Learn about buffering

In an AVEVA™ PI System™ implementation, buffering protects against data loss for PI interface and PI client computers when Data Archive is unavailable. Buffering is implemented as a service that stores and forwards events to a Data Archive server or collective.

If the Data Archive server becomes unavailable for any reason, the buffering service temporarily stores the data from applications running on the buffered computer. When the connection with Data Archive is restored and the server becomes available, the buffering service sends all the stored data from the buffer to Data Archive in chronological order.

In this manner buffering provides continuous data collection during network outages or other unexpected failures, or when a Data Archive server is down for maintenance, upgrades, or backups. N-way buffering supports high availability and redundancy by ensuring that all members in a collective receive the same data.

We recommend using buffering on any computer with software that writes data to Data Archive, including PI interfaces and connectors, and from manual data entry applications including Manual Logger, DataLink, and custom data entry applications developed with AF SDK or PI SDK.