RFC2544 Test Indicators

RFC2544 RFC2544 RFC2544 provides a benchmark for network device testing. It defines a series of testing processes and methods so that service providers and users can use the same benchmark, reach consensus on the implementation and results of the test. RFC2544 standard frame length: 64 byte, 128 byte, 256 byte, 512 byte, 768 byte, 1024 byte 1, Throughput (Throughput) the maximum rate at which the device can transmit the received frame signal without any loss. Simply put, the device transmits the received frame signal from the source sender to the Target receiver without packet loss, the maximum amount of data that can be transferred per unit time. For an Ethernet system, the absolute large throughput rate should be equivalent to the interface speed. In fact, because different frame lengths have different transmission rates, these absolute throughput rates cannot be achieved. Smaller frames have lower transmission efficiency due to the frame interval of the forward guide code. The maximum Data Throughput (Data Throughput) of a 64-byte frame is 76.19 MBit/s, and 148809 frames can be transmitted per second. For 1518byte frames, 98.69 MBit/s and 8127 frames/s respectively. Test requirements: Generally, the test duration is set to 20 s. To overcome the random impact, the number of tests in each test case is set to 20 times. The test granularity is set to 1% of the theoretical rate. 2. Latency refers to the total transmission time of a frame from the source point to the destination point. This time includes the processing time of the network node and the propagation time on the transmission media. The general test method is to send a frame with a timestamp. After the network is passed, the receiver compares the current time with the timestamp carried by the frame to obtain the latency value. Considering the time synchronization problem, it is generally called a two-way latency to compare the frames sent back to the sender. There are two defining methods: store and forward latency, S & F, and cut through latency (CT ). Storage and forwarding latency refers to the time interval between the last bit of the data frame arriving at the input port of the device and the time when the first bit of the data frame appears on the output port of the device. The latency is calculated based on the following method; the time when the first bit of a data frame arrives at the input port of the device and the time when the first bit of the data frame appears on the output port of the device. This is calculated in FIFO mode. RFC2544 requires that the latency test be repeated for at least 20 times for at least 120 s. The average value of all test results is obtained. 3. Frame Loss Rate is the number of frames sent by the sender but not reached the receiver. It is generally expressed as the frame loss rate. That is, a percentage relative to the total number of sent frames. Calculation formula: Packet Loss Rate = number of packets not received by the receiver/total number of packets sent by the sender * 100% 4. Back-to-back (Back) belongs to the Boundary Value Test scope, sends N frames with the minimum frame interval (the minimum frame interval is 0.96 microseconds) to the tested device without packet loss, counts the number of frames sent by the tested device. If the number is equal to the number of sent messages, N is added and the test process is repeated. Until the number of frames sent by the tested device is smaller than the number of frames sent by the test device. Otherwise, the number of sending frames is reduced, and the packet sending time is reduced until no frame loss occurs. It is mainly used to measure the maximum storage and forwarding capabilities of tested devices with storage and forwarding capabilities. It is mainly related to the following factors: the size of the buffer inside the network device; the speed difference between the inbound and outbound network devices; the size of the forwarding capability of the network device; scheduling Algorithms for network device switching networks. In the standard, the sending time must not be less than 2 s. We recommend that you repeat at least 20 times and obtain the average value of the result. 5. System recovery is used to test the System recovery capability of a device under overload. The test process first sends data of at least 60 s based on 1.1 times of the maximum throughput of the device to be tested, and then drops the statistical rate between 50% and no frame loss, that is, the system recovery time. 6. the Reset test (Reset) is used to test the time between system Reset and normal recovery. The testing process first sends the minimum frame length based on the maximum throughput, and then resets the device to be tested, calculate the difference between the time stamp of the last frame sent before resetting and the time stamp of the first frame received after resetting, that is, the reset test time.