failover – How to Fail-over of MYSQL 8.0 on Red Hat Cluster without Application to Re-login & avoiding service interruption


Getting the error after every fail-over test case is run :- host is blocked because of many connection errors . We need to login back again to database and use Flush Host command to allow new connection to the database ,but this disrupts the ongoing services and defeats the whole purpose of HA .We have bought Red Hat HA Add-on licenses to create the Cluster on which MYSQL 8.0 DB has been installed . App connects to the DB via a VIP and during any DB transaction if we test the Cluster node Fail-over ,it disrupts the ongoing App jobs . Tried using Connection Time-out set to 30 secs (by this time the fail-over completes and DB resources successfully move to the second node ) and setting Connection life-cycle to 10 sec on the connection string ,but same behaviour .Set the SET GLOBAL max_connect_errors=10000 as well . Please suggest a way forward .