Maintaining a secondary connection is useful, but redundancy alone does not guarantee continuity. Organizations also need mechanisms that detect degraded or failed paths and reroute traffic to an available connection without requiring manual intervention by administrators.
SD-WAN products address this requirement by continuously assessing WAN links and applying routing policies when conditions change. Depending on the architecture, redundancy may extend across circuits, appliances, gateways, data centers, or provider infrastructure. The following six products illustrate different approaches enterprises can evaluate when automatic failover and resilient connectivity are priorities.
1. Fortinet
Fortinet Secure SD-WAN combines software-defined networking with security and centralized management for distributed enterprise environments. Its architecture can use multiple WAN transports and health checks to determine whether network paths continue to meet established requirements.
Enterprises evaluating SD-WAN products with failover capabilities can consider how application-aware steering, link monitoring, centralized policies, and high-availability configurations fit their continuity requirements.
Redundancy can extend beyond having a second internet circuit. Fortinet documentation describes high-availability clustering and fast failover when failures are identified through devices, monitored links, routes, or health checks. Designs can also incorporate geographically redundant gateways and steer branches according to predefined service-level objectives.
Organizations should assess circuit diversity, appliance redundancy, application behavior during failover, security inspection, monitoring, and recovery after the preferred connection becomes available again. Testing these conditions before production deployment can reveal dependencies that are not apparent during normal operation.
2. Aryaka
Aryaka delivers SD-WAN through a managed network architecture that combines customer-edge connectivity with a global backbone. Its approach provides several potential layers of redundancy rather than concentrating resilience solely on individual branch circuits.
The company’s broader enterprise connectivity perspectives describe first-mile and last-mile redundancy, redundant customer-premises equipment, and redundancy within its points of presence.
This architecture may appeal to enterprises that want operational responsibility for portions of the WAN handled as a managed service. Organizations can consider whether redundancy needs to extend from individual branch connections through the provider’s wider network.
Evaluation should cover geographic reach, edge-device configurations, ISP diversity, backbone availability, service commitments, monitoring, and escalation procedures. Enterprises should also establish how failures are detected and which party is responsible for investigating problems at each layer.
3. Bigleaf Networks
Bigleaf Networks takes an internet-centric approach to SD-WAN, using multiple available connections and cloud-based routing intelligence to maintain application connectivity. Its architecture monitors circuit conditions and adjusts how traffic uses available links when performance or availability changes.
A discussion of wider network resilience practices explains its active-active circuit model and same-IP failover. The provider also describes a high-availability configuration using paired routers that can switch when the active device encounters a problem.
Maintaining addressing through a circuit transition can matter for applications that depend on persistent sessions. Enterprises should therefore evaluate more than whether traffic eventually reaches a backup connection. The effect of failover on active voice, video, SaaS, and transactional sessions can be equally important.
Buyers should examine supported circuit types, health monitoring, failover behavior, hardware redundancy, application handling, and centralized visibility.
4. Ericsson Enterprise Wireless Solutions
Ericsson Enterprise Wireless Solutions incorporates cellular connectivity into branch and SD-WAN architectures, making wireless links an option for primary, secondary, or additional network paths. This can provide physical path diversity when wired circuits share infrastructure that could fail simultaneously.
Its broader connectivity industry research describes automatic switching from a primary WAN source to a secondary connection and explains how LTE or 5G can provide an alternative path when wired connectivity is unavailable.
Wireless redundancy can be relevant to branches, stores, temporary locations, and other sites where obtaining two physically diverse wired connections is impractical. Additional cellular carriers can further diversify connectivity where business requirements justify it.
Enterprises should evaluate coverage, signal conditions, data plans, link prioritization, application steering, hardware redundancy, and failback behavior. They should also test how quickly applications recover when traffic transitions between connection types.
5. FatPipe
FatPipe provides SD-WAN technology designed around multiple WAN connections, traffic management, and continuity when individual links deteriorate or become unavailable. Its approach monitors connection characteristics and can move traffic according to configured conditions.
The company’s general WAN reliability perspectives discuss active-active connectivity and stateful session failover across heterogeneous links such as dedicated internet access, MPLS, cellular, and satellite connections.
This approach illustrates an important distinction between basic backup connectivity and application continuity. A secondary circuit may be operational, but enterprises also need to determine what happens to sessions established before a failure.
Organizations should compare link monitoring, failover thresholds, session behavior, transport support, load balancing, centralized management, and recovery processes. Different applications may also require different tolerances for latency, packet loss, or brief interruptions.
6. Open Systems
Open Systems delivers managed networking and security services that include SD-WAN for distributed enterprise environments. Its approach uses traffic steering across different connection types and can respond when an existing path no longer meets defined performance conditions.
Its wider networking architecture commentary explains how policy-based traffic steering can react to measured availability and performance. One example describes moving application traffic away from a connection when conditions such as latency, jitter, or packet loss deteriorate.
This capability is relevant because WAN problems are not limited to complete outages. A circuit may technically remain connected while performing too poorly for voice, video, or other latency-sensitive applications.
Enterprises should examine link-health measurements, application policies, transport diversity, managed-service responsibilities, monitoring, geographic coverage, and escalation processes. They should also determine how policies distinguish complete failures from performance degradation.
What to Evaluate in SD-WAN Failover
Automatic failover should be evaluated as part of a broader resilience strategy. Enterprises need to identify potential single points of failure across circuits, edge appliances, power supplies, providers, gateways, and upstream infrastructure.
Modern enterprise network architecture guidance addresses the changing enterprise network landscape, including geographically distributed resources, evolving WAN infrastructure, cloud access, and the security considerations associated with modern connectivity.
Link diversity deserves particular attention. Two circuits do not provide meaningful physical redundancy if both depend on the same local infrastructure. Enterprises may therefore combine different carriers or transport technologies to reduce correlated failures.
Hardware resilience is another consideration. Multiple WAN circuits cannot maintain connectivity if the single appliance handling them becomes unavailable. Depending on business requirements, organizations may need active-active or active-standby edge devices in addition to redundant network paths.
Established router redundancy protocol standards illustrate the underlying principle of eliminating a single default-router failure point by dynamically transferring forwarding responsibility when an active router becomes unavailable.
Failback should receive the same scrutiny as failover. Returning traffic to a preferred connection too quickly can create instability if a circuit repeatedly transitions between healthy and unhealthy states. Monitoring thresholds and recovery policies should therefore reflect actual application requirements.
Enterprises should ultimately test link failure, device failure, degraded performance, recovery, and active-session behavior. A resilient SD-WAN deployment depends not simply on having backup resources, but on detecting problems reliably and moving traffic in a way that maintains acceptable application availability.
FAQs
What is automatic failover in SD-WAN?
Automatic failover detects an unavailable or degraded WAN path and redirects affected traffic to another suitable connection according to configured policies.
Is a backup internet circuit enough for SD-WAN redundancy?
Not necessarily. Enterprises may also need diverse carriers, redundant edge appliances, resilient gateways, and policies that protect active application sessions during failures.
What should enterprises test before deploying SD-WAN failover?
Organizations should test circuit outages, degraded links, appliance failures, application sessions, failback behavior, monitoring thresholds, and recovery across their expected network conditions.
