volumes with one specific"},{"line_number":30,"context_line":"backup simultaneously."},{"line_number":31,"context_line":""},{"line_number":32,"context_line":"Proposed change"},{"line_number":33,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":34,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_737658a7","line":31,"in_reply_to":"bf8cb3f7_08dd7941","updated":"2017-12-26 08:52:16.000000000","message":"done!","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":2243,"name":"John Griffith","email":"john.griffith8@gmail.com","username":"john-griffith"},"change_message_id":"8b6edbc4f5de0eb33cf51c2ac3d80e46106244b2","unresolved":false,"context_lines":[{"line_number":40,"context_line":"being restoring to. To support multiple requests, the mentionable changes"},{"line_number":41,"context_line":"are below:"},{"line_number":42,"context_line":""},{"line_number":43,"context_line":"1. New table ``backup_operations`` with two foreign keys ``volume_id``"},{"line_number":44,"context_line":"   and ``backup_id`` will be introduced, please read data model impact"},{"line_number":45,"context_line":"   section for more detail. Once Cinder starts or ends restoring action,"},{"line_number":46,"context_line":"   new record will be correspondingly added or deleted."},{"line_number":47,"context_line":""},{"line_number":48,"context_line":"2. When service starts, the backup cleanup operation will both use"},{"line_number":49,"context_line":"   ``backup_operations`` table and existing ``restore_volume_id`` attribute"}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_c8947147","line":46,"range":{"start_line":43,"start_character":3,"end_line":46,"end_character":55},"updated":"2017-12-22 16:27:41.000000000","message":"So we\u0027re introducing the idea of using a table to represent a running process?  Looking at the data model notes below it seems you\u0027re only using this to build a relationship between backup-id and volume-id for the period a backup is active; is that correct?\n\nIf that\u0027s the case, would it perhaps be more efficient to rework the backup table and it\u0027s status?  I\u0027m having difficulty understanding the need for the new table, other than treating backup processes independent, but it seems the backup table that already is in place could be used to do that no?","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"1d2352b55ebfe87dd3a2101e65a585da81c22b22","unresolved":false,"context_lines":[{"line_number":40,"context_line":"being restoring to. To support multiple requests, the mentionable changes"},{"line_number":41,"context_line":"are below:"},{"line_number":42,"context_line":""},{"line_number":43,"context_line":"1. New table ``backup_operations`` with two foreign keys ``volume_id``"},{"line_number":44,"context_line":"   and ``backup_id`` will be introduced, please read data model impact"},{"line_number":45,"context_line":"   section for more detail. Once Cinder starts or ends restoring action,"},{"line_number":46,"context_line":"   new record will be correspondingly added or deleted."},{"line_number":47,"context_line":""},{"line_number":48,"context_line":"2. When service starts, the backup cleanup operation will both use"},{"line_number":49,"context_line":"   ``backup_operations`` table and existing ``restore_volume_id`` attribute"}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_93ec3c2a","line":46,"range":{"start_line":43,"start_character":3,"end_line":46,"end_character":55},"in_reply_to":"bf8cb3f7_c8947147","updated":"2017-12-26 08:52:16.000000000","message":"Yes, I had thought about this, and my first propose is to add a foreign key and relationship in volume table within which we can get all of the volumes that are been used for one specific backup resource via backup.restore_volumes see [1]. But Duncan worried about the performance impact. I think the benefits of using new table is it\u0027s easy to expand and we can record more detail of the restoring actions other than one simple volume id.\n\n[1]: https://review.openstack.org/#/c/524441/3..7/specs/queens/support-restore-backup-simultaneously.rst line 40","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":11904,"name":"Sean McGinnis","email":"sean.mcginnis@gmail.com","username":"SeanM"},"change_message_id":"941ba01e6fbfffe3d8cb1c225d6e32bf83ad2881","unresolved":false,"context_lines":[{"line_number":129,"context_line":"There would be a performance impact on single restoring operation for"},{"line_number":130,"context_line":"some of the backup driver when multiple requests are handled."},{"line_number":131,"context_line":""},{"line_number":132,"context_line":"Discussion: Maybe we should add a config option to turn on/off this"},{"line_number":133,"context_line":"feature or set a maximal number on how many volumes that a single backup"},{"line_number":134,"context_line":"can support restoring."