Systems and methods for distributing traffic across cluster nodes

摘要

The present application is directed towards systems and methods for distributing traffic across nodes of a cluster of intermediary devices through distributed flow distribution (DFD). Upon receipt of network traffic, a cluster node, such as an intermediary computing device or appliance, may internally steer a portion of the traffic via an inter-node communications backplane to one or more other nodes in the cluster so that the load is equally handled by all of the nodes in the cluster. A cluster node may determine whether to process the traffic steered via the backplane by computing a hash of packet parameters of the network traffic. Hash keys may be selected such that uniformity is assured, and the key used in hash computation may be synchronized across all of the nodes so that only one node determines that it should process the particular packets or traffic flow.

主权项

1. A method for distributing network traffic across a cluster of intermediary devices, the method comprising:
(a) receiving, by a first intermediary device of a cluster comprising a plurality of intermediary devices, a network packet; (b) transmitting, by the first intermediary device, the network packet on a common data backplane of the cluster to each of the other intermediary devices of the cluster; (c) determining, by a second intermediary device of the cluster responsive to receipt of the network packet from the common data backplane of the cluster, not to process the network packet responsive to a hash of a tuple of the network packet; and (d) determining, by a third intermediary device of the cluster responsive to receipt of the network packet from the common data backplane of the cluster, to process the network packet responsive to the hash of the tuple of the network packet.