Ethernet Access Services Definition and Implementation

Co-presented by: Arie Goldberg Ralph Santitoro Omnitron CEO and Founder MEF Board Member agoldberg@omnitron-systems.com Ralph Santitoro Director of Carrier Ethernet Solutions Ralph@Marcom-Services.net

Ethernet Service Delivery over Different Access Network Technologies 10/100BaseT 10/100BaseT Ethernet over Fiber (EoF) EoDSL Ethernet Service Provider Ethernet over SDH (EoS) 10/100BaseT Ethernet over PON/WDM Ethernet over HFC Ethernet over PDH (EoPDH) 10/100BaseT Coax 10/100BaseT Ethernet can be delivered over many different types of access network technologies

Fundamental Ethernet Service Components UNI I-NNI Access/Aggregation Network E-NNI E-NNI Metro Transport / Core Network Access/Aggregation Network EVC UNI (User-to-Network Interface) Physical interface/demarcation between service provider and subscriber Service start/end point NNI (Network-to-Network Interface) Demarcation/peering point between service providers (E-NNI) between service provider internal networks (I-NNI) Ethernet Virtual Connection (EVC) Logical representation of an Ethernet service as defined by the associate between 2 or more UNIs These Ethernet Service components are independent of the Access or Transport Network Technology to delivery the service

PDH (E1/T1) Access Network Provider MPLS Transport Network Example of Ethernet Service Delivery over Different Access/Transport Technologies Ethernet Service Provider UNI1 E-NNI I-NNI PDH (E1/T1) Access Network Provider E-NNI UNI2 MPLS Transport Network Fiber Access Network 10Mbps EVC PDH Access Network EoPDH: N bonded E1s MPLS Transport Network Ethernet over MPLS Fiber Access Network Ethernet over Fiber (EoF) PDH Access/Aggregation Network UNI1: PDH access network bonds 5 E1s for 10Mbps EoPDH E-NNI: 1Gbps Ethernet over Fiber (802.1ad / QinQ) MPLS Transport Network 802.1ad frames mapped to MPLS transport network I-NNI: 1Gbps Ethernet over Fiber (802.1ad / QinQ) Fiber Access Network UNI2: 10Mbps Ethernet over Fiber (EoF) Subscriber is unaware of the various transport network technologies used to deliver this 10Mbps Ethernet Service

Ethernet Services and Ethernet Access to IP Services Ethernet Services are Connectivity Services Multi-site E-LAN Ethernet Private LAN Transparent LAN Service VPLS Site-to-Site E-Line Ethernet Private Line Ethernet Virtual Private Line Ethernet Access to IP Services Provide Layer 2 connectivity to IP Service Networks Ethernet access to Internet Ethernet access to Managed VoIP service E-LAN E-Line Internet E-Line VoIP Ethernet Service: Layer 2 Service End-to-End Ethernet Access to IP Service: Layer 2 access to IP Service

Comparison of Ethernet L2 VPN and IP VPN Services in Access Network Ethernet L2 VPN Service UNI EVC (service) identified via VLAN ID CoS identified via PCP Forwarding on MAC or VLAN ID IP VPN Service Port Service identified via IP address CoS identified via DSCP Forwarding on IP Address Service Provider Service Attribute Ethernet Service IP Service Customer Handoff Ethernet UNI Ethernet port (or PDH circuit) Service Identification VLAN ID / EVC IP Address CoS Identification PCP DSCP/ToS Packet/Frame Routing/Forwarding MAC Address (E-LAN) VLAN ID (E-Line) Fault Management Link Trace, Continuity Check (Layer 2 Ping), Loopbacks Traceroute, ICMP Ping Performance Management Frame Delay, Frame Delay Variation, Frame Loss Ratio, Service Availability Packet Delay, Packet Delay Variation, Packet Loss, Fundamental difference between Ethernet and IP Services in the access network is the information used to identify, forward and process service frames

Ethernet Access Service Attributes Ethernet Service Performance Metrics, Bandwidth Profiles, CoS, Service Performance, and Fault Management

Ethernet Service Performance Management Key Metrics Frame/Packet Delay (Latency) Frame/Packet Delay Variation (Jitter) Frame/Packet Loss Ratio Service Availability Frame/Packet Goodput ITU-T has defined measurement framework/metrics for items 1-3 IP packet-based measurements ITU-T Y.1731 defines how to use 802.1ag to measure service performance MEF has defined measurement framework for items 1-4 Ethernet Frame-based measurements MEF 10.1 defines formulae for: Frame Delay (FD), FD Variation, Frame Loss Ratio, Service Availability The combination of IEEE 802.1ag, ITU-T Y.1731 and MEF 10.1 define Ethernet service performance

