Dynamic placement of computing tasks in a distributed computing environment

A method for placing a computing task within a distributed computing system that comprises first and second servers, the method comprising:

• with a first server, receiving a first request, the first request originating from a client;
• in response to receiving the first request, the first server determining whether to perform one or more computing tasks at the first server to generate particular content to provide in response to the first request;
• upon a determination by the first server to perform the one or more computing tasks:
• the first server performing the one or more computing tasks to generate the particular content, and,
• the first server sending towards the client a response to the first request, and including in the response both the particular content and a computing task locator that identifies any of the first server and a group of servers of which the first server is a member, as having performed the one or more computing tasks;
• upon a determination by the first server not to perform the one or more computing tasks:
• the first server sending a forward request to the second server for the particular content,
• receiving from the second server the particular content, and,
• the first server sending towards the client a response to the first request, and including in the response both the particular content and a computing task locator that identifies any of the second server and a group of servers of which the second server is a member, as having performed the one or more computing tasks;
• wherein the response to the first request causes the client to generate a beacon that includes performance data related to delivery of the particular content to the client and includes the computing task locator that the client received, and further causes the client to send the beacon back to the distributed computing system;
• and,
• the distributed computing system:
• receiving the performance data, and, based at least in part thereon,
• adjusting a probability that the first server determines to perform the one or more computing tasks.

A method performed in a distributed computing system that comprises a set of two or more nodes deployed in one or more networks between a client and an origin node, the method comprising:

• receiving a client request at a first node in the set, wherein the client request requires performance of one or more computing tasks;
• at the first node in the set, determining whether to perform the one or more computing tasks at the first node;
• subsequent to a determination by the first node to perform the one or more computing tasks, the first node performing the one or more computing tasks to generate a result of the performance of the one or more computing tasks;
• subsequent to a determination by the first node not to perform the one or more computing tasks, a second node in the set performing the one or more computing tasks to generate a result of the performance of the one or more computing tasks;
• sending, to the client and in response to the client request, the result of the performance of the one or more computing tasks, and one of the following:
• (a) when the first node performed the one or more computing tasks, a computing task locator that identifies any of the first node and a group of nodes of which the first node is a member as having performed the one or more computing tasks, and
• (b) when the second node performed the one or more computing tasks, a computing task locator that identifies any of the second node and a group of nodes of which the second node is a member as having performed the one or more computing tasks;
• wherein the response to the client request causes the client to generate a beacon that includes performance data related to the response to the client, the beacon further including the computing task locator the client received;
• the response further causing the client to send the beacon back to the distributed computing system;
• and,
• the distributed computing system:
• receiving the performance data, and, based at least in part thereon, adjusting a probability that the first node determines to perform the one or more computing tasks.

A distributed computing system including a set of two or more nodes deployed in one or more networks between a client and an origin node, the distributed computing system comprising:

• the two or more nodes in the set each comprising circuitry forming a processor and memory storing computer program instructions executable on the processor to operate as specified below;
• a first node in the set storing in its memory computer program instructions to, upon execution:
• receive a client request, wherein the client request requires performance of one or more computing tasks;
• determine whether to perform the one or more computing tasks at the first node;
• subsequent to a determination by to perform the one or more computing tasks, perform the one or more computing tasks to generate a result of the performance of the one or more computing tasks;
• a second node in the set storing in its memory computer program instructions to, upon execution:
• subsequent to a determination by the first node not to perform the one or more computing tasks, perform the one or more computing tasks to generate a result of the performance of the one or more computing tasks;
• the first node in the set storing in its memory computer program instructions to, upon execution:
• send, to the client and in response to the client request, the result of the performance of the one or more computing tasks, and one of the following:
• (a) when the first node performed the one or more computing tasks, a computing task locator that identifies any of the first node and a group of nodes of which the first node is a member as having performed the one or more computing tasks, and
• (b) when the second node performed the one or more computing tasks, a computing task locator that identifies any of the second node and a group of nodes of which the second node is a member as having performed the one or more computing tasks;
• wherein the response to the client request includes code to cause the client to generate a beacon that includes performance data related to the response to the client, the beacon further including the computing task locator that the client received;
• the response further including code to cause the client to send the beacon back to the distributed computing system;
• and,
• any node in the set of two or more nodes in the distributed computing system storing in memory computer program instructions to, upon execution:
• receive the performance data, and, based at least in part thereon, adjust a probability that the first node determines to perform the one or more computing tasks.