},{"line_number":135,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_54bebe65","line":132,"updated":"2017-12-22 12:10:01.000000000","message":"My initial thought is yes, we should protect against overloading things. But I also wonder if we would be able to pick a limit value that makes sense for every deployment as a default. I would think the number of times this would be used would be relatively small, and I would hope no one would try to restore the same backup to 20 volumes simultaneously.\n\nMaybe another alternative is we build in some kind of system aware throttling.","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"1d2352b55ebfe87dd3a2101e65a585da81c22b22","unresolved":false,"context_lines":[{"line_number":129,"context_line":"There would be a performance impact on single restoring operation for"},{"line_number":130,"context_line":"some of the backup driver when multiple requests are handled."},{"line_number":131,"context_line":""},{"line_number":132,"context_line":"Discussion: Maybe we should add a config option to turn on/off this"},{"line_number":133,"context_line":"feature or set a maximal number on how many volumes that a single backup"},{"line_number":134,"context_line":"can support restoring."},{"line_number":135,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_f3dca875","line":132,"in_reply_to":"bf8cb3f7_54bebe65","updated":"2017-12-26 08:52:16.000000000","message":"Ok, I would like to put this into proposed changes section to get more feedback.","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":2243,"name":"John Griffith","email":"john.griffith8@gmail.com","username":"john-griffith"},"change_message_id":"8b6edbc4f5de0eb33cf51c2ac3d80e46106244b2","unresolved":false,"context_lines":[{"line_number":132,"context_line":"Discussion: Maybe we should add a config option to turn on/off this"},{"line_number":133,"context_line":"feature or set a maximal number on how many volumes that a single backup"},{"line_number":134,"context_line":"can support restoring."},{"line_number":135,"context_line":""},{"line_number":136,"context_line":"Other deployer impact"},{"line_number":137,"context_line":"---------------------"},{"line_number":138,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_88ac298d","line":135,"updated":"2017-12-22 16:27:41.000000000","message":"We should clarify here, this isn\u0027t an option like \"cinder backup-restore \u003cbackup-id\u003e [v-1,v-2,v-3...]\" but rather the ability to do things like:\n  \"cinder backup-restore \u003cbackup-id\u003e v-1\"\n  # wait some time while backup operation is in progress\n  \"cinder backup-restore \u003cbackup-id\u003e v-2\"\n  # repeat...\n\nA parallel restore to multiple volumes simultaneously would be easier and less concerning in terms or resource consumption.  I\u0027m concerned what this proposal will require from a node but maybe that\u0027s unwarranted?","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":23083,"name":"TommyLike","email":"tommylikehu@gmail.com","username":"TommyLike"},"change_message_id":"1d2352b55ebfe87dd3a2101e65a585da81c22b22","unresolved":false,"context_lines":[{"line_number":132,"context_line":"Discussion: Maybe we should add a config option to turn on/off this"},{"line_number":133,"context_line":"feature or set a maximal number on how many volumes that a single backup"},{"line_number":134,"context_line":"can support restoring."},{"line_number":135,"context_line":""},{"line_number":136,"context_line":"Other deployer impact"},{"line_number":137,"context_line":"---------------------"},{"line_number":138,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"bf8cb3f7_50d46abd","line":135,"in_reply_to":"bf8cb3f7_88ac298d","updated":"2017-12-26 08:52:16.000000000","message":"Thanks for your reminder:) I will add this in the other end user impact secion.","commit_id":"0bc4a49be0796c80441f53366e27cae32c3558c8"},{"author":{"_account_id":9535,"name":"Gorka Eguileor","email":"geguileo@redhat.com","username":"Gorka"},"change_message_id":"2fcdc285786e95e253cdda51407df2bb54419011","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":"In some clouds, backup is one of the resources used for quick deploy data"},{"line_number":29,"context_line":"volumes across available zones. Add the ability to restore multiple volumes"},{"line_number":30,"context_line":"with one specific backup simultaneously will boost the administrator\u0027s"},{"line_number":31,"context_line":"deployment process."},{"line_number":32,"context_line":""}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f93b717_4833ce62","line":29,"range":{"start_line":29,"start_character":15,"end_line":29,"end_character":24},"updated":"2018-02-02 11:29:16.000000000","message":"nit: availability","commit_id":"191261277e6fd282e5cb0f01aac2b5c0daf96a12"}]}
