by Reza Vaez-Ghaemi, Ph.D.
Overview
For Ethernet to become carrier grade, a number of improvements must be developed and deployed. This paper introduces operation, administration, and maintenance (OAM), and examines some test applications for its qualification before deployment in the field.
Competition for market share among telecom, cable/multimedia service operator (MSO), and mobile operators has further fueled the need for network investments. When compared to voice-only service delivery, average revenue per user is expected to rise drastically with IPTV and higher-speed data services. Telecom operators are investing heavily in triple-play services, particularly IPTV, and related access networks capable of carrying increased data rates of typically 40-50 Mbps minimum, which will grow over time as newer high-bandwidth services are offered, such as three-dimensional video.
The Ethernet community has been continuously adding capabilities to native Ethernet to deliver a carrier-grade transport alternative to legacy TDM technologies. Provider Backbone Bridge/Provider Backbone Transport (PBB/PBT); Transport Multiprotocol Label Switching/Multiprotocol Label Switching Transport Profile (T-MPLS/MPLS-TP); Ethernet operation, administration, and maintenance (OAM); Synchronous Ethernet; and TDM over IP (TDMoIP) are some examples of major initiatives for carrier-grade Ethernet. These technologies require new methods and tools for qualification in labs and in the field.
Carrier Ethernet networks offer transport for millions of end users and connectivity among final users as well as among client networks. Native Ethernet was designed with the target of servicing much smaller groups of users in campuses and enterprises, but carrier-grade Ethernet must deliver scalability similar to conventional carrier networks (see Figure 1).
The simultaneous use of services by a large number of subscribers and applications leads to resource conflicts and congestion. Some applications are more critical and, therefore, are delivered at premium prices. Carrier Ethernet must provide service differentiation so that those critical applications receive the necessary quality.
Another attribute related to these critical applications revolves around protection and restoration. In the event of hard failures, users expect protection schemes that can deliver performance at or below 50 msec. Also in the event of failures, rapid fault identification and localization are essential in large networks, driving the demand for carrier-grade OAM functions.
Finally, while packet traffic represents the fastest-growing segment in user traffic, the demand for TDM traffic remains robust. Thus, any carrier-grade technology must transport TDM-based traffic at performance levels similar to circuit-based technologies.
Ethernet OAM provides a key challenge that must be overcome before Ethernet is ready for mass marketing in carrier networks. However, standards exist to address most of the issues. The new standards will provide new OAM capability to the customer premises demarcation point, potentially reducing operating expenses (opex) by more than half. Low Ethernet capital expenditures (capex) shift the profitability focus to opex, and Ethernet OAM is key to managing Ethernet service opex.
Ethernet OAM functions are provided in various layers and segments of the network, as Figure 2 shows:
- Access/link layer OAM (802.3ah)
- End-to-end connectivity fault management (CFM/802.1ag)
- Service layer (Y.1731)
- Two-port MAC relay (TPMR/802.1aj).
The IEEE 802.1aj TPMR standard defines protocols for interactions between provider bridges and remote two-port relay devices that might be used for demarcation.
The Ethernet in the First Mile (EFM) 802.3ah standard provides link-layer OAM functionality in the first/last mile. It is media independent and operates at a slow rate of 10 frames per second. Ethernet OAM packet data units (OAMPDUs) only work in point-to-point full-duplex networks and are not forwarded by peer devices. They require minimal configuration and deliver the following functions:
- device discovery
- remote failure indication
- remote loopback
- link monitoring.
During the network initialization, adjacent devices exchange identification information and OAM capabilities. With remote failure indication, network devices can notify peer devices in the event of failures. The remote loopback is a link-layer mechanism that operates at the frame level. Link monitoring delivers event notifications, such as status and diagnostics information, that are stored in local management information bases (MIBs), where peers can pull them.
The Connectivity Fault Management (CFM) standard specifies protocols and protocol entities within the architecture of VLAN-aware bridges that enable the detection, verification, and isolation of connectivity failures in virtual bridged LANs (VBLANs). These capabilities can be used in networks operated by multiple independent organizations, each with restricted management access to each other's equipment.
