Having redundancy built into networks is essential for a business’s success. The majority of business-critical functions today, starting from real-time services to point-of-sale transactions take place over the wide-area network, shaping the lifeline of an organization.
SD-WAN has made the deployment of multiple ISP circuits easier so that the network can remain in service in case of circuit failures. The gateway at the edge remains a single point of failure yet. Building redundancy at the router level continues to be expensive and complex to manage, requiring dedicated IT staff. At Sproute, we have just rolled out a new feature, PAIR for a comprehensive, automated, and active-active router redundancy.
You can enable PAIR with a single-click from the central dashboard and by connecting two Sproute routers with a direct link as shown in the following example topology. The ISP circuits can be segmented across these routers for maximal availability.
PAIR automatically discovers the other router, synchronizes necessary data and makes sure all your policies work in unified fashion. These, for example, include traffic steering rules. If you had a policy for voice traffic to go through one of the links on one of the routers and bulk traffic to go through another uplink through the other router, PAIR ensures that traffic always goes through the correct uplink regardless of which router the hosts send the packets to.
The simplest approach is to run VRRP on the Sproute routers. This is again, a one-touch operation against the interfaces tab to enable VRRP. The routers negotiate master <-> slave relationship and failovers are seamless.
A more involved network may require OSPF to be enabled.
Sproute routers build point-to-point overlays with other configured sites over each possible path, i.e. across all reachable uplinks. The reachability here refers to paths that can be NAT-traversed (the routers perform automatic NAT traversal to punch holes through any set of NAT devices). In a PAIR setup, the overlays will thus include going via the peer router and its uplink to build tunnels. This has a particularly large advantage that traffic switchover on link failures is seamless.