Systems and methods for optimizing a software allocation to shared resources based on a dynamic mapping of resource relationships

A system comprising:

• a plurality of nodes, each node of the plurality of nodes comprises resources that are available for executing different applications;
• and
• a controller comprising one or more processors configured to:
• map relationships between the resources of the plurality of nodes;
• generate a graph with a plurality of vertices that arrange the plurality of nodes into different groups based on the mapping of the relationships between the resources of the plurality of nodes, wherein generating the graph comprises connecting a first set of elements to the graph with a first set of vertices based on a common first relationship between the resources of a first set of the plurality of nodes, and connecting a second set of elements to the graph with a second set of vertices based on a common second relationship between the resources of a second set of the plurality of nodes, and wherein the common first relationship is different than the common second relationship;
• receive a request to run a particular application;
• determine different sets of the plurality of nodes that each comprise a particular set of resources for running the particular application;
• determine a set of optimization constraints associated with running the particular application;
• traverse the graph based on different vertices of the plurality of vertices satisfying different constraints of the set of optimization constraints;
• select a particular set of nodes from the different sets of nodes based on said traversing ending at a group comprising one or more of the particular set of nodes, wherein the relationships between the resources of the particular set of nodes satisfy more of the set of optimization constraints than the relationships between the resources of other sets of nodes from the different sets of nodes;
• and
• execute the particular application using the resources of the particular set of nodes.

A non-transitory computer-readable medium, storing a plurality of processor-executable instructions to:

• map relationships between resources of a plurality of nodes;
• generate a graph with a plurality of vertices that arrange the plurality of nodes into different groups based on the mapping of the relationships between the resources of the plurality of nodes, wherein generating the graph comprises connecting a first set of elements to the graph with a first set of vertices based on a common first relationship between the resources of a first set of the plurality of nodes, and connecting a second set of elements to the graph with a second set of vertices based on a common second relationship between the resources of a second set of the plurality of nodes, and wherein the common first relationship is different than the common second relationship;
• receive a request to run a particular application;
• determine different sets of the plurality of nodes that each comprise a particular set of resources for running the particular application;
• determine a set of optimization constraints associated with running the particular application;
• traverse the graph based on different vertices of the plurality of vertices satisfying different constraints of the set of optimization constraints;
• select a particular set of nodes from the different sets of nodes based on said traversing ending at a group comprising one or more of the particular set of nodes, wherein the relationships between the resources of the particular set of nodes satisfy more of the set of optimization constraints than the relationships between the resources of other sets of nodes from the different sets of nodes;
• and
• execute the particular application using the resources of the particular set of nodes.