Metro Ethernet: How Providers Link Customer Sites

Metro Ethernet is a service that uses Ethernet standards to connect a customer’s multiple sites across a metropolitan area, acting as a single Layer 2 network. Providers link these sites by building a fiber or copper network around the city and using Carrier Ethernet technologies to create virtual point-to-point or multipoint connections between each customer location. This article explains the technology, the different service types, how to choose one, and common issues you may face.

Background: From LAN to MAN

Ethernet was originally a Local Area Network (LAN) technology, designed to connect computers within a single building or campus. It is simple, cheap, and well understood by IT staff. Metro Ethernet extends the same Ethernet frames and switching logic across a Metropolitan Area Network (MAN), which typically covers a city or a large campus. The key idea is to make the provider’s network look like a giant Ethernet switch to the customer. This means that the customer’s existing routers, switches, and firewalls do not need major changes; they just see a standard Ethernet interface, often at speeds of 1 Gbps, 10 Gbps, or even 100 Gbps.

The Metro Ethernet Forum (MEF) standardized these services to ensure interoperability between different providers and equipment vendors. The MEF defines several service types, the most common being E-Line (point-to-point) and E-LAN (multipoint). These standards also define how the provider can offer Quality of Service (QoS), bandwidth profiles, and performance metrics like latency and frame loss.

How Metro Ethernet Works

When you order a Metro Ethernet service, the provider installs a device at each of your sites, often called a Network Interface Device (NID) or a customer premises equipment (CPE). This device connects to your router or switch using a standard Ethernet cable (e.g., RJ45 or SFP+). The provider then builds a virtual connection across their network between these NIDs. The core network may use MPLS (Multiprotocol Label Switching), VPLS (Virtual Private LAN Service), or even pure Ethernet switching with VLANs. From your perspective, the provider’s network is transparent: you can send any Ethernet frame, including VLAN-tagged frames, and it will appear at the other site with the same MAC addresses and VLAN tags.

There are two main service types:

• E-Line (Ethernet Private Line): A point-to-point connection between two sites. It behaves like a virtual cable. You can use it to connect two data centers, or a branch office to a head office.
• E-LAN (Ethernet LAN): A multipoint-to-multipoint connection that connects three or more sites in a full mesh. All sites can communicate with each other directly, as if they were on the same LAN switch.

Additionally, providers offer E-Tree, which is a rooted multipoint service where a root site can communicate with all leaf sites, but leaves cannot communicate with each other. This is useful for hub-and-spoke topologies.

Service Attributes and Bandwidth Profiles

Metro Ethernet services are defined by several attributes. The most important are the bandwidth profile and the Class of Service (CoS). The bandwidth profile specifies the Committed Information Rate (CIR) and the Excess Information Rate (EIR). The CIR is the guaranteed bandwidth, while the EIR is the additional bandwidth that can be used if the network is not congested. Frames within the CIR are delivered with high priority, while EIR frames may be dropped under congestion.

You can also choose the number of VLANs supported. A port-based service passes all VLANs transparently, while a VLAN-based service only passes specific VLANs. Some providers offer a service that supports QinQ (VLAN stacking) for more complex network designs.

Latency and frame loss are also specified. A typical metro Ethernet service has a latency of less than 5 milliseconds within a city, and frame loss of less than 0.01% for CIR traffic. These attributes are important for real-time applications like VoIP or video conferencing.

How to Choose a Metro Ethernet Service

When selecting a Metro Ethernet provider, consider the following factors:

1. Geographic coverage: Does the provider have fiber in the buildings you need to connect? If not, they may need to build a new fiber extension, which can take time and cost extra.
2. Service type: Do you need point-to-point (E-Line) or multipoint (E-LAN)? If you have more than two sites, E-LAN is often more cost-effective than multiple E-Lines.
3. Bandwidth: What is the required CIR? Consider both current needs and future growth. Some providers allow you to increase bandwidth easily.
4. Quality of Service: If you have real-time traffic, ensure the provider offers CoS with low latency and jitter guarantees.
5. Redundancy: Does the provider offer a redundant path? This could be a diverse fiber route or a backup link. This is critical for business continuity.
6. Cost: Obtain quotes from multiple providers. Be aware of installation fees, monthly recurring charges, and any contract terms.

