| CPC H04L 47/125 (2013.01) [H04L 47/825 (2013.01); H04L 67/1004 (2013.01)] | 9 Claims |

|
1. A scalable network for routing information between a source and a destination, the scalable network comprising:
a first transit gateway cluster including at least two pairs of transit gateways, each transit gateway of the first transit gateway cluster comprising a processor configured to control routing of one or more messages received via at least a first communication link through a first plurality of communication links;
a first pair of spoke gateways, each spoke gateway of the first pair of spoke gateways comprising a gateway routing table configured to determine which peer to peer communication link of a plurality of peer to peer communication links to use when forwarding a message from a cloud application instance assigned to the spoke gateway to a first destination instance;
a second transit gateway cluster including at least two pairs of transit gateways, each transit gateway of the second transit gateway cluster comprising a processor configured to control routing of the one or more messages received from each of the first plurality of communication links to a second plurality of communication links for subsequent routing to the destination; and
a second pair of spoke gateways, each spoke gateway of the second pair of spoke gateways comprising a gateway routing table configured to determine which peer to peer communication link of a plurality of peer to peer communication links to use when forwarding a message from a cloud application instance assigned to the spoke gateway to a second destination instance,
wherein, in response to a failure of a peer to peer communication link of the plurality of peer to peer communication links of the first pair of spoke gateways, a spoke gateway of the first pair of spoke gateways that relies on the failed peer to peer communication link updates the gateway routing table autonomously by disabling a tunnel interface corresponding to the failed peer to peer communication link without reliance on activity by a controller that manages operability of the scalable network.
|