Methods to configure multiple instances of PI APS

Two general methods for configuring multiple instances are described here. Although both methods should be generally applicable, some adaptation may be required for your specific circumstances.

Method 1

Method 1 involves partitioning the data source tags into distinct subsets. This method splits one physical data source into several smaller virtual data sources, which can then be treated in most respects as if they were independent data sources. One interface instance is associated with each subset of data source tags. The main advantage of this method is that PI APS can completely manage all aspects of synchronization for the interface instance associated with a specific subset of data source tags. That is, other than the configuration settings that define the subset of data source tags for each instance, the PI APS instances are independent of each other. As a result, Method 1 is simpler to understand and manage than Method 2.

See Example: Configure PI CIMPLICITY interface with multiple instances for an example of the configuration procedure for Method 1.

Method 2

Method 2 designates one PI APS instance to manage available points and its own existing points. The other PI APS instances associated with the same data source are restricted to updating their respective existing points. The essential element in this method is the use of the hidden points feature to configure the instance that manages available points to exclude the existing points of the other instances. The main disadvantage of this method is the difficulty of understanding the subtleties of managing the hidden points (and other APSPoints tables) for the available points instance. Another disadvantage is that the list of hidden points is likely to be large. As a result, memory and CPU usage can be adversely affected during synchronizations, and operations on the hidden points table in the PI APS Configuration Utility can be time consuming.

See Configure, add, register and synchronize interface instances for the procedural steps for Method 2.