In recent years, Software-Defined Wide Area Networking (SD-WAN) has created a buzz in the networking world. But what about MPLS (Multiprotocol Label Switching)? Has this routing technique been made redundant by the increasing uptake of SD-WAN? What are the differences and similarities between SD-WAN and MPLS? And are there valid reasons to use both?
What is SD-WAN?
Software-Defined Wide Area Networking (SD-WAN) is a software-defined approach toward bringing wide area networking to organizations seeking more diversity and control over their enterprise WAN. SD-WAN incorporates a traditional hardware-based networking model and adds a software-defined virtual network overlay on top. This overlay is managed and provisioned centrally by a controller, removing the need for device-by-device network configuration and management. The underlay, or data plane, is then left with the responsibility to process and transit packets between devices.
The overlay can run over a range of standard network transport services, including the public internet, 4G, 5G, and MPLS. Based on the performance of the underlying network transport, application-aware routing will control where and when an application uses a specific service to maintain the performance of real-time and sensitive applications.
What is MPLS?
MPLS is a label-switched-path network model in which data packets take a pre-defined, private route straight to their destination from the provisioning of Layer 2 Ethernet or Layer 3 Virtual Private Networks (VPNs). These label-switched paths can be statically defined to direct traffic around congested parts of a network on an end-to-end, low-latency route.
MPLS services are segregated from the internet and from other MPLS services on the carrier’s network, and can instead be viewed as dedicated services bound by Service Level Agreements (SLAs) for packet loss, jitter, and latency standards.
Benefits of SD-WAN
SD-WAN is the newer, and arguably more popular, player in town. Cisco, Fortinet, and VMware are just some of the networking giants that now offer SD-WAN solutions to simplify increasingly remote and complex enterprise networks. SD-WAN provides its users with a range of benefits:
Centralized management system
Most SD-WAN solutions by default provide a centralized management system with automation, security, and application-level visibility built in. This removes the need to integrate SD-WAN devices into another vendor’s network management system, saving significant time and cost.
Flexible private overlay
SD-WAN can build a private overlay over any network transport type, whether it be public internet, private MPLS, or a combination of both. By bundling different types of network transport within this flexible model, users can achieve higher bandwidth at a lower overall cost with improved security.
Zero Touch Provisioning (ZTP)
Zero Touch Provisioning or ZTP refers to the ability to provision a device or network without the need for local configuration. Some SD-WAN devices support ZTP, meaning they can be shipped to site directly from the factory without the need to apply configurations. When connected to the internet, a ZTP device will connect back to a centralized controller, receive verification against its serial number, download its configuration, and join the overlay network — it’s that simple.
Benefits of MPLS
MPLS is the legacy technology in this comparison, and while its higher price tag and lower flexibility places it at a comparative disadvantage against SD-WAN, it shouldn’t be dismissed. MPLS has its own set of benefits:
MPLS can be used as a dedicated private service with SLAs for throughput, latency, and jitter. End-to-end connectivity across an MPLS network is assured by the network operator to perform within the bounds of this SLA, ensuring reliability and consistency.
Proactive network management
MPLS network capacity is proactively managed, and end-to-end latency is closely monitored by the provider to ensure paths are clear of congestion and faults are addressed as soon as they occur. This attentive approach to network management minimizes downtime and maximizes productivity for your enterprise.
WAN designs with static and predictable requirements can be easily deployed with MPLS; an MPLS-enabled site with a simple redundant design of two MPLS links is easy to both implement and support.
The similarities between SD-WAN and MPLS
While SD-WAN and MPLS have some key differences, they also share some important similarities. SD-WAN and MPLS both:
- deliver high-performance, reliable, and private WAN.
- provide a type of private overlay — SD-WAN with the use of IPSec VPNs; MPLS with labels.
- use private IPv4 addressing to communicate between devices connected to the WAN within their private overlays.
- support classification of traffic into different priority and importance levels.
SD-WAN and MPLS are often treated as separate network models, but as our comparison shows, they aren’t directly competing configurations; you don’t have to face an ultimatum between which model is right for you.
If you need to streamline connectivity between multiple endpoints and cloud providers, SD-WAN is a proven, cost-effective solution. If consistency, reliability, and simplicity is what you’re after, MPLS alone may be your preferred option. But if you have larger or more complex networking requirements, you can actually use MPLS together with SD-WAN to build a hybrid WAN design. The application-aware routing benefit of SD-WAN can ensure critical traffic like Voice over Internet Protocol (VoIP) is directed over your reliable MPLS transport, with non-critical traffic directed over internet transport. Using MPLS and SD-WAN together is a great way to lay the foundation for your organization as cloud connectivity continues to grow and evolve in the years to come.