)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"0fd7bcf75bb08b48ced53d9446bffaeac0fc0c0a","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"8c957bb9_a8bd2234","updated":"2022-02-19 20:37:02.000000000","message":"recheck","commit_id":"7cce929e6fc9d1f02464f0d6032b354d5953bbb2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"02882e6c833a9754c42878911579dbe0a76b0c29","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"fbc46726_281a3020","updated":"2022-02-20 10:05:15.000000000","message":"recheck","commit_id":"7cce929e6fc9d1f02464f0d6032b354d5953bbb2"}],"nova/objects/compute_node.py":[{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"873153cb1a7f0808720936dda6e27c41ad72b7e0","unresolved":true,"context_lines":[{"line_number":147,"context_line":"        else:"},{"line_number":148,"context_line":"            super(ComputeNode, self).obj_make_compatible("},{"line_number":149,"context_line":"                primitive, str_target_version)"},{"line_number":150,"context_line":""},{"line_number":151,"context_line":"        if target_version \u003c (1, 17):"},{"line_number":152,"context_line":"            if \u0027mapped\u0027 in primitive:"},{"line_number":153,"context_line":"                del primitive[\u0027mapped\u0027]"}],"source_content_type":"text/x-python","patch_set":2,"id":"09904c81_e1594630","line":150,"updated":"2022-02-21 09:50:28.000000000","message":"So this whole thing is ugly like hell. I don\u0027t see an easy fix from the OVO side either but maybe others can. \n\nAnyhow a totally different alternative would be not to implement this backport at all here. But instead make a minimum compute service level check in the new compute before publishing HVSpec objects with the new enum values. That way we can avoid ever triggering the need of this backport.","commit_id":"7cce929e6fc9d1f02464f0d6032b354d5953bbb2"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"001841d96964bcebcadb154417cb7515607d2092","unresolved":true,"context_lines":[{"line_number":147,"context_line":"        else:"},{"line_number":148,"context_line":"            super(ComputeNode, self).obj_make_compatible("},{"line_number":149,"context_line":"                primitive, str_target_version)"},{"line_number":150,"context_line":""},{"line_number":151,"context_line":"        if target_version \u003c (1, 17):"},{"line_number":152,"context_line":"            if \u0027mapped\u0027 in primitive:"},{"line_number":153,"context_line":"                del primitive[\u0027mapped\u0027]"}],"source_content_type":"text/x-python","patch_set":2,"id":"3369c1ee_2f411645","line":150,"in_reply_to":"09904c81_e1594630","updated":"2022-02-21 17:13:57.000000000","message":"Does making the thing a ListObject help? Then you get the opportunity to do the backports as a context-related ListObject of objects, which would not run that generic \"if it\u0027s a list, assume backports of each\" code in o.vo.\n\nThat said, you\u0027re right that this is right on the border of a nova-specific upgrade issue and purely a schema versioning issue.","commit_id":"7cce929e6fc9d1f02464f0d6032b354d5953bbb2"}]}