Common Problems and Troubleshooting

Metro Ethernet is generally reliable, but issues can occur. The most common problem is a mismatch in VLAN configuration between your equipment and the provider’s. For example, if the provider expects a specific VLAN tag and your switch sends untagged frames, the link may not work. Always verify the VLAN settings on both ends.

Another issue is frame size. Standard Ethernet MTU is 1500 bytes, but some services support jumbo frames up to 9000 bytes. If one site sends jumbo frames and the other does not, packets will be dropped. Ensure that all devices on the service use the same MTU.

Latency and packet loss can also be a problem, especially if the provider has oversubscribed their network. Use ping and traceroute to test the link. If you see high latency or loss, contact the provider with the test results. They may need to adjust the bandwidth profile or investigate their core network.

Metro Ethernet vs. Other WAN Technologies

Metro Ethernet competes with other WAN technologies like MPLS VPN, dark fiber, and broadband internet with VPN. Compared to MPLS VPN, Metro Ethernet is simpler and often cheaper for a metro area, but MPLS VPN can be more scalable for national or global networks. Dark fiber gives you full control but requires you to manage your own equipment and redundancy. Broadband internet with a VPN is cheaper but offers no quality of service guarantees and has variable latency.

Metro Ethernet is often the best choice when you need high bandwidth, low latency, and a simple Layer 2 connection between sites within the same city. It is also a good option for connecting to a data center where you need to extend your LAN to the cloud.

Costs and Pricing Models

The cost of Metro Ethernet varies widely depending on the provider, the distance between sites, the bandwidth, and the contract term. Typically, you pay a monthly recurring charge based on the bandwidth (e.g., per 1 Gbps). There is often a one-time installation fee, which can be significant if the provider needs to build new fiber. Some providers offer a lower monthly rate if you sign a 3-year contract. Be sure to ask about any additional charges for maintenance, static IP addresses, or VLAN support.

To estimate costs, you can use an online calculator that compares prices from different providers, but it is best to get a formal quote. Remember that the cheapest option may not provide the reliability or quality of service you need.

Frequently Asked Questions

What is the difference between Metro Ethernet and Ethernet over MPLS?

Metro Ethernet is a service that uses Ethernet as the interface, while the underlying transport can be MPLS, VPLS, or even dark fiber. Ethernet over MPLS specifically uses MPLS in the core to create the virtual connections. In practice, many Metro Ethernet services use MPLS in the backbone, so the terms are often used interchangeably. The key is that you get an Ethernet interface, not an MPLS interface.

Can I use Metro Ethernet for internet access?

Yes, you can. Many providers offer a service called Ethernet Internet Access, which is a Metro Ethernet connection to the provider’s internet gateway. This gives you a dedicated, symmetrical bandwidth connection to the internet, with a public IP address. It is often more expensive than cable or DSL but offers better performance and reliability.

Do I need a router at each site?

Yes, you need a router or a Layer 3 switch at each site to route between your LAN and the Metro Ethernet network. The Metro Ethernet service is Layer 2, so it does not route IP packets. Your router will handle IP addressing, routing protocols, and security.

What is the maximum distance for Metro Ethernet?

Metro Ethernet is designed for metropolitan areas, typically up to 50-100 km in diameter. However, some providers can extend the service across longer distances using optical transport technologies like Dense Wavelength Division Multiplexing (DWDM). For longer distances, you would typically use a WAN service like MPLS VPN or dedicated internet.

How do I test if my Metro Ethernet link is performing as promised?

You can use tools like iperf to measure throughput, ping to test latency and loss, and manual checks for jitter. Compare the results to the Service Level Agreement (SLA) from your provider. If the provider offers a performance dashboard, use it to monitor the link. For a formal test, you can request a third-party test or use a certified testing tool.

Related

• Manhattan in Control Resonant: Charge Central Checkpoint
• ESP32 Dual Core Scheduling: How to Use Both Cores
• Beats Solo 4 Controls: Buttons & Mic Guide