As I tested, primary and secondary node receives TXs at same time working in parallel.
But if primary node crash, secondary will not take place and web app will be offline, and will not help to continue validation.
So at present secondary node is not a "fail over", just working in parallel.
As I tested, primary and secondary node receives TXs at same time working in parallel. But if primary node crash, secondary will not take place and web app will be offline, and will not help to continue validation. So at present secondary node is not a "fail over", just working in parallel.