Ethernet Service Performance Metrics Frame Delay (FD) and Frame Delay Variation (FDV) Measured between UNI pairs via transmission and reception of a percentile of service OAM frames over a measurement period Example: FD: 30ms (99th percentile over 15 minute interval) FDV: 10ms (99the percentile over 15 minute interval) FD and FDV measurements can be One Way or Round Trip Frame Loss Ratio (FLR) Percentage of Green (in-profile for CIR) Ingress frames received at Egress UNI over a measurement period Example FLR: 2% ( over 15 minute interval) Service Availability Percentile Amount of time, FLR meets its service level objectives over a measurement period Example Availability: 99.9% (over 1 month interval)

Bandwidth Profiles (BWP) and CoS Bandwidth Profiles per EVC (service) and per CoS CIR (Committed Information Rate) CIR assured via Bandwidth Reservation and Traffic Engineering EIR (Excess Information Rate) EIR bandwidth is considered ‘excess’ Traffic dropped at congestion points in the network CBS/EBS (Committed/Excess Burst Size) Higher burst size results in improved performance EVC-2 EVC-1 EIR EIR CIR CIR EVC-3 EIR CIR CoS 6 1Mbps CIR for VoIP EVC1 BWPs can divide bandwidth per EVC (service) over a single UNI Multiple services over same port (UNI) CoS markings enable the network to determine the network QoS to provide CoS 2 10Mbps UNI (port) 6Mbps CIR for VPN data traffic UNI EVC2 3Mbps for Internet Access CIR defines the assured bandwidth EIR improves the network’s Goodput

Ethernet Service Connectivity Fault Management IEEE 802.1ag for EVC Connectivity Fault Management For Protection/Restoration Connectivity Check Messages (CCMs) for heartbeats For diagnostic purposes Connectivity Check (Layer 2 Ping), Link Trace, Loopbacks CCMs initiated between: Management Endpoints (MEPs / UNIs) Management Intermediate Points (MIPs / NNIs) Management Endpoints and Intermediate Points (UNI-NNI) Connectivity Fault Management provided by Carrier Ethernet network elements and NIDs

Ethernet Link Fault Management IEEE 802.3ah for Link Fault Management for (first mile) Ethernet access connection Link Fault Detection and Remote Loopbacks Layer 1 fault management used for transport connections used to deliver Ethernet services Ethernet over PDH Ethernet over SDH Ethernet over DSL Ethernet over HFC Coax Ethernet Link Fault Management performed by Carrier Ethernet network elements and NIDs

Ethernet Service Continuity Check UNI IP/MPLS Core Network UNI Metro Aggregation Network Metro Aggregation Network Access Network Access Network MEP#1 MEP #2 MIP MIP MIP MIP MIP Maintenance Association (MA) MIP MEP #1 CCMs sent every 10ms on working/protect paths Check for CCMs received from MEP #2 on working/protect paths MEP #1 No CCMs received from MEP #2 within 30ms (3 x 10ms) MEP #1 Report CC fault to management system MEP #1 Send CCMs with RDI flag set MEP #2 Remote alarm detected by received RDI Report alarm to management system MEP #2 Initiate protection switchover MEP #2 Report on CC fault if no CCMs are received from MEP #1 for 30ms MEP #2 CCMs sent every 10ms Check for CCMs received from MEP #1 Ethernet continuity check is analogous to IP’s Ping but does much more

Metro Aggregation Network Ethernet Link Trace UNI IP/MPLS Core Network UNI Metro Aggregation Network Metro Aggregation Network Access Network Access Network MEP MIP MIP MIP MIP MEP Link trace Reply Ethernet Link Trace is analogous to IP’s Traceroute

Some Benefits of Ethernet L2 VPN versus IP VPN L2 VPN Service Benefit Enterprise Benefits Service Provider Benefits No IP awareness required Especially important with IPv6 applications No service provider coordination of IP addresses No concerns of service provider’s IPv6 capabilities No need for service provider to have IPv6-capable routers Fault management Ethernet OAM versus IP, MPLS and Ethernet OAM Simpler OAM (lower OpEx) by managing a single network layer Perform just Ethernet OAM instead of Ethernet OAM + MPLS (OAM) + IP OAM for fault management Multi-site connectivity Simpler to add new location to an E-LAN service (auto learning) In many cases, simpler to deploy L2 multipoint services

Summary Ethernet Service are agnostic of the Access or Transport Network Technology Facilitates a common service portfolio for all markets Ethernet Service can be: A Layer 2 Connectivity Service (E-Line or E-LAN L2 VPN) A Layer 1 Private Line Replacement (EPL) A Layer 2 access connection to an IP Service Ethernet Access to Internet Managed IP VPN or VoIP service

Presentation available for download at: http://MetroEthernetForum