Round-trip delay time

The Round Trip Time (RTT; " round trip time " ) or round-trip time indicates the time it takes for a data packet ( datagram ) in a computer network to travel from source to destination and back. It is the sum of running time from point A to point B and the duration from point B to point A.

Application of the measurement

This method of measurement is therefore much more commonly used in the practice of network technology as the ( measured only in one direction) latency ( or lag time, or engl. "Delay" ) because it does not elaborate time synchronization of the two terminals involved is required. In the RTT measurement is to be noted, however, that in many cases by asymmetric routing ( eg, due to BGP policies ) can occur unbalanced delay times and half the RTT thus does not necessarily provide a good approximation for the time delay in one direction.

RTT is used, for example, by the Transmission Control Protocol (TCP), to determine when packets should be re-sent after the absence of an acknowledgment. This measure is used to adapt the protocol behavior depends on the available transport capacity and under changing load conditions. A simple way of determining the RTT is measured from a plurality of packets, the time difference between transmission and receipt of the acknowledgment, and then to average over several measurements. Packets that had to be sent multiple times ( Timeout) should be ignored in the calculation, as it is not always clear which package was the confirmation.

Contrary to the definition in the Introduction section, the RTT is therefore not usually measured with a back and returned package. Instead, one usually measures so that host A sends a special packet to Host B (eg ICMP ping ). When receiving host B immediately sends back a response packet to Host A. The time between transmission of the first and receiving the response packet is then the RTT. This minimum difference is naturally rather academic in nature and then plays at most a role when sending the response packet significantly delayed at host B, for example by greatly increased CPU load.

End users can measure, for example with the Ping tool included with most modern operating systems, the RTT.

Typical measured values

Typical RTT values:

RTT vs. Ping

Ping test paths in the network and the routing in the network and returns the status usually in the form of ICMP error messages, packet loss rates, sequence numbers, information about packet sizes / fragmentation, time-to -live and, inter alia, also the round- trip time ( round trip time) ( the exact information depends on the used ping program or operating system). If you want to measure the RTT sense should not rely on Ping alone, since the one used by Ping ICMP protocol in many networks over standard TCP traffic is routed or prioritized differently.