As part of our work on aggregation performance, we have added mongo-perf micro-benchmarks which test the aggregation equivalents of our find, count, and distinct benchmarks. The goal is to ensure that the aggregation path performs as well as our find, count, and distinct paths.

To provide an easy mechanism to run these new micro-benchmarks, we should create a new Evergreen task in the performance project called agg-query-comparison. This should run all of the agg benchmarks which are meant to be compared with the analogous tests run as part of the query and misc tasks.