AUTOMATED LEARNING OF PEERING POLICIES FOR POPULARITY DRIVEN REPLICATION IN CONTENT DELIVERY FRAMEWORK

摘要

A computer-implemented method includes collecting information about at least one resource, the information having been determined based on (i) a set of features of the at least one resource and on (ii) information about previous requests for the at least one resource; determining a computable function of the set of features of the at least one resource, the computable function having been determined based on the information about the at least one resource, the function defining a peering policy for the at least one resource; and providing the function to at least one service endpoint in a cluster or supercluster.

主权项

1. A computer-implemented method, the method operable on one or more devices comprising hardware including memory and at least one processor, the method comprising:
(A) collecting information about at least one resource, said information having been determined based on (i) a set of features of said at least one resource and on (ii) information about previous requests for said at least one resource; (B) determining a computable function of said set of features of said at least one resource, said computable function having been determined based on said information about said at least one resource, said function defining a peering policy for said at least one resource; and (C) providing said function to at least one service endpoint in a cluster or supercluster.