Library: Test Plans

10 Gigabit Ethernet Testing

The test plans presented in this section include several examples demonstrating how Ixia's solution addresses the challenges of 10GE switch testing.

1. Throughput Testing: RFC 2544 Throughput Test
2. Testing for Latency: RFC 2544 Latency Tests
3. Testing how 10GE ports handle prioritized streams based on QoS parameters
4. 10GE OSPF Convergence Test

1. Throughput Testing: RFC 2544 Throughput Test

Objective.
To characterize the 10GE switch data plane performance in forwarding traffic. IETF RFC 2544 defines a test methodology for this characterization, divided into the following three tests:

• The back-to-back test determines how the DUT responds to different quantities of frames with the minimum gap allowed by the protocol specification.
• The frame loss test determines how the DUT responds to streams with different loading.
• The throughput test finds the highest rate at which the DUT can forward frames.

Figure 1. RFC 2544: Throughput Test setup.

Setup.
A minimum of two 10GE test ports is required for this test. Ixia's IxScriptmate application can be used to run RFC 2544 tests.

Parameters.
Select the test port pairs, frame size mode and frame sizes, and number of iterations.

Methodology.
This test should involve several test ports to match the DUT's port density. Ideally, the test should flood traffic to every input port of the DUT. A number of Ixia load modules will be connected to the DUT. Ixia's IxScriptMate is used to perform the RFC 2544 benchmark test.


Figure 2. RFC 2544: Throughput Test results.

Result.
The results of the test show the throughput rates, in frames per second, obtained for each frame size. The results also show the average throughput rates for all the trials.

2. Testing for Latency: RFC 2544 Latency Tests

Objective.
To determine the latency of the DUT, and how much it varies with different frame sizes.

In this test, frames are transmitted for a fixed duration. Once per second, the test tags a frame and transmits it halfway through the duration time. The test compares the tagged frame's timestamp when it was transmitted with the timestamp when it was received. The difference between the two timestamps is the latency.

Setup.
A minimum of two 10GE test ports will be used for this test, in conjunction with the IxScriptmate RFC2544 test. Latency will be measured in both directions.

Parameters.
Select the test port pairs, frame size mode and frame sizes, and number of iterations.

Methodology

1. Packets will be sent to the DUT from each 10GE load module starting with the selected minimum frame size for an interval that lasts for selected test duration.
2. The average latency is then measured.
3. These two steps will continuously repeat, using selected frame sizes every interval until it reaches the maximum frame size selected.

Results.
The results of the test show the latency in nanoseconds (ns) obtained for each frame size and test port pair. The results also show the average latencies across the pairs.


Figure 3. RFC 2544: Latency Test results.

3. Testing how 10GE ports handle prioritized streams based on QoS parameters

Objective.
To determine the maximum rate at which the DUT can forward frames correctly according to their priority settings.

The test sets up a configuration in which many ports, each with a different priority, sends traffic to one single port. The receive port may be overloaded in order to test the receipt of the highest priority frames.

This test supports both MAC and IP layer frames; priority bits may be specified either in the IP precedence bits or in the 802.1p header. Latency may optionally be calculated per priority level. Test results are: the transmit and receive rates per priority, percent loss, and optionally, latency, for each priority per port.

Setup.
A minimum of two test ports will be used for this test, in conjunction with the IxScriptmate QoS Many-to-One test. In this example, one Ixia 10GE test port and two Ixia 1 GE test ports connected to the DUT all traffic directed to a single 10GE port on the DUT.



Figure 4. Testing how 10GE ports handle prioritized streams based on QoS parameters setup.

Parameters
Select the three test port pairs, frame sizes, QoS parameters for each port, and number of iterations.

Methodology

1. Packets will be sent from all ports. Multiple streams with varying QoS parameters will be sent to the one egress port.
2. Total frames received, latency, and frame forwarding rate are measured per priority.
3. Steps 1 and 2 continuously repeat, using selected frame sizes every interval until the test reaches the maximum frame size selected.

Results.
The results of the test show, on a per priority stream, the total frames received, the receive rate, and percentage loss for each frame size.


Figure 5. Testing how 10GE ports handle prioritized streams based on QoS parameters results.

4. 10GE OSPF Convergence Test

Objective.
Verify the ability of a router to switch between preferred and lesspreferred routes on its 10GE ports when the preferred routes are withdrawn and readvertised. This will test the control plane function on the DUT's 10GE line card.

In this test, two 10GE ports simulate OSPF routers, Router 1 and Router 2. Both routers advertise the same route prefixes to the simulated network. However, the routes advertised by Router 1 will have smaller metrics (lower costs), which should cause the DUT to forward traffic over them instead of Router 2's routes. After advertising the routes, the transmit port begins transmitting a stream of packets to an address in each of the advertised route prefixes. The DUT should forward all the packets over the route with the lower metric (Router 1). After a time interval has elapsed, the receive port simulating Router 1 withdraws its routes. The DUT should detect that the preferred route has gone down and switch traffic to Router 2's routes. Router 1 then re-advertises its routes. The DUT should again detect a change re-route traffic to the receive port simulating Router 1.

Setup.
This test uses three test ports one to transmit and two to receive (see Figure 6 below). One Ixia test port (acting as the transmit port) and two Ixia 10GE (test ports 1 and 2 acting as receive ports) connect to the DUT. Receive ports will emulate OSPF routers. The traffic is unidirectional. The DUT must have three ports utilized with two enabled for OSPF. All three ports should be configured for IP and have unique subnets in which to communicate with the tester ports.



Figure 6. 10GE OSPF Convergence Test setup.

Results.
The test results provide an average convergence time for all routes. Figure 7 below displays example results for the automated OSPF convergence test in IxScriptMate. In addition to convergence time, this test also indicates the amount of lost packets caused by the convergence.


Figure 7. 10GE OSPF Convergence Test results.

Acknowledgements

Authors: Ted Fornoles: Product Manager, Alireza Safari: Product Manager

Editor, Illustrator: Elliott Stewart

[ back | top of page | back to test plans ]