FAQ : Connectivity (Tunnels and Subnets) : Tunnel endpoint didn't ping

Tunnel endpoint didn't ping

This basically means your side is not replying pings. This can be for several reasons.

You are blocking ICMPv6.
You are rerouting the ICMPv6 echo-reply through another route.
Your IPv4 path to the PoP is broken for proto-41 or AYIYA (depending on what is used).
Due to Connection tracking inbound packets are not always allowed.
You have just upgraded your Linux kernel to 2.4.21+.

The PoP you are connected to tests the latency of your connection by so-called pinging using standard ICMPv6 Echo Request packets from the PoPs side of the tunnel (tunnel::1) toward the user side of the tunnel (tunnel::2).

Every minute the PoP sends a single ICMPv6 Echo Request towards all tunnel endpoints that are configured (that is: admin & userstate enabled, and in case of heartbeat/ayiya that a heartbeat was received recently).

One can see the Latency check counters in the Live Tunnel Status. Go to your tunnel details and click on the link to monitor it.

Your endpoint should reply to these by sending an ICMPv6 Echo Reply along with a copy of the contents of the packet that the PoP sent. The time taken for the reply to come back determines the latency.

One can easily observe these pings by running tcpdump or Wireshark and looking at icmp6 packets coming to/from the PoP side of the tunnel.

When you have a static tunnel and the endpoint doesn't ping it costs credits. If you can't manage to fix this, your tunnel will be automatically disabled after 7 days.

The results of these latency tests are show on the Tunnel Details page with the latency graphs. When your tunnel shows gaps in the latency it means that during those times it was not reachable. The Tunnel Details page also shows when the system thinks your system was last reachable (alive) and previously unreachabled (dead) along with the amount of time it has been up.