Sie sind auf Seite 1von 4

ACCEDIAN

PACKET PERFORMANCE ASSURANCE TM

White Paper

NETWORKS

Deploying Ethernet OAM using NIDs

Strategies for Deploying Ethernet OAM


The Ethernet Service OAM capabilities defined in the IEEE 802.1ag & ITU-T Y.1731 standards have been designed to help service providers efficiently monitor and maintain carriergrade Ethernet services over multi-vendor and multi-carrier networks. When properly implemented at key points in the network, service OAM provides robust connectivity fault management (CFM) and performance monitoring (PM) capabilities for Ethernet services. By defining a framework of multiple maintenance domains and levels, service OAM allows providers to determine if faults lie within their own network or a customers or partner carriers network, providing to each party only the degree of detail necessary to restore or circumvent service issues.

Executive Summary
Establishing end-to-end Ethernet Operations, Administration and Maintenance (OAM) is a key part of providing Carrier Grade, SLA-backed services for demanding wireless backhaul and business services applications. This white paper explores strategies to introduce Ethernet OAM into existing networks using Network Interface Devices (NIDs), enabling the connectivity fault management and performance monitoring features of the IEEE 802.1ag and ITU-T Y.1731 OAM standards over multi-technology, multi-carrier and multi-vendor networks.

The Multi-Domain OAM Model


Service Demarc

Customer
Customer Site 1 Customer Site 2

Degrees of Compatibility & Performance


Multi-vendor and multi-platform interoperability is one the most common issues hindering functional, uniform OAM deployment. The various elements that constitute a typical network often have varying levels of Ethernet awareness, from legacy SONET/SDH transport elements and more

Provider
802.1ag / Y.1731

Operator 1

Operator 2

Operator 3

Ethernet OAM offers promising capabilities that enable providers to deliver large-scale, resilient Ethernet-based services but in practice, the deployment of OAM functionality can be complex or impractical depending on the approach taken. Certain implementations reduce the benefits of OAM performance monitoring with inconsistent measurements that can lead to misleading or false Quality of Service (QoS) alarms and incomplete troubleshooting functionality.

recent MSxPs, to more packet-savvy IP/MPLS routers and pure Carrier Ethernet access platforms. Each of these elements play different roles in the end-to-end transport and delivery of Ethernet services, with varying degrees of layer 2 processing and transparency. Where possible, many network element vendors are starting to introduce Ethernet OAM upgrades new software and firmware releases that implement selected OAM features by leveraging the elements network processors.

2008 Accedian Networks, Inc.

ACCEDIAN.com

ACCEDIAN
PACKET PERFORMANCE ASSURANCE TM

White Paper

NETWORKS

Deploying Ethernet OAM using NIDs

Cell Site

MSC
Wireless Backhaul

Core / Transport Network


Aggregation Node

Hand-Off Point

Metro / Access
Customer LAN

CO

Partner Carrier

Customer LAN

Typical Demarcation / Ethernet OAM Points:

But limits in processing capabilities and the increasing feature richness required to provision, switch and condition Ethernet traffic (such as shaping and rate limiting) place increasing demands processors, often leaving OAM functionality partly implemented. In addition to partial compliance, the constant evolution of Ethernet OAM standards leads to interoperability issues as different versions of 802.1ag / Y.1731 are deployed throughout the network.

Network elements typically only provide round trip latency and jitter measurements with a relatively coarse ~1ms resolution, insufficient when monitoring Ethernet services carrying delay-sensitive VoIP, IP-Video, transactional and high-performance 3G & 4G wireless backhaul traffic. These demanding applications normally require one-way latency in the 3-5ms range, and jitter performance between 1-3ms. To accurately monitor these services, delay measurements need to be at least 10X more precise than SLA-based specs: 0.1ms or better resolution (and accuracy) is required. The hardware-based performance monitoring capabilities available in NIDs provide one-way latency and jitter measurements with microsecond precision and accuracy on the order of 20 microseconds. These measurements are up to 1,000X more precision than network elements, while enabling the one-way delay test specified in the Y.1731 standard. The same processing limitations that affect delay

Deploying Ethernet OAM Using NIDs


A scaleable and uniform alternative to deploying OAM network-wide is to use Ethernet Network Interface Devices (NIDs) as OAM maintenance end and intermediate points (MEPs & MIPs). Many providers already have NIDs installed as demarcation and monitoring units at key maintenance locations throughout the network, typically customer and cell sites, aggregation and inter-carrier hand-offs points. NIDs address OAM performance and interoperability concerns by providing uniform, hardware-based OAM functionality independent of the elements installed in the network.

Performance & Accuracy


The software-based OAM implementations offered by network elements often lack the performance to conduct the precise measurements required to effectively monitor Service Level Agreement (SLAs) and real-time applications

measurements in network elements also affect frame-loss monitoring. Unlike in-line NIDs that provide Real Frame Loss (RFL) metrics, network processor-based designs can only provide synthetic frame loss estimates, several of orders of magnitude less accurate than RFL.

