[rincebrain]

I think we're basically agreeing loudly over everything except the example being a demonstration of mitigating the impact of BPs being immutable while adding a feature that requires that statement be less than true - and I agree, the permanent overhead of a mini-DDT is a non-starter for anything other than the example case of "oops I added a device, time to evac it before $TONS_OF_DATA gets landed".

Certainly, it would be much less exciting to send|recv from poolA to poolB, and require no code changes and no GB per TB of data indirection overhead.

But this was intended as an example of how many caveats and problems are involved in even a "simple" feature involving shuffling data on-disk, and thus, why "defrag" is a horrendously hard problem in this environment.