Three node system. RAC 1 is down and RAC 2 and 3 is up. The Oracle VIP for Node 1 fails over to Node 3. When you go to a client server and do a tnsping to the rac 1 VIP address, it answers no listener. It's the VIP address and not the physical IP address. Is there something missing on the Oracle cluster software/config for the three rac node system that is not redirecting the node 1 VIP to node 3
a) clients (11.2) connect to the DB using the SCAN address - on the event of failover, the service is removed from the SCAN listener, so it will not forward requests to the failed VIP - No problem here.
b) clients (pre 11.2, but works for 11.2 also of course) have a list of VIP addresses - When a client tries to connect through a failed VIP, it will immediately get the reply that there is no listener there, so it can try another address without wasting time.
Anything else than the above is not a proper HA configuration.
"A VIP address is an alternate public address that client connections use instead of the standard public IP address. If a node fails, then *the node's VIP address fails over to another node but there is no listener listening on that VIP, so a client that attempts to connect to the VIP address receives a connection refused error (ORA-12541) instead of waiting for long TCP connect timeout messages. This error causes the client to quickly move on to the next address in the address list and establish a valid database connection.*"