Benchmark data generator utility

[samricotta]

We propose a benchmark data generator utility in order to aid the reproduction recovery of archive node data sets as we are finding it hard to do this from an arbitrary chain is too much effort.

It must be configurable with the following parameters:

• seed
• store key
• initial tree size
• max tree size
• delete fraction
• update fraction
• mean/std_dev key size (use https://go.dev/src/math/rand/normal.go)
• mean/std_dev value size

[kocubinski]

Started work here: kocubinski/iavl-bench@800771a

[alexanderbez]

Started work here: kocubinski/iavl-bench@800771a

Just a small nit -- this should probably be in the cosmos org as it'll have high utilization for us and others going forward. We can move it once it's in a good enough shape.

[tac0turtle]

closing as i believe this has been delivered