Managing computer server capacity

摘要

Systems and methods are disclosed for using machine learning (e.g., neural networks and/or combinatorial learning) to solve the non-linear problem of predicting the provisioning of a server farm (e.g., cloud resources). The machine learning may be performed using commercially available products, such as the SNNS product from The University of Stuttgard of Germany. The system, which includes a neural network for machine learning, is provided with an identification of inputs and outputs to track, and the system provides correlations between those. Rather than static rules, the machine learning provides dynamic provisioning recommendations with corresponding confidence scores. Based on the data collected/measured by the neural network, the provisioning recommendations will change as well as the confidence scores.

主权项

1. A system comprising:
a plurality of server machines, each comprising at least one computer processor, one or more computer memories, and an input-output bus; a data store storing corresponding load and health measurements of the plurality of server machines; a neural network configured to receive the corresponding load and health measurements as inputs; a memory of a computing device storing a learn module that, when executed by a processor of the computing device, causes an update to the neural network using inputs of corresponding load and health measurements of the plurality of server machines; the memory of the computing device storing a predictor module that, when executed by the processor of the computing device, generates a predicted health value of the plurality of server machines, using the updated neural network, given a hypothetical increased load value, wherein the predicted health value comprises a confidence score, wherein the predicted health value indicates a linear response when the load on the system is at a first value, and the predicted health value indicates a non-linear response when the load on the system is at a second value greater than the first value; and the memory of the computing device storing a simulator module that, when executed by the processor of the computing device, generates a simulated load on the plurality of server machines in accordance with hypothetical increased load value.