1. 2
  1.  

  2. 1

    In many cases I would rather a slower database and storage engine that behaves in predictable ways over longer periods of time than one that is really fast in the beginning but can degrade really badly later on in production.

    You can flip this though: the database is slow all the time (“really bad in production”) but has a small-data optimization. That is, if you capacity plan for what performance when the levels fill up, performance when the levels are not full is just gravy.