It would be great to have a lightweight version of the --stats argument during creation of backups.
Currently the --stats argument makes the backup much more expensive due to what it prints as part of statistics.
Could you consider printing a subset of the current --stats output so that it is very light weight but still gives some useful metrics including files changed/deleted, total files, size of the archive and whatever other cumulative stats that can be printed without making it "expensive"?
The "All archives" line can definitely be skipped (assuming it is expensive) since that can also be derived from a repo call (as it is independent of a specific archive).
Part of the statistics is also very expensive time wise if there is a high latency connection.
It would be great to have a lightweight version of the
--statsargument during creation of backups.Currently the
--statsargument makes the backup much more expensive due to what it prints as part of statistics.Could you consider printing a subset of the current
--statsoutput so that it is very light weight but still gives some useful metrics including files changed/deleted, total files, size of the archive and whatever other cumulative stats that can be printed without making it "expensive"?The "All archives" line can definitely be skipped (assuming it is expensive) since that can also be derived from a repo call (as it is independent of a specific archive).
Part of the statistics is also very expensive time wise if there is a high latency connection.