Expand performance testing of large volume sizes

[awoods187]

Describe the problem
While we have #24277, I think we should split up disk and cpu testing. I'll focus this on disk size.

I ran an experiment this weekend comparing SSD under TPC-C 2500, EBS under TPC-C 2500, EBS with 1TB under TPC-C 2500, EBS with 1TB under TPC-C 10k:

I'm not really sure what to make of it as its a bit inconsistent and it matters the runs you put in. My main conclusion (besides the variability of the runs) is that we do have some degradation when you load large amounts of data on a node but only run a smaller active amount. I can't run this enough times because of #30213

The tpcc load btw is 2.1tb of 2.7tb on the last two as opposed to only 555gbs on the first two.

We need to expand our testing under load/large storage volume