)]}'
{"nova/tests/functional/test_servers.py":[{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"0cbde4883b4455d95c0fdabf19d3ef4027beb08e","unresolved":false,"context_lines":[{"line_number":1577,"context_line":"        dest_allocation \u003d allocations[dest_rp_uuid][\u0027resources\u0027]"},{"line_number":1578,"context_line":"        self.assertFlavorMatchesAllocation(self.flavor1, dest_allocation)"},{"line_number":1579,"context_line":""},{"line_number":1580,"context_line":"        self._run_periodics()"},{"line_number":1581,"context_line":""},{"line_number":1582,"context_line":"        source_usages \u003d self._get_provider_usages(source_rp_uuid)"},{"line_number":1583,"context_line":"        self.assertFlavorMatchesAllocation(self.flavor1, source_usages)"}],"source_content_type":"text/x-python","patch_set":1,"id":"9f436f4f_b32eeea2","line":1580,"updated":"2017-08-10 14:45:25.000000000","message":"I think this doesn\u0027t do anything now because of:\n\nhttps://review.openstack.org/#/c/491850/\n\n^ will skip removing the allocation for the instance if it\u0027s no longer tracked on that compute node.\n\nor\n\nhttps://review.openstack.org/#/c/491012/\n\n^ will skip overwriting allocations tracked on the dest node if everything is now running pike code","commit_id":"e3899d2615a73e8178a69a03e7c3f599f001054e"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"0cbde4883b4455d95c0fdabf19d3ef4027beb08e","unresolved":false,"context_lines":[{"line_number":1599,"context_line":""},{"line_number":1600,"context_line":"        self._run_periodics()"},{"line_number":1601,"context_line":""},{"line_number":1602,"context_line":"        # this looks like a bug  as the source compute should delete the claim"},{"line_number":1603,"context_line":"        # after it detects that the instance is evacuated away but it doesn\u0027t"},{"line_number":1604,"context_line":"        source_usages \u003d self._get_provider_usages(source_rp_uuid)"},{"line_number":1605,"context_line":"        self.assertFlavorMatchesAllocation(self.flavor1, source_usages)"},{"line_number":1606,"context_line":""}],"source_content_type":"text/x-python","patch_set":1,"id":"9f436f4f_6e4d3d71","line":1603,"range":{"start_line":1602,"start_character":8,"end_line":1603,"end_character":77},"updated":"2017-08-10 14:45:25.000000000","message":"yeah see above\n\nMaybe https://review.openstack.org/#/c/491850/ needs to take into account self.tracked_migrations? Although that might be empty when we just started up the service again.\n\nMaybe we\u0027d have to do some cleanup on init_host when we process evacuations:\n\nhttps://github.com/openstack/nova/blob/9a66d039a14afd591f4a3b6e655580aeeed17d29/nova/compute/manager.py#L649","commit_id":"e3899d2615a73e8178a69a03e7c3f599f001054e"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"32477cb07aa29364999483512d5581db345c4866","unresolved":false,"context_lines":[{"line_number":1599,"context_line":""},{"line_number":1600,"context_line":"        self._run_periodics()"},{"line_number":1601,"context_line":""},{"line_number":1602,"context_line":"        # this looks like a bug  as the source compute should delete the claim"},{"line_number":1603,"context_line":"        # after it detects that the instance is evacuated away but it doesn\u0027t"},{"line_number":1604,"context_line":"        source_usages \u003d self._get_provider_usages(source_rp_uuid)"},{"line_number":1605,"context_line":"        self.assertFlavorMatchesAllocation(self.flavor1, source_usages)"},{"line_number":1606,"context_line":""}],"source_content_type":"text/x-python","patch_set":1,"id":"9f436f4f_6e841d5f","line":1603,"range":{"start_line":1602,"start_character":8,"end_line":1603,"end_character":77},"in_reply_to":"9f436f4f_6e4d3d71","updated":"2017-08-10 14:49:02.000000000","message":"We could potentially lump this in with bug 1679750 where we\u0027d need to cleanup allocations for a compute when it starts back up and we\u0027re processing evacuated instances.\n\nWe had talked about fixing bug 1679750 by removing the allocations when deleting the instance in the API during the \u0027local delete\u0027 case, but this is a bit different in that the instance isn\u0027t deleted, it\u0027s evacuated, so we\u0027d need to cleanup the stale allocations on the source compute on startup.","commit_id":"e3899d2615a73e8178a69a03e7c3f599f001054e"}]}