We modified the code to reduce the overhead of checkpoint configuration parsing as part of WT-5042. We would like to measure the overhead of checkpoint configuration parsing.

Construct a new Genny workload that creates ~50K collections and indexes. The LargeScaleLongLived.yml workload gives an example of creating 1000 collections, each with 10 indices, and creating steady read and write traffic to each. 1000 is a parameter, we can dial that up. Sue suggests a varying amount of work. I don't know how to do that with the existing actors (either varying with time or varying by collection) without modifying the actors. But we can control the overall workload by specifying updates per second - and we can control which fields are updated.