Hi,
I have noticed a bug in the way NPS logs to SQL in the schema and stored procedure that it creates.
At present, it uses a database schema that uses ints (which are fundamentally 32-bit signed) for data that is in the RADIUS specifications 32-bit unsigned and therefore has a larger range of potential positive values. This means that bigints (signed 64-bit)
must be used to correctly store the values.
This bug means that I frequently see insert failures where a value from a NAS overflows an int and failure events from NPS in the Event Log.
How would I go about getting a fix from Microsoft to correct this behaviour or at least making the product team aware of this bug?
Nick