IP QoS Tester
Quality of Service (QoS) measurement for IP transmissions
IP QoS Tester is a very easy solution to measure the most important parameters concerning Quality of Service (QoS) over IP networks.
IP Qos Tester is a solution containing three applications:
- the Sender application
- the Receiver application
- the Reflector application
The Sender application sends UDP datagrams to the Receiver application (identified by its IP address and its user-defined port number).
Between the Sender and Receiver applications, the Reflector application can serve as an intermediary by reflecting the data it receives towards another machine.
The user can choose the size of the datagrams and the number of the datagrams to send during a test.
If the size of a datagram is lower or equal to the MTU, then Sender and Receiver work on a packet basis (instead of a datagram basis).
The Receiver checks the received packets and measures the most important parameters concerning QoS:
- Packet (or datagram) loss : number of lost packets (or datagrams) and percentage of lost packets (or datagrams)
- Packet (or datagram) corruption : number of corrupted packets (or datagrams) and percentage of corrupted packets (or datagrams)
- Packet (or datagram) delay
- Packet (or datagram) jitter
- Packet (or datagram) reordering
IP QoS Tester is a set of 3 tools: Sender, Receiver and Reflector.
IP QoS Tester measures the delay, packet loss, jitter, reordering, etc... of packets sent from one machine to a second machine.
There are two ways to use IP QoS Tester:
- You install the Sender tool on one machine (a server for example) and the Receiver tool on the other (a client). But to do that, you have to make sure that the two machines have their clocks synchronized with an NTP server.
- You install the Sender tool and the Receiver tool on the same machine (a server for example) and the Reflector tool on the other (a client). When it receives packets (from the Sender tool), the Reflector tool will re-send them to the Receiver tool. Of course, the packets received by the Receiver tool will have been transmitted twice (Sender -> Reflector, and Reflector -> Receiver). Then you have to assume that the real delay between server and client is the half of the measured delay. Same things for the number of lost packets, reordering...