Network redundancy vs. resiliency: What's the difference?

What is network redundancy?

Network redundancy is the use of redundant physical or virtual hardware and/or interconnections. A common example of this is to deploy a pair of network firewalls with duplicated cabling connecting to the inside, outside and demilitarized zone networks.

In many cases, redundant firewalls are configured in active and standby modes. This means a single firewall device handles all network routing and security policy enforcement while the secondary unit sits idle, waiting for the active firewall to fail.

The active firewall sends the standby firewall health status information every few seconds. If too many health checks are missed, the standby firewall assumes the active unit has failed in some capacity and takes over all duties.

Another network redundancy method is to place both firewalls in active states. With this architecture setup, the firewalls split routing and security policy enforcement tasks while continuing to check the health status of the other. If either firewall were to fail, the other would take on the full load of duties.

What is network resilience?

While network redundancy is an all-or-nothing approach to provide continuous network operation, network resilience is the recovery from smaller system faults on a single network device or between devices without needing duplicate hardware and software.

Software-defined WAN path selection is an example of adding resilience into a network infrastructure. As primary WAN links become congested, the SD-WAN controller has the intelligence to identify mission-critical traffic and reroute it over a less-congested path.

Thus, while both redundant links to a remote site may be fully operational, bottlenecks can occur that render one path less desired. This is where resilient intelligence can step in to maintain higher levels of network performance for traffic that is deemed most important to the business.

Network failover vs. self-recovery

Network redundancy focuses on duplicated components to achieve higher levels of network uptime. This is a more simplistic method -- although usually more costly and complex from a deployment standpoint.

Network resilience, on the other hand, uses built-in system intelligence to recover from faults that occur from time to time. In most cases, redundancy can be thought of as the ability to fail over, while resilient network functions focus on self-recovery methods.