To measure the Network Performance between two agents, agents must communicate together. When a monitoring session is created, the system automatically select one agent as the generator (G) and the other one as the reflector (R). The generator is sending a packet to the reflector every 500ms and the reflector replies back when it receives the packet. The measurement information is added inside the packet.

When both agents are on the same private network (i.e. they can communicate using private IPs), they must be on the same Network inside the App. If this is the case, agents should be able to communicate together. Learn more about Networks at Network Configuration - Under the hood.

If the two agents are not on the same private network, configuration is required to let one agent receive packets from the other agent. You have to decide which agent you would prefer as the Reflector.

In the Agent Advanced Parameters, it is possible to increase or decrease the Generator Preference. In case both agent can be the generator, the agent with the highest preference will be selected.

It's a best practice to have the agent that participate in many monitoring sessions to be the Reflector (the Hub in a Hub-and-Spoke model). It's a better configuration because it scales more easily. The Reflector is using less CPU processing than the generator and it also reduce the configuration required.

Once you selected which agent you prefer as Reflector, you must do the following configurations to make sure the system selects it:

  1. Enable Accept Incoming Public Connection in the Agent Advanced Settings.
  2. Write down the port and Private IPv4 Address used by this agent. The default port is 23999. The agent is using both TCP and UDP.
  3. Make sure your router/firewall allows/redirects that TCP and UDP ports to the agent IP address.
Did this answer your question?