Autonomous intelligent workload management
View Patent ↗Apparatus, systems, and methods may operate to create a hypergraph of weighted vertices comprising computing resources and storage resources, and nets comprising workloads; to receive a plurality of requests to be addressed by a network associated with the hypergraph, at least some of the requests associated with data objects; to calculate partition schemes for the network based on the requests and the data objects according to an ant colony optimization heuristic; and to autonomously reallocate the workloads to the computing resources and/or the storage resources according to the partition schemes. The workloads may act as ants following a path defined by the vertices of the hypergraph. Further activities may thus include depositing pheromones along hyperedges of the hypergraph, wherein the hyperedges are used for swapping the vertices between the workloads. Additional apparatus, systems, and methods are disclosed.
1. An apparatus, comprising:
a hardware reception module to receive a plurality of requests having a request type to be addressed by a network, at least some of the requests associated with data objects; and
a hardware processing module to:
create a hypergraph associated with the network, the hypergraph including weighted vertices comprising computing resources and storage resources, and nets comprising workloads represented as ants following a path defined by the vertices, the hypergraph including hyperedges used for swapping the vertices between the workloads to reduce a metric based on internal costs and external costs associated with the workloads;
calculate partition schemes for the network based on the requests and the data objects according to an ant colony optimization heuristic; and
autonomously reallocate the workloads to the computing resources and/or the storage resources according to the partition schemes by ant swapping, wherein to autonomously reallocate the workloads includes controlling movement of the workloads as the ants according to an amount of pheromones deposited along the hyperedges of the hypergraph, wherein the probability of movement of one of the workloads from one of the vertices to another of the vertices depends on a product divided by a sum of products, wherein the product comprises an amount of pheromone on an edge of the hypergraph times a desirability of the edge.
2. The apparatus of claim 1 , further comprising:
a display to display the network and the vertices.
3. The apparatus of claim 1 , further comprising:
a storage node to store the hypergraph.
4. The apparatus of claim 1 , further comprising:
a capacity monitor module to monitor capacity of the computing resources and capacity of the storage resources.
5. A hardware processor-implemented method to execute on one or more processors that perform the method, comprising:
creating a hypergraph of weighted vertices comprising computing resources and storage resources, and nets comprising workloads;
receiving a plurality of requests having a request type to be addressed by a network associated with the hypergraph, at least some of the requests associated with data objects;
calculating partition schemes for the network based on the requests and the data objects according to an ant colony optimization heuristic, wherein the workloads act as ants following a path defined by the vertices of the hypergraph;
depositing pheromones along hyperedges of the hypergraph, the hyperedges used for swapping the vertices between the workloads to reduce a metric based on internal costs and external costs associated with the workloads; and
autonomously reallocating the workloads to the computing resources and/or the storage resources according to the partition schemes, wherein reallocating the workloads includes controlling movement of the workloads as the ants according to an amount of the pheromones, wherein the probability of movement of one of the workloads from one of the vertices to another of the vertices depends on a product divided by a sum of products, and wherein the product comprises an amount of pheromone on an edge of the hypergraph times a desirability of the edge.
6. The method of claim 5 , wherein the depositing occurs after the swapping.
7. The method of claim 5 , further comprising:
evaporating the pheromones after the swapping.
8. The method of claim 5 , further comprising:
randomly assigning the workloads to the vertices of the nets prior to the receiving, without minimizing cross-location data communication overhead.
9. The method of claim 5 , further comprising:
controlling the movement of the workloads as ants according to a combination of weight constraints associated with capacity of the computing resources and/or capacity of the storage resources.
10. The method of claim 5 , wherein the requests are video file requests, and wherein the request type defines a type of video file format.
11. The method of claim 5 , wherein individual ones of the vertices are assigned weights that correspond to the workloads.
12. The method of claim 5 , wherein individual ones of the vertices are assigned weights that correspond to a size of data to be stored using the storage resources.
13. The method of claim 5 , wherein the workloads are each associated with a net cost based on an expected number of the requests, the request types associated with the plurality of requests, and size of neighboring ones of the data objects.
14. The method of claim 5 , further comprising:
monitoring a threshold associated with exhausting at least one of the computing resources and/or the storage resources; and
reallocate the workload to prevent exhaustion of the at least one of the computing resources and/or the storage resources, responsive to the threshold being exceeded.
15. A computer readable storage device including instructions that, when executed by a machine, cause the machine to perform operations comprising:
creating a hypergraph of weighted vertices comprising computing resources and storage resources, and nets comprising workloads;
receiving a plurality of requests having a request type to be addressed by a network associated with the hypergraph, at least some of the requests associated with data objects;
calculating partition schemes for the network based on the requests and the data objects according to an ant colony optimization heuristic, wherein the workloads act as ants following a path defined by the vertices of the hypergraph;
depositing pheromones along hyperedges of the hypergraph, the hyperedges used for swapping the vertices between the workloads to reduce a metric based on internal costs and external costs associated with the workloads; and
autonomously reallocating the workloads to the computing resources and/or the storage resources according to the partition schemes, wherein reallocating the workloads includes controlling movement of the workloads as the ants according to an amount of the pheromones, wherein the probability of movement of one of the workloads from one of the vertices to another of the vertices depends on a product divided by a sum of products, and wherein the product comprises an amount of pheromone on an edge of the hypergraph times a desirability of the edge.
16. The computer readable storage device of claim 15 , wherein the depositing occurs after the swapping.
17. The computer readable storage device of claim 15 , further comprising:
evaporating the pheromones after the swapping.
18. The computer readable storage device of claim 15 , wherein individual ones of the vertices are assigned weights that correspond to a size of data to be stored using the storage resources.
19. The computer readable storage device of claim 15 , wherein the workloads are each associated with a net cost based on an expected number of the requests, the request types associated with the plurality of requests, and size of neighboring ones of the data objects.
20. The computer readable storage device of claim 15 , wherein the operations further include:
monitoring a threshold associated with exhausting at least one of the computing resources and/or the storage resources; and
reallocate the workload to prevent exhaustion of the at least one of the computing resources and/or the storage resources, responsive to the threshold being exceeded.