Workload balancing between nodes in a cluster as required by allocations of IP addresses within a cluster

摘要

Balancing workload distributed between nodes in a cluster through allocations of IP addresses by determining, for each node, an imbalance metric for each possible pairing of the plurality of an IP address assignable to the node based upon the distances between the IP addresses in each pairing and selectively distributing a plurality of IP addresses between the nodes in a cluster based upon minimizing the sum of the imbalance metrics of the IP addresses for each of the nodes in the cluster resulting from the selected distribution.

主权项

1. In a cluster of nodes in a network, a method for balancing workload distributed between the nodes in the cluster in response to a workload allocation request from a source IP address to the nodes in the cluster comprising:
determining, for each node in the cluster, an imbalance metric for each possible pairing of each source IP address allocatable to said node based upon the distances between said source IP addresses in each pairing; wherein said distance between a pair of source IP addresses is the longest prefix, taken bitwise, common to each of said pair of IP addresses; said imbalance metric being determined by squaring the distances of all paired source IP addresses between the node; and calculating the sum of all paired source IP addresses allocated to said node; and calculating the sum of ail of said pairings for said node; and selectively allocating said source IP addresses between the nodes in cluster based upon minimizing the sum of the imbalance metrics of the nodes resulting from the source IP addresses allocated to each of the nodes in the cluster.