task: benchmark storage performance

[ndushay]

related to / blocked by #30

For each of the types of reads and writes for the PCC, we will want to make sure our storage choice is performant:

• given the (large) quantity of data already in storage
• given the volume of read/write requests expected

Can think of this from perspective of:

• Trusted Checksum Repo
• Object Inventory
• Object Provenance

and/or from

• Audit process
  • Inventory process?
• Archive process
• Ingest process

[ndushay]

closed in favor of sul-dlss/preservation_core_catalog/issues/11