vdbench could benefit from more control over how replay workloads are grouped. For example, it would be nice to be able to setup multiple (say 64), workloads, each of which uses a different replay file. Each workload could then be run against its own device.
That way, in one configuration file it would be possible to run multiple replays, against multiple targets. Presently, this entails creating multiple config files, and running multiple copies of vdbench against a specific config parameter file. This becomes particularly cumbersome with many replays.
Additionally, a "for" loop construct to repeat a replay file multiple times would be another great addition.