SD-WAN vs Traditional WAN: Key Differences Explained
This article compares SD-WAN (Software Defined WAN) and Traditional WAN, highlighting the key differences between the two technologies.
Introduction
As we know, traditional WANs typically consist of multiple single-function devices connected via various WAN links. This can lead to complexity in branch IT management, requiring a larger number of devices to be maintained. SD-WAN offers several benefits over traditional WANs, leading to its increasing adoption by companies already using traditional WAN infrastructure.
Traditional WAN
- In traditional WANs, each instance of the data plane contains its own control plane, making programming difficult.
- The traditional WAN architecture is static and private, inhibiting migration to dynamic and public cloud environments.
- Branch deployments can be complex due to the dependency on various transport mediums like DSL, coaxial cable, fiber, and 4G LTE.
- Traditional WANs require separate security and management configurations for disparate traffic flows, increasing operational complexity and cost due to the need for skilled resources.
- Figure 1 depicts a block diagram of a traditional WAN system and its interface components.
Challenges with Traditional WAN
Traditional WAN solutions face the following challenges, which SD-WAN aims to address:
- Manual QoS Configuration: QoS parameters (e.g., bandwidth) are configured manually and don’t adapt to changes in link conditions.
- Manual Routing Configuration: Routing protocols are manually tuned to load balance traffic.
- Inefficient Resource Utilization: WAN circuits often remain in passive standby mode, used only during hard failover situations.
- Slow Restoration Times: Link failures can take several seconds to restore, leading to a poor end-user experience.
- Difficult Key Management: Traditional WANs use IPSec for authentication, requiring pre-shared keys that are difficult to manage across a large number of devices.
SD-WAN (Software Defined WAN)
SD-WAN separates the functionality of networking devices into a control plane and a data plane. The control plane handles signaling traffic and routing decisions, while the data plane carries application and user data. Unlike traditional WAN devices where each data plane instance has its own control plane, SD-WAN uses one logical control plane instance to serve multiple data plane instances (typically switches and routers). This separation simplifies the programming and control of various transport mediums. Figure 2 shows a block diagram of a typical SD-WAN system and its interface components.
SD-WAN is a technology that distributes traffic between branch offices and data center sites via the most appropriate transport mediums using Software Defined Networking (SDN) concepts, making it transport-agnostic.
Key Features of SD-WAN
- Centralized Control Plane: SD-WAN uses a centralized control plane architecture, managing a large and diverse set of data plane components.
- Open Flow Protocol: It uses standard Open Flow Protocol for communication between the control plane and data plane components. Viptela (acquired by Cisco) developed the Overlay Management Protocol (OMP) specifically for SD-WAN systems, eliminating the need for protocols like OSPF and BGP.
- Transport Independent, Secure and Reliable Network: SD-WAN delivers a transport-independent, secure, and reliable network.
- Standard Encryption: It uses standard encryption such as AES, providing secure connectivity over any transport medium. Some SD-WAN services also use IPsec to authenticate network traffic between branch offices and data center sites.
- Easy Migration: SD-WAN allows for easy migration from traditional WANs using different approaches like inline path, out of path, and thin branch. All these approaches use BGP to interface the legacy branch with the SD-WAN branch.
- User-Friendly Management Console: SD-WAN provides users with an easy-to-use management console to manage traffic, assign policies, and configure devices (switches, routers, firewalls, etc.).
- Simplified Network Management: SD-WAN simplifies network management by automating tasks such as deployments, configurations, and operations.
- Cost-Effectiveness and Flexibility: SD-WAN improves cost-effectiveness and flexibility by leveraging commercially available hardware and network devices or servers.
SD-WAN vs. Traditional WAN: Key Differences
The following table summarizes the major differences between SD-WAN and Traditional WAN:
Feature | Traditional WAN | SD-WAN |
---|---|---|
Control Plane | Each instance of the data plane contains its own control plane. There is no central controller, making programming and management difficult. | Each logical instance of the control plane serves multiple instances of the data plane. The centralized SD-WAN controller makes programming and management easy. |
Transport Agnostic | Not transport agnostic. | Transport agnostic. |
Overlay Networks | Uses overlay networks such as MPLS. | Uses overlay networks such as VXLAN, IPsec, GRE, DMVPN, etc. |
Performance | Delivers higher performance. | Delivers very high performance. |
Routing Protocols | Uses traditional routing protocols such as BGP, OSPF, etc. | Uses flow-based traffic management protocols such as OMP (Overlay Management Protocol). |
CPE Model | Uses a vertically integrated CPE model as hardware models. | Uses hardware models deployed on X86 virtualization. |
Bandwidth Cost | Offers expensive bandwidth-based solutions. | Offers low-cost WAN solutions. |
Infrastructure | Uses complex infrastructure. | Uses simple infrastructure. |
Security | Offers high security. | Offers very high security. |
Reliability | Offers WAN with good reliability. | Offers WAN with excellent reliability. |
Automation & Analytics | Automation and analytics features are not supported. | Automation and analytics features are available. |