Automated capacity management in distributed computing systems

A computing device for performing capacity management in a distributed computing system having multiple hosts interconnected by a computer network, the computing device comprising:

• a processor;
• and
• a memory operatively coupled to the processor, the memory containing instructions executable by the processor to cause the computing device to:
• predict one or more future usage levels of a computing resource in the distributed computing system based on data representing historical usage levels of the computing resource;
• derive an inventory level of the computing resource based on the predicted one or more future usage levels and data representing a currently available capacity of the computing resource in the distributed computing system;
• determine whether the derived inventory level exceeds a preset threshold associated with a capacity shortage of the computing resource;
• and
• in response to determining that the determined inventory level does not exceed a preset threshold, instruct one or more out-for-repair hosts that provide the computing resource to perform recovery before performing recovery of other out-for-repair hosts that provide other computing resources.

A method of capacity management in a distributed computing system having one or more fabric controllers, cluster controllers, and hosts interconnected by a computer network, the method comprising:

• receiving, via the computing network, data representing usage levels of a computing resource of a resource type from the one or more fabric controllers and cluster controllers;
• predicting one or more future usage levels of a computing resource in the distributed computing system based on the received data representing historical usage levels of the computing resource;
• determining whether a currently available capacity of the computing resource in the distributed computing system is depleted beyond a threshold time period based on the one or more future usage levels;
• and
• in response to determining that the currently available capacity of the computing resource in the distributed computing system is depleted before the threshold time period, instructing one or more out-for-repair hosts that provide the computing resource to perform recovery before performing recovery of other out-for-repair hosts that provide other computing resources.