This standard specifies protocols, procedures, and managed objects in support of connectivity fault management. These enable discovery and verification of the path through bridges and LANs taken from frames addressed to and from specified network users. They also enable detection and isolation of a connectivity fault to a specific bridge or LAN. Some of the more important functions are described below.
Continuity check: Figure 3 illustrates that continuity check message (CCM) verification is among the most important CFM tests because it:
- checks the CCM interval rate
- compares the received CCM interval rate against the expected CCM interval rate
- checks for missing messages
- when emulating an Ethernet device, the tester collects the CCM and, based on this message, can then calculate the CCM interval/rate.
If the tester misses three CCMs, it declares loss of continuity (LoC), which is cleared when the tester detects two consecutive CCMs. LoC can be used to issue traps to the management system, to update the alarm log, and optionally to initiate a switchover to a protection link.
Figure 4 shows that loopback tests are necessary when conducting connectivity and diagnostic tests. The latter includes verifying bandwidth throughput or detecting bit errors. For example, the user can send a loopback message (LBM) as a single event or repetitively. The two types of loopback categories are unicast and multicast.
Loopback tests can be performed both in service and out of service. During in-service tests, the loopback test is performed in the presence of user traffic. The in-service loopback only applies to a configured Ethernet virtual circuit (EVC). For instance, the LBM causes only the configured EVC to loop back at the demarcation device, while the other EVC/VLAN continues to pass through the demarcation device.
One application for the LBM is providing connectivity to a remote demarcation device. A tester can be used to send an LBM to the demarcation device. The demarcation device then provides a loopback response (LBR) within a specified period of time. If no response is received within that period of time, it declares a LoC.
The link trace traces the path to a target MAC address once a link trace message (LTM) initiates the action.
The Y.1731 International Telecom-
munication Union-Telecom (ITU-T) recommendation was developed in cooperation with the IEEE 802.1ag, which specifies mechanisms required to operate and maintain the network and service aspects of the Ethernet layer. The recommendation also specifies the Ethernet OAM frame formats as well as the syntax and semantics of OAM frame fields.
Y.1731 defines two sets of functions for fault management and performance monitoring. The fault management functions include many components described in the previous CFM section, such as continuity check (ETH-CC), loopback (ETH-LB), and link trace (ETH-LT).
The performance management functions include:
- Frame loss (ETH-LM): Used to collect counter values applicable for ingress and egress service frames, where the counters tally the number of transmitted and received data frames between a pair of MEPs.
- Frame delay (ETH-DM): Used for on-demand OAM to measure one- and two-way frame delay as well as frame delay variations.
1. IEEE 802.3ah. Part 3: Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Access Method and Physical Layer Specifications Amendment: Media Access Control Parameters, Physical Layers, and Management Parameters for Subscriber Access Networks
2. IEEE P802.1aj/Dx.x. Virtual Bridged Local Area Networks-Amendment 08: Two-Port Media Access Control (MAC) Relay
3. ITU Y.1731. OAM functions and mechanisms for Ethernet based network
4. MEF 17. Service OAM Requirements & Framework-Phase 1
5. MEF 21. Abstract Test Suite for UNI Type 2 Part 1 Link OAM.
Attributions: The descriptions of standards in this article are reprinted with permission from IEEE Std. 802.1ag 2007, Fault Management, Copyright 2007, by IEEE. The IEEE disclaims any responsibility or liability resulting from the placement and use in the described manner.
* From IEEE Std. 802.1ag 2007, Fault Management, Copyright 2007, by IEEE. All rights reserved.
Reza Vaez-Ghaemi is currently emerging markets, alliances, and technology research manager at JDSU, where he manages a product line while conducting research into new technologies. To receive a more detailed version of this article, please contact the author at [email protected].
LINKS TO MORE INFORMATION
Metro Ethernet Forum: The Technical Specifications of the Metro Ethernet Forum