IPFE Failure and Recovery

If IPFE-A1 fails, the system handles it in the following manner:

• IPFE-A1's mate, IPFE-A2, detects the failure.
• IPFE-A2 takes over IPFE-A1's TSA, TSA1.
• There are no changes to the application servers in TSA1. TSA1 continues to comprise Server1, Server2, and Server3
• Traffic for TSA1 continues to go to TSA1, which is now managed by IPFE-A2
• IPFE-A2 continues to route TSA1 traffic to Server1, Server2, and Server3 - no different than they were before the failure.
• IPFE-A2 also continues to route traffic for TSA2 to Server4, Server5, and Server6.
• No disruption of service occurs.
• New connection requests for TSA1 is routed to Server1, Server2, or Server3.
• New connection requests for TSA2 is routed to Server4, Server5, or Server6.

When IPFE-A1 recovers, the following happens:

• IPFE-A2 detects that IPFE-A1 has recovered and relinquishes control of TSA1.
• IPFE-A1 assumes control of TSA1.
• Traffic that went to TSA1 continues to go to TSA1.
• The clients are unaware that a recovery has occurred.
• New connection requests for TSA1 continue to be routed to Server1, Server2, or Server3.
• New connection requests for TSA2 continue to be routed to Server4, Server5, or Server6.