Why Migrate to Ethernet Services

In order to support applications that require higher network performance and scalability, CIOs and IT Directors should migrate away from legacy networks to next generation networks, like WAN Ethernet. Ethernet services are:

- High performance
- Easy to manage
- Cost-effective

 

The Problem: Legacy Network Constraints

The problem with corporate WANs is that old technologies, like frame relay, ATM and private lines, were not designed to handle the typical applications that run across corporate networks. Furthermore, end users expect their applications to be as responsive across the WAN as they are across the LAN.

 

Frame relay, was designed to handle up to 45 Mps (DS3) of bandwidth. In reality, the vast majority of frame relay ports are 1.5 Mbs (T1) and use a fraction of that bandwidth between offices. Permanent Virtual Circuits (PVCs) must be configured between each site to form a full mesh. Most frame relay networks are designed as hub and spoke as a more cost-effective way to configure a network. In short, frame relay networks lack the scalability and flexibility to support the modern enterprise.

 

ATM (asynchronous transfer mode, not the cash machine) is a protocol designed to provide higher reliability and performance than frame relay. It is also used in corporate networks where bandwidth needs exceed a DS3 or 45 Mbps. Similar to frame relay, ATM networks must be configured using virtual circuits between each site, complicating network design. Also, since many companies want a full mesh (each office being able to connect to every other directly), many, many virtual circuits need to be configured.

 

In fact, to interconnect five offices using ATM or frame relay, you would have to configure (and pay for) ten virtual circuits. To connect 20 offices, you would need to configure 190 virtual circuits! The calculation is [n x (n-1)] / 2, where n is the number of sites.

 

Potential Solutions

Corporate WAN managers are migrating away from legacy networks to three types of networks:

- IP VPN
- IP MPLS VPN
- Ethernet VPLS VPNs

 

IP VPNs, while scalable to 155 Mbps (OC3), require encryption and the configuration of secure tunnels (similar to frame relay and ATM) between each site. These requirements make IP VPNs an unlikely option for mid to large-size enterprises.

 

Both IP MPLS services and Ethernet VPLS services rely on MPLS technology.

 

Multi-protocol Label Switching (MPLS) is an IETF standard that integrates Layer 2 information about network links into Layer 3 to simplify and improve IP packet exchange. Virtual Private LAN Service (VPLS) is an MPLS application that allows geographically dispersed sites to share an Ethernet broadcast domain. When VPLS is deployed over an MPLS platform, the result is a best-of-breed approach that provides the reliability, security and ease of use of Ethernet along with the scalability, multipoint-to-multipoint support, resiliency and QoS features of MPLS.

 

Ethernet services that use VPLS provide all of the benefits of an MPLS-based network: class of service, prioritization, resiliency, high performance, and any-to-any connectivity. Unlike IP MPLS services, Ethernet VPLS are easy to implement, easy to manage and scalable.

 

Learn more about IP MPLS versus Ethernet VPLS.

 

Or click here to learn how to migrate to Ethernet services.