Understanding the Network Underlay: An Unsung Hero of Enterprise Connectivity

After prioritizing overlay services for years, enterprises are now renewing interest in carrier-grade underlay services for connectivity.

Railway Track
Railway Track_Arelion

In recent years, network overlay solutions such as cloud applications and SD-WAN have taken the technology industry by storm as enterprises seek to enhance flexibility and scalability. But what about the underlying infrastructure these services depend on? After focusing on overlay services for years, many enterprises are showing a renewed interest in carrier-grade underlay services to meet their substantial connectivity requirements.  

Amid these developments, the network underlay is becoming the unsung hero of enterprise connectivity to enable cloud migration and more, with some enterprises purchasing connectivity directly from network operators. However, to unlock the network underlay’s full potential, enterprises must stay educated about the many options available due to progress made at the foundation of the network.  

The Many Flavors of Network Underlay: More Than Just MPLS 

Unfortunately, many companies have traditionally overlooked the network underlay. As a result, enterprises may mistakenly believe it primarily provides MPLS (multiprotocol label switching) services, a technique developed in the 1990s for efficient data packet routing. But, as we all know in the telecommunications industry, the 1990s were a long time ago, and network communications have evolved significantly since then. 

While MPLS has its place in certain use cases due to its high quality of service, underlay solutions such as optical and Ethernet services are increasingly popular to serve enterprises’ capacity and latency requirements. In addition, direct cloud connectivity based on optical transport services is useful for organizations handling immense amounts of critical traffic, including enterprises in financial services or manufacturing industries. However, many enterprises can also route their non-critical data over the public Internet for additional cost savings.  

For example, suppose you’re a multinational bank that needs to route critical data between numerous global data centers. In that case, you probably shouldn’t choose a border gateway protocol (BGP) solution that utilizes the public Internet. But, if much of your data is not subject to critical security concerns, BGP-based solutions may provide cost savings while meeting or exceeding your performance requirements. So, amid these options, how can enterprises choose the best underlay connectivity for their business needs once they assess their network requirements? 

3 Questions When Considering Enterprise Connectivity

  1. How critical is my data? 

While virtualized network services allow enterprises to store more of their data in off-premises infrastructure, enterprises should also consider how much non-sensitive data they can route over the public Internet without security or performance concerns. By evaluating their critical data sensitivity carefully, enterprises can get the best of both worlds and route data more cost-efficiently over the public Internet if criticality is less of a concern for their business needs. 

2. How reliable is my underlay? 

Reliability is a critical consideration for enterprises when choosing their connectivity option and is defined by several networking qualities. In an always-on business world, redundancy and diversity are crucial in preventing operational downtime and subsequent financial losses from customers frustrated with outages. Latency is also a chief consideration, with many network operators making progress in this area. For example, some transport services such as Ethernet now utilize segment routing based on the lowest-latency path available at any given time. Additionally, as enterprises increasingly integrate Artificial Intelligence (AI) applications, reliable, low-latency, high-capacity connectivity is essential to accommodate these applications’ real-time functionality and immense data processing requirements. 

3. How secure is my underlay? 

Amid continuing DDoS attacks and geopolitical tensions, network security remains paramount to minimizing downtime and ensuring operational efficiency. Due to these realities, enterprises should prioritize security on the network along with endpoint security. Automatic mitigation on the backbone network ensures enterprises stay connected to their global customers by scrubbing malicious traffic before it even reaches an enterprise’s network or IT systems. This consideration is critical for high-risk industries such as financial services or automotive manufacturing, where network interruptions due to malicious traffic can severely impact revenue and costs. 

One Size Does Not Fit All 

Enterprises can make informed decisions on connectivity options by understanding how each option serves (or does not serve) their business-critical networking requirements. For larger enterprises with critical data requirements, it may make sense to choose an optical transport service or direct cloud connectivity option that allows them to scale their infrastructure without sacrificing latency, security or capacity.  

On the other hand, maybe an enterprise is comfortable taking a hands-off approach to the network underlay, so a BGP-based routing solution utilizing the public Internet may suit its needs just fine. No matter an enterprise’s decision, staying informed on the benefits and drawbacks of each option ensures enterprises align their connectivity services to their company’s network requirements, positioning their business for continued growth as they scale their infrastructure. 

About the Author(s)

Mattias Fridström, Chief Evangelist, Arelion


Mattias Fridström has held numerous senior roles in Arelion, a global Internet carrier, including CTO. He has been Arelion’s Chief Evangelist since July 2016. 

Stay informed! Sign up to get expert advice and insight delivered direct to your inbox

You May Also Like

More Insights