Level 3 Communications, LLC

Netflix Inc.

David Pfitzner

Jeffrey G Koller

Maksim Yevmenkin

Mark Brady

Paul Ernest Stolorz

Michael S Warren

Ted Middleton

John K Salmon

Aric Hagberg

Andrew Berglund

Jeffrey Davis

The disclosed computer-implemented method includes accessing a hard drive to measure operational characteristics of the hard drive. The method next includes deriving hard drive health factors used to control the hard drive that are based on the measured operational characteristics. The derived hard drive health factors include an average per-seek time indicating an average amount of time the hard drive spends seeking specified data that is to be read and an average read speed indicating an average amount of time the hard drive spends reading the specified data. The method next includes determining, based on the hard drive health factors and the operational characteristics, an amount of load servicing capacity currently available at the hard drive, and then includes regulating the amount of load servicing performed by the hard drive according to the determined amount of available load servicing capacity. Various other methods, systems, and computer-readable media are also disclosed.

Systems and methods for optimizing hard drive throughput

One embodiment of the present invention sets forth a technique for generating one or more hash data structures. The technique includes generating a hash data structure having entries that correspond to a plurality of content servers, and, for each file included in a first plurality of files, allocating the file to one or more content servers included in the plurality of content servers by comparing a hash value associated with the file to one or more entries included in the entries. The technique further includes comparing a network bandwidth utilization of a first content server to a network bandwidth utilization associated with one or more other content servers included in the plurality of content servers to generate a result, and modifying a first number of entries associated with the first content server and included in the entries based on the result to generate a biased hash data structure.

Hash data structure biasing

A network includes a first tier of servers, a second tier of servers and a third tier of servers. The first tier of servers comprises a plurality of primary servers capable of responding to requests. The plurality of primary servers includes at least one dynamic server having a load share characteristic that is dynamically altered during a first period in time and at least one static server having a constant load share characteristic during the first period in time. The second tier of servers includes at least two first layer overflow servers. Each of the first layer overflow servers is configured to provide response duties in the event that one or more of the primary servers is unavailable or overloaded. The third tier of servers comprising at least two second layer overflow servers. Each of the second layer overflow servers is configured to provide response duties in the event that one or more of the primary servers is unavailable or overloaded and one or more of the first layer overflow servers is unavailable or overloaded.

Multi-tiered server network

A method includes monitoring the operational status of at least some name servers in a name server network. Based on the monitoring, when an error or anomalous condition is detected at a particular name server, creating a policy to indicate the status of the particular name server. The policy is taken into account when resolving hostnames using the name server network.

Detecting anomalous conditions in a name server network

A method, operable in a framework wherein resources of a content provider may be delivered to clients from at least two distinct content delivery networks (CDNs). Responsive to a client request, one of the CDNs is selected, the selection being based at least in part on one or more policies set by the content provider. The client is provided with information associated with the selected CDN. The information may be a domain name such as a CNAME for the selected CDN. The policies may include a load share policy that specifies a relative load share for each CDN, and the selecting may attempt to satisfy the load share policy. Updated content provider policies may be obtained, at regular intervals.

Distributing requests across multiple content delivery networks based on subscriber policy

In a framework wherein resources of a content provider may be delivered to clients from different domains, a method distributes the requests based on content-provider policies. In some cases, the domains include at least two distinct content delivery network (CDN) domains. The domains may include a content provider domain. Responsive to a request, either the content provider domain or one of the two CDN domains is selected, the selection being based at least in part on one or more policies set by the content provider.

Policy-based content delivery network selection

A framework for delivery of Internet content includes a subscriber server network; and at least one domain name server constructed and adapted to provide policy-based domain name service, wherein, in response to a request to resolve a hostname, the domain name server resolves the hostname to at least one address corresponding a server in the subscriber server network based on at least one policy consideration. The framework may include a content delivery network, distinct from the subscriber server network, wherein, in response to a request to resolve a hostname, the domain name server provides at least one addresses of a server in the content delivery network based at least on the location of the requestor and other policy considerations. An address returned by the domain name server is selected based on one or more of: (a) geographic policies; (b) load share policies; (c) overflow policies; and (d) network aware policies.

Growing community of inventors

College Park, Australia

David Pfitzner