)]}'
{"nova/virt/libvirt/driver.py":[{"author":{"_account_id":8768,"name":"Chris Friesen","email":"chris.friesen@windriver.com","username":"cbf123"},"change_message_id":"33742f4b2026fab5b4a329492bf7285cac3aaff6","unresolved":false,"context_lines":[{"line_number":4895,"context_line":"        # don\u0027t destroy disks in cleanup in case VIF plugging fails."},{"line_number":4896,"context_line":"        destroy_disks \u003d True"},{"line_number":4897,"context_line":"        if (instance.task_state \u003d\u003d task_states.REBUILD_SPAWNING and"},{"line_number":4898,"context_line":"                self.instance_on_disk(instance)):"},{"line_number":4899,"context_line":"            destroy_disks \u003d False"},{"line_number":4900,"context_line":""},{"line_number":4901,"context_line":"        guest \u003d None"}],"source_content_type":"text/x-python","patch_set":1,"id":"7af24918_1ee662e8","line":4898,"range":{"start_line":4898,"start_character":0,"end_line":4898,"end_character":49},"updated":"2016-03-04 20:52:47.000000000","message":"As per IRC, the logic here is a bit iffy.  With Ceph we don\u0027t want to delete the disks because we know they\u0027re shared.  With non-Ceph shared storage it\u0027s tricky because if we happen to be currently rebuilding on the same host as the instance was on previously then it\u0027ll look like it\u0027s shared even if it isn\u0027t.  (Which would then result in leaving the disks around if the rebuild fails.)\n\nHere\u0027s a question...do we ever need to delete the disks when rebuild fails?  If we don\u0027t would the get left around forever or would something eventually clean them up?","commit_id":"420dce3a4685dd1c772a8106ad799d86145c7b12"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"19008726467599ffa16744252d63ac16a92ce08c","unresolved":false,"context_lines":[{"line_number":4895,"context_line":"        # don\u0027t destroy disks in cleanup in case VIF plugging fails."},{"line_number":4896,"context_line":"        destroy_disks \u003d True"},{"line_number":4897,"context_line":"        if (instance.task_state \u003d\u003d task_states.REBUILD_SPAWNING and"},{"line_number":4898,"context_line":"                self.instance_on_disk(instance)):"},{"line_number":4899,"context_line":"            destroy_disks \u003d False"},{"line_number":4900,"context_line":""},{"line_number":4901,"context_line":"        guest \u003d None"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f7c97a3_910b827a","line":4898,"range":{"start_line":4898,"start_character":0,"end_line":4898,"end_character":49},"in_reply_to":"5f7c97a3_5c703541","updated":"2018-07-18 17:06:57.000000000","message":"But we don\u0027t want to limit this check only to evacuate, do we? Otherwise in the case of ceph, we would destroy the disks during a rebuild (same host) failure.\n\nI have the same question as Chris though, it seems like we should never destroy the disks if a rebuild fails -- if rebuild didn\u0027t complete, shouldn\u0027t the disk remain?","commit_id":"420dce3a4685dd1c772a8106ad799d86145c7b12"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"34c52fc77e68cdff8cfdd6e525fc85b37de0a07a","unresolved":false,"context_lines":[{"line_number":4895,"context_line":"        # don\u0027t destroy disks in cleanup in case VIF plugging fails."},{"line_number":4896,"context_line":"        destroy_disks \u003d True"},{"line_number":4897,"context_line":"        if (instance.task_state \u003d\u003d task_states.REBUILD_SPAWNING and"},{"line_number":4898,"context_line":"                self.instance_on_disk(instance)):"},{"line_number":4899,"context_line":"            destroy_disks \u003d False"},{"line_number":4900,"context_line":""},{"line_number":4901,"context_line":"        guest \u003d None"}],"source_content_type":"text/x-python","patch_set":1,"id":"7af24918_b96b0885","line":4898,"range":{"start_line":4898,"start_character":0,"end_line":4898,"end_character":49},"in_reply_to":"7af24918_1ee662e8","updated":"2016-03-18 20:13:45.000000000","message":"The _run_pending_deletes periodic task in the compute manager might clean them up eventually after the instance is deleted.","commit_id":"420dce3a4685dd1c772a8106ad799d86145c7b12"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"731e7d75bbb0794955e3d633bf792d9ec77c700f","unresolved":false,"context_lines":[{"line_number":4895,"context_line":"        # don\u0027t destroy disks in cleanup in case VIF plugging fails."},{"line_number":4896,"context_line":"        destroy_disks \u003d True"},{"line_number":4897,"context_line":"        if (instance.task_state \u003d\u003d task_states.REBUILD_SPAWNING and"},{"line_number":4898,"context_line":"                self.instance_on_disk(instance)):"},{"line_number":4899,"context_line":"            destroy_disks \u003d False"},{"line_number":4900,"context_line":""},{"line_number":4901,"context_line":"        guest \u003d None"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f7c97a3_5c703541","line":4898,"range":{"start_line":4898,"start_character":0,"end_line":4898,"end_character":49},"in_reply_to":"7af24918_b96b0885","updated":"2018-06-28 15:52:39.000000000","message":"\u003e if we happen to be currently rebuilding on the same host as the instance was on previously then it\u0027ll look like it\u0027s shared even if it isn\u0027t.\n\nFor evacuate we\u0027re rebuilding on a new host. Only rebuild (not evacuate) rebuilds on the same host. But it\u0027s true that in this code we aren\u0027t going to know the difference unless we look for something like an active evacuate migration record.","commit_id":"420dce3a4685dd1c772a8106ad799d86145c7b12"}]}