The existing "histogram" selector supports workload definition as "traffic distribution by object size", however it runs the %/obj size ranges in "parallel". There is a desire to also support "sequential" mode where the traffic ranges are generated sequentially.
first 4% between 0-4K
next 8% between 0-16K
next 50% between 1M-10M
...etc
The existing "histogram" selector supports workload definition as "traffic distribution by object size", however it runs the %/obj size ranges in "parallel". There is a desire to also support "sequential" mode where the traffic ranges are generated sequentially.
first 4% between 0-4K next 8% between 0-16K next 50% between 1M-10M ...etc