Operational overview

Save PDF

Last UpdatedMay 01, 2026
5 minute read

This section provides an overview of how AVEVA PI System Connector 3 (PSC3) replicates AF objects from the source PI System to the destination PI System.

AF Replication

The connector supports the following options to replicate AF objects: Replicate AF and Data Archive and Replicate AF only. Unless otherwise stated, the procedures below apply to both configurations.

Initial Check at Startup

Prior to copying AF data to the destination AF server, PSC3 checks to ensure that all standard (Table Lookup, Formula, PI point, PI point Array, and String Builder) and custom AF data references present on the source AF database also exist on the destination AF database. This is to prevent errors during lengthy import operations on the destination that would effectively cause the import to fail.

Caution: This is a critical step and prevents PSC3 from proceeding further if one or more data references do not exist on the destination. Any missing standard and custom data references must be installed on the destination AF server prior to starting the connector.

Upon successful completion of the AF data reference check, you can start a full export of the source AF database or start a full export of a source AF element.

Start full export of the source AF database

Export the source AF server UOM objects (UOM classes and UOMs) and send the UOM objects to the destination AF Server.
Export the source AF database.
Process the exported AF database to find all the referenced PI points.
Prepare the exported AF database for import on the destination. This includes altering PI point configuration strings to include any prefixing done by the connector and broker.
Send the exported AF database to the destination.
Prepare found PI point information for data collection (Replicate AF and Data Archive option).
Start the AF discovery task.

Start full export of a source AF element

Export the UOM objects (UOM classes and UOMs) and send the UOM objects to the destination AF Server.
Export the source AF element.
Export the referenced AF objects of the AF element and its children.
Send the exported referenced AF objects to the destination AF Server.
Process the exported source AF element to find all the referenced PI points.
Prepare the exported source AF element for import on the destination. This includes altering PI point configuration strings to include any prefixing done by the connector and broker.
Send the exported source AF element to the destination broker.
Prepare found PI point information for data collection (Replicate AF and Data Archive option).
Start the AF discovery task.

Connector Restart

The following occurs on a restart:

If no previous export exists for the connector, the connector performs a full export as described above. Otherwise, the connector exports only the changes that occurred since the last discovery event.
Start the AF discovery task. The connector resumes where it left off after the shut down and then begins AF discovery to capture real-time events.

Discovery

AF discovery is a task that periodically checks for changes that have occurred in the source AF database since the initial startup or the last discovery event. The discovery task uses a timestamp that is updated immediately on the request for changes. During this discovery operation, pertinent changes are sent to the destination AF database. This includes AF object additions, AF object deletions and AF object updates. The timestamp is persisted and is read on the restart of the connector.

PI Point Replication: Replicate AF and Data Archive and Replicate Data Archive only

This section applies to the replication of PI points from the source PI System to the destination PI System. The connector supports the following options to replicate PI point data: Replicate AF and Data Archive and Replicate Data Archive only. Unless otherwise stated, the procedures below apply to both configurations.

Initial Startup

PSC3 collects data for PI points associated with data source configurations. For Replicate AF and Data Archive configurations, the connector identifies PI points for data replication through AF Data References. For replication of data archive only configurations, the connector does not require connectivity to an AF server or database. The connector identifies PI points for data replication by querying the source Data Archive for PI points that match the query defined in the configuration.

It does not matter which replication option you select, a PI Point is only replicated once, regardless of how many times it is referenced in the AF or PI Data Archive.

Note: The PI Point Query only applies to Replicate Data Archive only option, and not to any other replication setting type.

The data collection involves the following steps:

Load PI points from their respective source Data Archive.
Send selected PI point attributes for PI point replication on destination Data Archive.
Backfill history data for the PI points.
Collect current archive data for the PI points after backfill completes.
Collect snapshot data for the PI points that have compression disabled/off as indicated by the PI point attribute compressing=0 after backfill completes.
If a PI Point query is defined for a specific source Data Archive, replicate snapshot data for the PI points that match the source Data Archive snapshot query regardless of compressing attribute.
Start the PI point discovery task.

PSC3 supports history recovery. On initial start, it starts to backfill from the History Recovery Start Time up to the time the connector was started. PSC3 collects current archive data by signing up on the PI points as a consumer of the source Data Archive. PI points that complete the backfill will be added to the consumer sign-up list and start collecting current archive data.

Likewise, PSC3 collects current snapshot data by signing up on the PI points as a consumer of the source PI Snapshot Subsystems. PI points that have compression disabled/off, or satisfy the snapshot query, and complete backfill will be added to the consumer sign-up list and start collecting current snapshot data.

Connector Restart

When restarting a connector, history recovery depends on the last value that the connector received before shutting down. When restarted, PSC3 verifies the last update time for all PI points and retrieves historical data starting from either each PI point's last update time or the set History Recovery Start Time, depending on which is more recent. It then transitions to real time data collection.

Discovery

The PI point discovery task auto-discovers the new points added to the configured source database or source AF element branch through PI Point Data References.

When a source Data Archive is specified in connector data source configuration, it also discovers the changes in the PIPOINT database of the source Data Archive. Any newly identified PI points will be generated on the destination PI Server, and an initial backfill will be conducted from the History Recovery Start Time prior to streaming the current data. History Recovery Start Time is specified in the Shared Data Source Options for the data source.

Operational comments

The following comments relate to general PSC3 operation.

PSC3 never creates or alters PI points on source Data Archives.
PSC3 never creates or alters the AF structure on the source AF.
PI points referenced by the AF structure referenced in the Replicate AF and Data Archive or matching the query associated with the Replicate Data Archive only data source configuration options are created and updated on the destination Data Archive.
PSC3 does not interpret the %Element%.%Attribute substitution parameters found in the AF export. The connector only checks for configuration strings that map to existing points.
PSC3 cannot delete PI Points on the destination Data Archive. It is capable of creating and updating PI Points, but not removing them.

PI System Connector 3