Abstract: Aspects of provisioning computing units based on improved failure resiliency are described. In one embodiment, an infrastructure component shared between a pair of computing units is identified. A failure rate for the infrastructure component is obtained, and a failure probability for a class of assigned computing units is computed based in part on the failure rate. A spread request related to the class of assigned computing units is also received. In response to the spread request, an altered composition of computing units is determined, and a difference between a failure probability for the altered composition of computing units and the failure probability for the class of assigned computing units is computed. In one embodiment, when a spread score improvement value associated with the difference meets a spread criteria of the spread request, the altered composition of computing units may be provisioned for use.

Abstract: A fact generation engine generates facts in response to fact requests submitted by clients. The fact generation engine also predictively generates additional facts prior to receiving a request for the additional facts from a client. The fact generation engine might determine whether to predictively generate additional facts based upon a determined probability that a client will request the additional facts. The estimated cost of predictively generating additional facts might also be utilized to determine whether to predictively generate the facts.

Abstract: Aspects of provisioning computing units based on improved failure resiliency are described. In one embodiment, an infrastructure component shared between a pair of computing units is identified. A failure rate for the infrastructure component is obtained, and a failure probability for a class of assigned computing units is computed based in part on the failure rate. A spread request related to the class of assigned computing units is also received. In response to the spread request, an altered composition of computing units is determined, and a difference between a failure probability for the altered composition of computing units and the failure probability for the class of assigned computing units is computed. In one embodiment, when a spread score improvement value associated with the difference meets a spread criteria of the spread request, the altered composition of computing units may be provisioned for use.

Abstract: Disclosed are various embodiments for provisioning computing units. A spread request is received. The spread request relates to a class of assigned computing units residing within a plurality of networked computing units. The spread request is associated with a spread criteria. In response to the request, a plurality of networked computing units is provisioned based on failure correlation data and in accordance with the spread criteria, to produce a final spread score. Success is indicated in response to the request if the final spread score meets the spread criteria.

Abstract: A host may provide a client-managed hierarchical resource allocation. In some aspects, a host may analyze a computing system including a plurality of computing resources to identify, for various tasks, dependencies between the computing resources. The host may receive a new task to be performed by the computing system. The new task may be associated with a similar task performed by the computing system to enable modeling the new task using dependencies between computing resources which are determined from the similar task. The host may receive an estimate of computing resources required by the new task. The host may then allocate at least a portion of the computing resources to perform the new task based on the estimated workload of the new task and similarities between the new task and the similar task.

Abstract: Techniques described herein provide a self-service, run-time configuration store for interaction between a business owner and a Web service. For instance, the techniques may receive entity-specific configuration data from a business entity, may validate and store the entity-specific configuration data in a central location, and may receive a service-specific schema definition from a Web service. In response to a command to push the validated configuration data, or in response to a request to pull the validated configuration data, the techniques serve the entity-specific configuration data to configure the Web service for use with customers of the business entity.