Skip navigation
Digitally composite image of technological interface against illuminated buildings in city in 3d Sean Prior / Alamy Stock Photo

Why We Need a New Network Architecture to Connect Data Centers in the Cloud

Author Rodney Dellinger outlines a compelling argument for a new network architecture to connecting data centers in the cloud.

As enterprises accelerate their digital transformation efforts, they are increasingly turning to cloud applications and services running on servers located in off-premises data centers, including public cloud, hosted private cloud and multi-tenant colocation data centers. 

To provide the sophisticated connection and interconnection capabilities required, the many participants in this evolving cloud ecosystem need a new network architecture, one that not only interconnects edge and core cloud data centers but also gives enterprises more control over how they connect to cloud services and deploy new applications. Using cloud data centers with rich interconnection ecosystems provides several important benefits for enterprises, including improved security, reduced transport costs, higher performance, lower latency and increased flexibility.

DCI Architecture Byline Fig 1.jpg

Figure 1. The many participants in the cloud ecosystem require a new network architecture
 

Equinix’s most recent Global Interconnection Index  and TeleGeography’s Internet Map forecast a compound annual growth rate of 44% for global data center interconnection and predict that interconnection bandwidth will be 15x larger than internet bandwidth by 2024. For network and data center operators looking to capitalize on this opportunity, three fundamental networking capabilities deserve careful consideration: data center connection, interconnection and automation.

Connecting Data Centers to Maximize Fiber Capacity at the Lowest Cost Per Bit 

Cloud applications require vast amounts of data to be transported between data centers, all of which must be replicated using synchronous and asynchronous backup between primary, secondary and tertiary data centers. This applies for both on- and off-premises business continuity and disaster recovery, and often between regions.

For operators, achieving this can be challenging. To do so, they need to plan for greater network resiliency through regional and international route diversity and deploy agile, scalable and resilient optical mesh networks and point-to-point DCI solutions that maximize fiber capacity and reach the lowest cost per bit.

Interconnecting Data Centers in the Cloud Ecosystem

The ability to interconnect data centers with multiple service providers, cloud providers and internet providers is equally important. This involves routing traffic between multiple networks, and connecting customers with cloud services, applications and workloads wherever they are located. 

DELLINGER Rodney Head Shot 2020 Color.jpeg

Rodney Dellinger, CTO and Head of Architecture of Nokia’s Webscale Business

Routing traffic from one network to another and between different operators’ data centers is known as peering. Peering enables two networks to connect and exchange traffic directly without having to pay a third party to carry traffic across the internet. In some cases, it involves multiple networks, each peering with the other to pass traffic from the source to the destination; for example, from a customer office to an application running in a core cloud data center.

To meet these interconnection requirements, implementing secure peering is a must for network operators, as is deploying data center gateway and IP routing solutions that offer comprehensive IP edge routing features such as segment routing, traffic engineering and network-based DDoS mitigation.

Flexible and Resilient Data Center Interconnection through Automation

Automation of data center connection and interconnection is the third critical requirement for enabling a dynamic cloud ecosystem. Today’s networks are static, multi-layered and costly, and need to be more integrated and automated to offer flexible and resilient connectivity between edge and core data centers. Automation of data center connection and interconnection needs to be optimized and efficient, with the ability to use multiple infrastructures, such as illuminated or dark fiber, managed wavelengths and multi-technology IP/optical networks. 

The benefits of network automation for customers are significant. It gives them the control to bring up applications, access services and spin up workloads when needed to serve their needs in any location, and they don’t have to contact the network operator and ask for more capacity in a specific data center or between data centers. 

By integrating customer web portals with network interconnection platforms, operators can automate the configuration, provisioning, and deployment of network resources to make the whole process as seamless as possible. 

Operators have a significant role to play in creating a highly interconnected data center infrastructure to deliver the next generation of cloud applications. By offering advanced data center interconnection capabilities, built on a foundation of network automation, they can provide exceptional experiences to their customers while enabling digital transformation and the digital economy.


Rodney Dellinger is the CTO and Head of Architecture for Nokia’s Webscale Business and has spent 24 years in the telecom industry in various roles ranging from network architecture and software engineering to leading business development and sales engineering teams.

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish