Why should you configure send-flush-on failure on your PE routers when using active/standby pseudowires?
It flushes the local FDB speeding recovery
It discards all packets in the queue and sends icmp-discad messages to the sender. The packets are re-transmitted resulting in no loss of data.
On failure, the PE sends an LDP flush-all-from-me message to all PE devices in the VPLS. This speeds convergence to the new topology.
It is not important to use send-flush-on-failure as Active/Standby signalling automatically converges the FDB in all VPLS instances
Submit