2008 Accedian Networks, Inc.

ACCEDIAN.com

ACCEDIAN
PACKET PERFORMANCE ASSURANCE TM

White Paper

NETWORKS

Deploying Ethernet OAM using NIDs

EVC2 EVC3

EVC1

Multiple providers & equipment vendors OAM Maintenance End-Point (MEP)

High-Capacity, Multi-Flow OAM Performance Monitoring

Density & Scalability


Network elements software-based OAM functionality offers limited capacity compared to a NIDs hardware-based implementation, able to monitor several dozen services, flows or SLAs simultaneously without adding any processing delay to the traffic. This level of scalability is required in many realworld Carrier Ethernet applications, including wireless backhaul (where up to 100 cell towers may home into a single mobile switching center), and large scale WAN connectivity (E-LANs with several service levels connecting a large number of sites over national or international geographies). Aside from performance monitoring restrictions, softwarebased OAM implementations also face limitations in fault verification and isolation, with little ability to process any reasonable stream of OAM loopback packets without impacting actual traffic. Extensions to Ethernet OAM standards also provide IP (layer 3) performance monitoring and connectivity fault management, while providing layer 2 transparency over routed / MPLS networks where standard Ethernet OAM is interrupted.

Complementary QoS-Enabling Features


In addition to OAM functionality, NIDs often include a wide range of complementary service creation and traffic conditioning capabilities that simplify Ethernet service deployment and optimize QoS. Integrated service mapping allow NIDs to replace access platforms in many applications, while hardware-based filtering, rate limiting and low-latency traffic shaping functionality conditions traffic at the edge to ensure critical, real-time applications are given the priority and bandwidth they require.

Advanced Functionality & Transparency


The silicon-based packet processing employed in latestgeneration NIDs enables service assurance capabilities that go beyond existing OAM standards, providing the additional measurements required to fully validate service performance and monitor SLAs. As an example, NIDs can provide inservice throughput testing, allowing an operator to confirm the committed or excess information rates (CIR, EIR) of a service while its in use, without affecting live traffic.

Combining these capabilities with OAM in a NID at the service demarcation point has many advantages, while also meeting a key deployment requirement: OAM must be applied at the traffic conditioning point to ensure performance monitoring results are valid.

2008 Accedian Networks, Inc.

ACCEDIAN.com

ACCEDIAN
PACKET PERFORMANCE ASSURANCE TM

White Paper

NETWORKS

Deploying Ethernet OAM using NIDs

Extending OAM Coverage End-to-End


In addition to providing comprehensive, precise OAM capabilities more readily than network elements, NIDs also extend OAM capabilities beyond access platforms or edge routers to provide visibility directly to customer and cell sites. By including this mostvulnerable last-mile in the coverage of 802.ag / Y.1731 OAM, providers gain true end-to-end service management and performance visibility. First mile issues can also be addressed by the IEEE 802.3ah standards, but lack of interactivity with 802.1ag / Y.1731 and limited functionality make this Ethernet in the first mile standard less desirable than extending the full capabilities of service OAM directly to the network edge. 802.3ah lacks performance monitoring, link fault propagation beyond the first mile, link trace & loopback response for fault verification & isolation, and layer 2, perflow loopback capabilities. CPE
Demarcation Point Demarcation Point 802.3ah or no OAM 802.3ah or no OAM

CPE

Providers Network
802.1ag / Y.1731 OAM

CPE

Providers Network
802.1ag / Y.1731 OAM

CPE

Extending Service OAM to the Demarc Point

Provisioning Ethernet OAM Today


Addressing the issues complicating the deployment of Ethernet OAM in multi-vendor, multi-technology and multi-carrier networks, Ethernet NIDs offer a cost-effective, scalable, hardware-based approach that strategically covers the network from endto-end, while also providing OAM information at aggregation and inter-carrier handoff points. When also used for advanced service assurance, traffic conditioning, multi-port aggregation and Ethernet service mapping, NIDs offer a wide range of functionality that simplifies the delivery of QoS-optimized, efficiently managed carrier-grade Ethernet services.

EtherNID & MetroNID Demarc Units


Accedian Networks packet assurance demarcation units feature the full functionality of the latest 802.1ag, Y.1731 and 802.3ah OAM standards, and are capable of acting as maintenance endpoints (MEPs) or intermediate points (MIPs) in any topology, including multi-vendor, multi-carrier and multi-technology networks. This capability allows service providers to establish and benefit from Ethernet OAM visibility without the need to upgrade access platforms or expensive edge routers to those supporting the latest OAM standards.
Test Set Support EVC Creation Shaping, Policing
METER

Demarc & Loopback

In-Service RFC-2544

Assure SLAs

Accedian Networks 4878 Levy St, suite 202 St-Laurent, QC, Canada, H4R 2P1 +1 (514) 331-6181, 1 (866) 685-8181

Accedian Networks Packet Performance Assurance solutions enable carrier-grade, packet-based wireless backhaul, business services & multi-carrier applications over wireless & wireline networks. 2008 Accedian Networks, Inc. Version: Sep 08

Accedian, EtherNID, MetroNID, SLA-meter, Performance Assurance Agent, PAA, Multi-SLA, Ether-PRO, EtherSHELF, Ethernet Service Assurance Platform (ESAP), EVC-Builder, Packet Performance Assurance, Delivering Service Performance with Assurance and Map-out the Health of your Network are trademarks of Accedian Networks Inc.

ACCEDIAN.com

Das könnte Ihnen auch gefallen