A Comparison of Carrier-Grade Ethernet Testing Methodologies
(RFC 2544 Vs ITU-T Y.1564)
Welcome to another September, 2016 issue of GL Communications' Newsletter providing information and insight into High-Performance Carrier-Grade Ethernet Testing Services:
- RFC 2544 - Benchmarking methodology for network interconnect devices
- ITU-T Y.1564 - Ethernet service activation test methodology
Service providers are shifting from only providing Ethernet pipes to enabling Ethernet services. Networks must support multiple services from multiple customers, and each service has its own performance requirements that must be met even under full load conditions and with all services being processed simultaneously.
IETF RFC 2544 standard, titled “Benchmarking Methodology for Network Interconnect Devices” has been designed to test network equipment in a lab environment. With the introduction of Carrier Ethernet, service providers started offering Ethernet as a transport service. Initially, the existing benchmark RFC 2544 was adopted to test such Ethernet transport services, but soon, it was found to be inadequate for the purpose.
RFC 2544 was designed as a performance tool with a focus on a single stream to measure maximum performance of a Device Under Test (DUT) or Network Under Test (NUT) and was never intended for multiservice testing. From a laboratory and benchmarking perspective, the RFC 2544 methodology is an ideal tool for automated measurement and reporting. From a service turn-up and troubleshooting perspective, RFC 2544, although acceptable and valid, does have some drawbacks.
The new ITU-T Y.1564 standard has been specifically created to test basic services on packets and to resolve gaps of the RFC 2544 standard. ITU-T Y.1564 is intended for multiservice testing to measure the maximum performance of the DUT or the NUT.
ITU-T Y.1564 standard is a set of procedures that test the ability of Ethernet-based services to carry a variety of traffic (voice, data, and video) at defined performance levels. In particular, it is aimed at addressing limitations of legacy RFC 2544 test procedures, especially for Service Level Agreements (SLA).
Aligned with the requirements of today’s Ethernet services, the ITU-T Y.1564 test methodology enables complete validation of all SLA parameters in a single test to ensure optimized quality of service (QoS).
GL offers PacketExpert™ a comprehensive testing tool for 1 Gbps/10 Gbps wirespeed Ethernet/IP networks. The hardware includes both Optical and Electrical interfaces capable of RFC 2544, and ExpertSAM™ (ITU-T Y.1564) functionalities.
Additionally, PacketExpert™ also includes other important functionalities such as BERT, Smart Loopback, Record and Playback, IP Link Simulator (IPLinkSim™), IP Network Simulator (IPNetSim™), PacketBroker and Multi-Stream Traffic Generator and Analyzer.
Testing Ethernet Line Vs Testing Ethernet Services
GL’s ExpertSAM™ test suite is designed to test a service per ITU-T Y.1564 standard rather than a line per RFC 2544 standard. The RFC 2544 is limited to test maximum output for a single service i.e. it treats the whole Ethernet Line as a single service without distinguishing the multiple services within the line, while the ExpertSAM™ allows to create and test more realistic multiple services with parallel flows from the bandwidth of 0 to maximum.
Sequential Vs Simultaneous Testing
RFC 2544 method tests and measures the Throughput, Frame Loss, Latency and Burstability parameters in sequential order, one after the other. This not only takes a lot of time to complete and makes it unsuitable for real world testing, where all the parameters have to be measured in parallel. On the other hand, ExpertSAM™ allows testing aforementioned critical parameters of all services simultaneously. This allows a significant time saving compared to RFC 2544.
ExpertSAM™ service configuration test is performed sequentially per service. For each service, it verifies that all the KPIs or SLA parameters are met, and that the service is provisioned properly.
ExpertSAM™ service performance test guarantees that the network can stand all the services at their lowest guaranteed output simultaneously on a significant period (a few minutes to longer testing hours) constantly validating KPIs or SLA parameters - Information Rate (Throughput), Frame Loss Ratio (Packet Loss), Frame Transfer Delay (Latency) and Frame Delay Variation (Jitter).
Testing Maximum Throughput Vs Testing the Bandwidth Profile
Each Ethernet Service is defined by a Bandwidth profile, which defines the normal and peak data rate that the service can handle. Bandwidth Profile is defined by a set of attributes:
Committed Information Rate (CIR) - this is the guaranteed bandwidth. Customer is assured of this rate.
Excess Information Rate (EIR) - sometimes during peak traffic hours, customers may generate traffic above the CIR. In such cases, the Bandwidth profile defines an EIR. This is the rate above the CIR, in which the service tries to deliver the frames on a best effort basis (meaning it could be dropped in case of congestion or other higher priority traffic).
Traffic Policing - Any traffic sent by the customer that is above the EIR is immediately dropped by the service provider.
RFC 2544 simply measures the maximum Throughput of the service, without any consideration for committed and excess traffic. It is a simplistic test that only measures the maximum capacity of the line. Whereas ExpertSAM™ allows defining the CIR, EIR and the Traffic Policing rates, and for each service, generates traffic at these rates, and measures SLA parameters for both CIR and EIR rates.
RFC 2544 provides Throughput, Latency, Frame Loss and Burstability measurements, while ExpertSAM™ measures each of these parameters simultaneously for all services and additionally measures the packet jitter.
To measure latency, RFC 2544 sends one probe packet per test, usually during the middle of the test duration. This will measure only the instantaneous delay without considering the variations in latency over the entire duration of the test. ExpertSAM™ measures Latency during the entire duration of the test. The Latency is measured on every frame transmitted and received, thereby capturing the variations in the delay over time. The Average, Minimum, Maximum and Instantaneous Latencies are reported.
- Frame Loss
RFC 2544 measures Frame Loss over a range of input rates. This is a separate test from the Throughput test, and Frame Loss is measured for a range of rates. ExpertSAM™ measures Frame Loss simultaneously along with the Throughput at the committed rate.
- Frame Delay Variation (FDV)
The variation of frames delay, also known as jitter is a critical parameter in packet networks. Jitter can severely affect the QoS provided, especially for voice and video services in real time. RFC 2544 does not measure jitter at all, whereasExpertSAM™ continuously measures jitter simultaneously along with other parameters at the CIR rate.