Search All Applications in Edge Computing
Application US20150358401
Published 2015-12-10
Load Balancing With Layered Edge Servers
An edge server node comprises an edge server, a DNS, and a load controller. The DNS is assigned to a layer of edge server nodes where each layer of edge server nodes has a DNS anycast IP address for the DNSs of that layer and a FQDN for that layer. The DNS either handles or offloads requests to resolve the FQDN for the assigned layer to balance the load on the edge server. A request is offloaded by returning the FQDN for another layer. The load controller monitors the load on the edge server and directs the DNS to offload requests to reduce the load on the edge server.
Much More than Average Length Specification
View the Patent Matrix® Diagram to Explore the Claim Relationships
USPTO Full Text Publication >
- 1. A computer-readable storage medium storing computer-executable instructions for controlling a node with a domain name server (“DNS”) and an associated edge server (“ES”) of a network to load balance traffic for the associated ES, the computer-executable instructions comprising instruction that when executed:
access a DNS anycast Internet Protocol (“IP”) address for the DNS and fully qualified domain name (“FQDN”), the node being assigned to a layer of nodes from an innermost layer to an outermost layer, the DNS anycast IP address being for the DNSs of the assigned layer, the FQDN being for the assigned layer; announce the DNS anycast IP address; and when a request to resolve the FQDN for the assigned layer is received by the DNS,
when the load on the associated ES indicates that the DNS should not be offloading load from the associated ES, handle the request by returning an ES anycast IP address for the associated ES; and
when the load on the associated ES indicates that the DNS request should be offloading load from the associated ES, offload the request by returning the FQDN of an inner layer.
- 12. An edge server node comprising:
an edge server (“ES”); a domain name server (“DNS”) that is assigned to a layer of edge server nodes, each layer of edge server nodes having a DNS anycast Internet Protocol (“IP”) address for the DNSs of that layer and a fully qualified domain name (“FQDN”) for that layer, the DNS including a component that either handles or offloads requests to resolve the FQDN for the assigned layer to balance load on the edge server, a request being offloaded by returning the FQDN for another layer; and a load controller that monitors load on the ES and directs the DNS to offload requests to reduce the load on the ES.
- 14. A method of load balancing for systems, comprising:
associating each system in a first group of systems and identifier servers (“NSs”) with a NS in the first group and each system in a second group of systems and NSs with a NS in the second group; assigning a first NS anycast identifier to the NSs in the first group and a first identifier to the systems in the first group, and a second NS anycast identifier to the NSs in the second group and a second identifier to the systems in the second group; and when a first receiving NS in the first group receives a request sent to the first NS anycast identifier for resolution of the first identifier,
when the load on the system associated with the first receiving NS should be handled by the associated system, handling the request by the first receiving NS by returning an identifier of the associated system; and
when the load on the system associated with the first receiving NS should not be handled by the associated system, offloading the request by returning the second identifier so that a second receiving NS of the second group can handle the request and return an identifier of the system of the second group associated with the second receiving NS.
