Our network currently forwards IP packets to and from the Internet at a maximum size of 1500 bytes
Our core network is capable of transmitting and receiving up to 1600 bytes of payload but customer connections are by default configured to maximum 1500 bytes. If you wish to transport more than 1500 bytes of payload over your circuits, please contact your account manager (business customers only) to verify that your circuit will support this.
Our core routers de-prioritise ICMP messages that are destined to, or originate from them. The result of which is that traceroutes through our network may display high RTT (round trip time) figures or, in extreme cases, packet loss from the intermediary core nodes. We would like to advise customers that this behaviour is completely normal and that customers only raise support issues if the final hop in the traceroute displays errant behaviour
Additionally, VPN:ng customers may notice that traceroutes contain many hops which do not respond (displaying stars, perhaps), this behaviour is intentional and a result of the product not supporting traceroute as a debugging tool, we would urge customers to contact our support to assist with debugging
Finally, we do not permit fragmented UDP traffic to or from the Internet. If you are wishing to use eDNS (RFC 2671) between the Internet and our network, please ensure that you set your maximum UDP payload size to the minimum MTU size, including packet overhead, since most implementations do not do path MTU discovery and the failure to deliver the UDP fragments will result in the maximum UDP size being reduced to 512 bytes as per RFC.
© 2012 Claranet Limited ISO9001