)]}'
{"nova/compute/api.py":[{"author":{"_account_id":10224,"name":"Feodor Tersin","email":"ftersin@hotmail.com","username":"ftersin"},"change_message_id":"6414ab1d116da380b66d8fa2128db33a5a5e9212","unresolved":false,"context_lines":[{"line_number":718,"context_line":"        \"\"\"Override any block devices from the image by device name\"\"\""},{"line_number":719,"context_line":"        device_names \u003d set(bdm[\u0027device_name\u0027] for bdm in block_device_mapping"},{"line_number":720,"context_line":"                           if bdm[\u0027device_name\u0027])"},{"line_number":721,"context_line":"        return (block_device_mapping +"},{"line_number":722,"context_line":"                [bdm for bdm in image_mappings"},{"line_number":723,"context_line":"                 if bdm[\u0027device_name\u0027] not in device_names])"},{"line_number":724,"context_line":""}],"source_content_type":"text/x-python","patch_set":2,"id":"fa32b979_d47ed5ad","line":721,"updated":"2015-06-15 13:25:04.000000000","message":"AWS can merge each bdm record instead of replace it:\n\n image: /dev/vdb, snapshot_id\u003d1, size\u003d10\n args: /dev/vdb, size\u003d100\n result: /dev/vdb, snapshot_id\u003d1, size\u003d100\n\nCould we have the same in Nova? (Certainly with the change you proposed we can emulate AWS behavior in ec2api side by parse image bdm and merge it with user bdm, but i think this is not good for long-term code support and in any case this feature is useful for Nova users as well)","commit_id":"c71e75ce7c6742ce9e34addd17c210b87ac5f9d0"},{"author":{"_account_id":5511,"name":"Nikola Dipanov","email":"ndipanov@redhat.com","username":"ndipanov"},"change_message_id":"4d2f80b1a18cf17f2e05a53f5cfeb8a72b0ac6e0","unresolved":false,"context_lines":[{"line_number":718,"context_line":"        \"\"\"Override any block devices from the image by device name\"\"\""},{"line_number":719,"context_line":"        device_names \u003d set(bdm[\u0027device_name\u0027] for bdm in block_device_mapping"},{"line_number":720,"context_line":"                           if bdm[\u0027device_name\u0027])"},{"line_number":721,"context_line":"        return (block_device_mapping +"},{"line_number":722,"context_line":"                [bdm for bdm in image_mappings"},{"line_number":723,"context_line":"                 if bdm[\u0027device_name\u0027] not in device_names])"},{"line_number":724,"context_line":""}],"source_content_type":"text/x-python","patch_set":2,"id":"fa32b979_8a5c087e","line":721,"in_reply_to":"fa32b979_d47ed5ad","updated":"2015-06-18 08:46:04.000000000","message":"Well as I pointed out on the commit message - this is how nova used to work before I brok^H^H^H^H changed it :) so it seemed like the right thing to do in this case.","commit_id":"c71e75ce7c6742ce9e34addd17c210b87ac5f9d0"},{"author":{"_account_id":10224,"name":"Feodor Tersin","email":"ftersin@hotmail.com","username":"ftersin"},"change_message_id":"6414ab1d116da380b66d8fa2128db33a5a5e9212","unresolved":false,"context_lines":[{"line_number":720,"context_line":"                           if bdm[\u0027device_name\u0027])"},{"line_number":721,"context_line":"        return (block_device_mapping +"},{"line_number":722,"context_line":"                [bdm for bdm in image_mappings"},{"line_number":723,"context_line":"                 if bdm[\u0027device_name\u0027] not in device_names])"},{"line_number":724,"context_line":""},{"line_number":725,"context_line":"    def _check_and_transform_bdm(self, context, base_options, instance_type,"},{"line_number":726,"context_line":"                                 image_meta, min_count, max_count,"}],"source_content_type":"text/x-python","patch_set":2,"id":"fa32b979_ccd3a937","line":723,"updated":"2015-06-15 13:25:04.000000000","message":"Also probably it\u0027s good not to return bdms with no_device flag is on. But it doesn\u0027t directly related to this review and could be done separately.","commit_id":"c71e75ce7c6742ce9e34addd17c210b87ac5f9d0"},{"author":{"_account_id":5511,"name":"Nikola Dipanov","email":"ndipanov@redhat.com","username":"ndipanov"},"change_message_id":"4d2f80b1a18cf17f2e05a53f5cfeb8a72b0ac6e0","unresolved":false,"context_lines":[{"line_number":720,"context_line":"                           if bdm[\u0027device_name\u0027])"},{"line_number":721,"context_line":"        return (block_device_mapping +"},{"line_number":722,"context_line":"                [bdm for bdm in image_mappings"},{"line_number":723,"context_line":"                 if bdm[\u0027device_name\u0027] not in device_names])"},{"line_number":724,"context_line":""},{"line_number":725,"context_line":"    def _check_and_transform_bdm(self, context, base_options, instance_type,"},{"line_number":726,"context_line":"                                 image_meta, min_count, max_count,"}],"source_content_type":"text/x-python","patch_set":2,"id":"fa32b979_0a49d8b6","line":723,"in_reply_to":"fa32b979_ccd3a937","updated":"2015-06-18 08:46:04.000000000","message":"Right - probably useless but all other block device code ignores it so it can\u0027t really hurt apart from some unnecessary db entries.","commit_id":"c71e75ce7c6742ce9e34addd17c210b87ac5f9d0"}]}