When to Scan
UNEXPECTED HISTORIC DATA VALUES
The driver retrieves historic data from the FloBoss using multiple transactions. Because of this, and the fact that slow radios are sometimes used to retrieve data from FloBoss devices, it is important that the driver is not configured to scan a historic point very close to the time that its value is scheduled to change.
Example:
On a particular system, a point that retrieves historic data is set to be scanned exactly on the hour. However, the host clock is 2 seconds faster than the clock in the FloBoss. The scanner starts a scan at 9:59:58 according to the FloBoss clock, and the new hour hasn’t started. The driver retrieves some data about the pointers. The hour then rolls over in the FloBoss and new records are created. The driver continues retrieving data from the FloBoss, but the pointer values that were retrieved in the first part of the scan are now wrong. This potentially causes incorrect data to be pushed into the database, because the data changed as the driver was in the process of retrieving it.