Use Atlas historical connection rate information from the data warehouse to determine what incoming connection rates are typical for clusters and how those rates vary by instance size, workload, and over time (e.g. after a server restart or an election).

Correlate connection rate information with resource utilization information (e.g. CPU utilization) if possible.

The goal is to determine rough initial values for the connection establishment rate limiter across different clusters, and to identify outliers that are worth studying more closely to inform the Mongotune policy's design.