Agent Instance

Definition

Agent instances are the central components of PCo. They establish the data flow between the data source and PCo and enable the processing and forwarding of notification messages. Moreover they process queries and method calls.

Agent instances are based on the agent that implements the special source system type and are each connected with exactly one source system.

However, you can also create agent instances without a source system. Then you can run the agent instance as an OPC UA server or as a Web server.

When you start agent instances, each instance is executed as a standalone Windows service or standalone program in the background. Therefore, you can run multiple agent instances at the same time and independently of each other. The number of agent instances that can be executed at the same time is only restricted by the capacity of the computer on which PCo is installed. You can monitor the status of the individual agent instances at runtime using the Management Console or the remote client. If there are fatal errors, the agent instance stops and is displayed as faulty.

Before you can start an agent instance, you need to make your settings. Configuration of an agent instance depends on its intended purpose and on the agent used. Depending on the functionality of the agent used, specific configuration options for the agent instance are hidden. For example, if you create an agent instance for an agent that does not support notifications, the tab pages for notification processing and for notification message queues are not available.

Structure

The following tabs are available for configuring an agent instance:

The following tabs are also available in the notification process for checking the notification messages:

  • Queued Messages tab

  • Message Failures tab

  • Expired Messages tab

  • Messages for Bundling tab

For more information about agent instances, see Starting and Stopping an Agent Instance and Agent Instance: Displaying Messages.