Documenting some feedback about this dashboard.

I believe the root cause here is a “workflow “problem.

In the dashboard “Dashboard MongoDB Collections Overview“ one will first select at the top of the page the cluster, the replica set and the database.

Some charts will be related to the selected database (expected) but some other charts, the ones included under the group “instance level” are more “global”, and include data from other replica set,specifically:

• Top 5 Hottest Collections by Read (Total)

• Top 5 Hottest Collections by Write (Total)

This is actually documented: https://docs.percona.com/percona-monitoring-and-management/details/dashboards/dashboard-mongodb-experimental_collection_overview.html however, it creates indeed confusion because a database was selected, so the expectations is that the 5 hottest collections by read and write are related to the selected database.

Please note - having such global charts in general is useful. The point here is about mixing the two things (global and database level).

Perhaps we can find a way to better separate / clarify things or have multiple versions (per database and “global”)

Thanks