solidDB also provides a High-Availability Controller (HAC, for short) software to automatically detect errors
in HSB and recover from them. Every HSB node includes a solidDB HSB server, and a HAC instance.
In addition to sub-second failure detection and fail over, HAC handles all single failures and several failure
sequences without downtime. Network errors are detected by using an External Reference Entity (ERE) [15].
External Reference Entity which can be any device sharing the network with HSB nodes, which responds to ping
command. When HSB connection gets broken, HAC first attempts to reach the ERE to determine the correct
failure type. If ERE responds, HAC concludes that the other server has either failed or become isolated. Only
after that the local HSB server can continue (if it was Master) or start (otherwise) to execute write transactions.