Feature request for system design : Byte Range Fetch and Prefix specification

[julienlau]

Hi, Thanks for this very useful tool !

Would it be possible to implement byte range fetch in order to use this tool to benchmark different partitioning scheme :

• use byte range fetch on 8MB chunks to read 1GB of data from a large single
• use multiples objects of smaller sizes with the same prefix
• use multiples objects of smaller sizes with the different prefix

In addition, given the importance of sharding on performance, it would be nice to get a grip on how to use prefix to populate the object store. Example : option to specify a target of ~100 objects per prefix

Regards

[harshavardhana]

Feel free to send a PR @julienlau