Broker health and flow control

AVEVA PI System Connector 3 (PSC3) uses broker health reporting and flow-control mechanisms to regulate data flow and prevent downstream components from becoming overwhelmed. These mechanisms help ensure reliable data delivery while maintaining system stability under high load or during partial outages.

Broker Health Status

The broker continuously reports its health status to upstream components based on the number of unread messages it processes. The connector uses this information to determine whether data can be sent immediately or must be cached.

The broker reports health using the following thresholds:

• Target (~1,000 messages)

  Normal operating range. The broker accepts new data and processes messages without restriction.

• Maximum (~1,500 messages)

  The broker reports a busy status. Upstream components, including the connector, stops sending new data and may begin caching to disk.

• Recovery (~500 messages)

  The broker reports a healthy status again and resumes accepting new data.

Broker to Broker Chained Deployments

In chained broker topologies, such as DMZ traversal scenarios:

• A middle broker continues accepting data from upstream components if its downstream connection is unavailable.

• Data is cached locally on the broker until connectivity is restored.

• If the broker’s own cache approaches capacity, it reports a busy status to upstream components.

• Normal forwarding resumes automatically after downstream connectivity and capacity are restored.