Hi @chris,
So you mean by default the secondary will reference the oplog from primary for replication but when the operation exceeds the configured slowms threshold then the secondary itself will log a separate oplog?