CAUSE 1
"Enable Dynamic Memory Allocation" is enabled in the Agent Status Properties
CAUSE 2
Duplicate monitoring.
If two agents are monitoring the same targets or the same replica set there will be an issues when the agent gets flipped back and forth on the topology object. One agent should be used to monitor one sharded cluster or replica set or standalone with a unique database connection list per agent.
RESOLUTION 1
Uncheck the "Enable Dynamic Memory Allocation" in the Agent Status Properties (ASP) page for all mongodb agents.
RESOLUTION 2
Check the Database Connections in the Agent Status Properties of the MongoDB agents for duplicate monitoring where there are different hosts with the same alias or different replica sets with the same aliases.
If the agents were created with the wizard, using discovery, even if the discovery was seeded with different nodes, if those nodes are in a single replica set, the agents will each be monitoring every object.
The DBConnectionsList Agent Status Properties are shared across all the agents. Users should clone the agent status property list and configure the replicas as per the agents.
