Hi, We need to read multiple locations in a big object(4MB) at the same time with librados and each part is 16KB. If librados provides a batch read api(vector<pair<offset, length>>) and issue multiple requests to SSD, will there be any improvement in performance?
