Understanding Metro Ethernet Principles For Data Center Networking

Data Center Networking

Today, users are geographically dispersed across large metropolitan areas in the United States. This trend has dictated how networks are managed. That is why it is common for network carriers to deal with metro environments. Oftentimes, connectivity is provided across an array of edge links. These links are interconnected via a high-capacity and high-speed metro area network (MAN). This network typically uses optical multiplexing technology, e.g. Dense Wavelength Division Multiplexing (DWDM). The goal is to pull traffic in, make distributed resources accessible via some transport or distribution layer. Now that you know that this is common in carrier environments, can the same metro ethernet principles be applied in data centers?

The Relationship Between Data Centers (Host) and Service Providers

Data centers for as-a-service players (*aaS) and hosting providers share some similar underlying requirements. When a carrier has to deal with connectivity to end users, a hosting provider has to service either applications or tenants. End-user profiles do no matter in these situations. There is still a need to provide connectivity and transport traffic between resources.

In a typical data center colocation environment, customers often place and store their servers within a data center. These servers often require high-speed access to Internet providers. Hosting providers are constantly figuring out how to pull traffic from these servers. Additionally, they need to develop solutions that can transport resources between carriers and data centers whilst providing uninterrupted Internet connectivity.

Although MPLS gateways and leaf-spine fabrics can facilitate traffic between servers as well as out of the data center, you still need to be mindful of resources that are not in close proximity.

Wide Area Networks Add Burden

Some hosting environments feature multiple data center sites. This is a non-trivial architecture. To successfully transport traffic across the wide area network and between data centers, the traffic transportation method has to be carefully engineered. In these situations, turning up new services and servers become complex. Additionally, the operational overhead increases and the time and cost ultimately fall on the end users’ shoulders.

Metro Ethernet Deployment Benefits

  • Business continuity

By following metro ethernet best practices, providers can achieve a distributed infrastructure that overcomes physical constraints, e.g. real estate, cooling, power, etc. This way, an issue in one location does not affect the entire infrastructure.

  • Enhance performance

A physically separate infrastructure can cause performance issues. Metro ethernet environments ensure high-speed transport and reliable end-user connectivity. Providers no longer need to stick to traditional optical connections and access ports.

  • Affordability

Building high-capacity optical interconnects can come with an exorbitant price tag. However, metro ethernet technology has allowed people to manufacture more advanced interconnects that share the same optical capabilities, at a more affordable price! They are already powering large carrier networks and used in a wide range of data centers.

  • Establish high-quality links between resource islands

Hosting providers can now house servers across the data center with the aid of dedicated wavelengths. This allows them to provide low-latency, high-capacity, and isolated links between resource islands.

  • Leveraging Layer 2 domain

With the ability to stretch a Layer 2 domain across an infrastructure (e.g. thousands of kilometers), providers can turn up applications and even move them without touching the gateway that connects these sites. Customers can enjoy greater flexibility at managing their networking services and applications.

To sign things off, metro carriers have successfully utilized this technology for years. Leveraging this technology is possible without breaking the bank.