Resource allocation using relational fuzzy modeling

A computer program product stored on a computer usable medium for allocating computing resources among a plurality of units, comprising:

• program code configured for providing a relational fuzzy model based on historical usage data collected from the units;
• program code configured for predicting and allocating resource requirements for each unit based on the relational fuzzy model;
• program code configured for ensuring that a sum of the predicted resource requirements is no more than a total available amount of actual resources available on the plurality of units;
• and
• program code configured for allocating extra resources when the sum of the predicted resource requirements is less than the total available amount of resources available on the plurality of units.

A method of allocating computing resources among a plurality of units, comprising:

• generating a relational fuzzy model based on historical usage data collected from each of the units;
• predicting and allocating resource requirements for each unit based on the relational fuzzy model;
• ensuring that a sum of the predicted resource requirements is no more than a total available amount of actual resources available on the plurality of units;
• and
• allocating extra resources when the sum of the predicted resource requirements is less than the total available amount of resources available on the plurality of units.

A method for deploying an application for allocating resources to a set of units in a distributed network, comprising:

• providing a computer infrastructure being operable to:
• generate a relational fuzzy model based on historical resource parameter data and historical system parameter data collected from each of the units;
• and
• predict and allocate resource requirements for each unit based on the relational fuzzy model.