Automatic failback of an Orace RAC database Service - 19c new feature

When a rac database instance fails, a service which is running on this failed instance (configured as preferred) will fail over to the available instance to continue application connectivity. Wheh the failed instance restarted again, Unfortunately, the service did not fail back to the original (preferred) instance as soon as the instance is up. So a DBA need to relocate the service again. This has changed with Oracle RAC Database 19c with a feature called "failback -yes" option. 

oracle@node1:~/ [RAC1] srvctl add service -db RACDB -service PRODSRV -preferred RAC1 -available RAC2 -failback YES 

When you fire above srvctl command, cluster will make sure that the service PRODSRV is always running on RAC1 instance as long as RAC1 instance is available.