Store-Forward

A store-forward event occurs when a data source is unable to communicate with the Wonderware Historian. When this occurs, the data source begins caching the data until the connection with the Wonderware Historian is reestablished. When the connection is reestablished, the data source sends the most current data followed by cached data in order of the oldest to the most recent. At the same time, the data source sends the most current data as it arrives.

The store-forward time span is defined to be the time the data source begins caching its data to the time the Wonderware historian has completely captured the cached data. The time span includes the loss of communication with the Wonderware Historian and reestablishment of the data source to the historian.

There is no reliable way for Wonderware Historian SDK 2014 R2 to determine whether the history data is currently in a store-forward event, or if the store-forward event is complete and all the backfill data is fully historized. Consequently, the Wonderware connector uses a complex algorithm to determine the current store-forward state of the collected Wonderware tags and when they are completed.

Following are the types of data changes that are encountered during a store-forward event.

Store-forward data collection speeds

• Fast-changing data

  Two or more values are updated during the store-forward time span.

• Unchanging data

  No change to the data occurs during the store-forward time span.

• Slow changing data

  Zero to one change occurs during the store-forward time span.

Store-forward data values

• Quality bad (1) and quality detail Connection Loss (10)

  Indicates that the historian has lost connection for the tag. The algorithm uses this state to ascertain if the tag is in store-forward.

• Quality detail First-Store-Forward data value (248)

  Indicates that the value is the first value in the stream for the store-forward data.

• Quality detail RPC Reconnect (44)

  Indicates that the first good value received by storage after the connection to HCAL (Historian Client Access Layer) has been restored.