This implements schema changes proposed in #57 and #63, namely:
Instead having ArchiveFileCopyRequest have a single instance for each (file, node_from, group_to) combination, and updating them in place as a new copy operation is requested after being completed or cancelled once, this changetracks the copying at a request level by:
adding new columns for start and end time of the transfer
once a copy request is completed or cancelled, requesting the copying of a file again will create a new instance of a copy request for the same (file, node_from, group_to).
With this semantics of copy requests, the n_requests column becomes redundant and is removed.
jrs65
commented
5 years ago
This implements schema changes proposed in #57 and #63, namely:
Instead having ArchiveFileCopyRequest have a single instance for each
(file, node_from, group_to)combination, and updating them in place as a new copy operation is requested after being completed or cancelled once, this changetracks the copying at a request level by:adding new columns for start and end time of the transfer
once a copy request is completed or cancelled, requesting the copying of a file again will create a new instance of a copy request for the same
(file, node_from, group_to).With this semantics of copy requests, the
n_requestscolumn becomes redundant and is removed.