Failure Detection and Redundant Remote Connections

You can configure keep-alive messages to detect loss of connection at the remote end of the VXLAN tunnel and redundant secondary IP addresses to restore the connection.

When you configure failure detection for a VXLAN site, keep-alive (ICMP echo) messages are sent to the remote end at a configured interval. If no response is received, the keep-alive message is resent as many times as configured for the associated retry count.

You can also configure more than one IP address for the remote site so that if the remote site does not respond to the keep-alive message within the specified time (interval multiplied by count), the next configured IP address in the configured series is tried. If a response is received from that IP address, it becomes the active connection for the remote site. One primary IP address and a maximum of 15 secondary IP addresses can be configured per VXLAN site.

Note: If all configured IP addresses are exhausted and all remote sites are unresponsive, the redundancy mechanism sets the keep-alive timer and retry count to their maximum values (20 seconds and 5 retries) and continues to loop through the IP address list checking for an ICMP reply from a remote VTEP. After an ICMP reply is received from a remote VTEP, the keep-alive timer and the retry count values are both reset to the values configured by the user.

Note: If an ICMP reply is received from another IP address after an active connection is established, the second response is dropped.