Connection Mirroring in BIG-IP LTM

Connection mirroring is a feature in BIG-IP LTM used to maintain session persistence and ensure high availability for long-term connections, such as FTP, TELNET, or SSH. It is an essential feature in systems that need to maintain the state of ongoing connections in case of failover between the active and standby units.

---

Key Aspects of Connection Mirroring:

1. Long-Term Connections:

   • Primarily used for long-lived connections that need to be mirrored between active and standby units, especially protocols like:

     • FTP

     • TELNET

     • SSH

   • These protocols are stateful and require the connection to remain intact, even if the system switches from the active unit to the standby unit.

2. Short-Term Connections:

   • Short-lived connections (like HTTP and UDP) should NOT be mirrored because they tend to have very brief lifespans, and the overhead involved in mirroring would not be justified.

   • Mirroring these types of connections can lead to performance degradation as it consumes unnecessary system resources.

3. CPU-Intensive Operation:

   • Connection mirroring is a CPU-heavy operation because it involves synchronizing the state of active connections between two systems (active and standby).

   • It requires substantial processing power, especially when dealing with many simultaneous long-term connections, which can affect the overall performance of the system.

4. TCP Ports Used for Mirroring:

   • Mirroring traffic between the active and standby units typically happens over the TCP port range 1029-1155.

   • This allows for communication between the units and the transfer of connection state information.

   • Up to 127 traffic groups can be mirrored from the active unit to the standby unit, enabling redundancy for large-scale deployments.