Technologies for assigning workloads based on resource utilization phases

One or more non-transitory machine-readable storage media comprising a plurality of instructions stored thereon that, in response to being executed, cause an orchestrator server to:

• assign a set of workloads to the managed nodes;
• receive telemetry data from the managed nodes, wherein the telemetry data is indicative of resource utilization by each of the managed nodes as the workloads are performed and wherein each managed node includes a set of sleds that define pools of different types of disaggregated resources;
• identify, as a function of the telemetry data, historical resource utilization phases of the workloads, wherein each historical resource utilization phase is indicative of a utilization of a particular type of managed node component that satisfies a predefined threshold amount over a time period;
• determine, as a function of the historical resource utilization phases and as the workloads are performed, predicted resource utilization phases for the workloads, wherein each predicted resource utilization phase is indicative of a predicted utilization of a particular type of managed node component that satisfies the predefined threshold amount over a second time period;
• and
• apply, as a function of the predicted resources utilization phases, an adjustment to the assignments of the workloads among the managed nodes to delay an execution of a first workload to temporally align a first predicted resource utilization phase of the first workload with a second predicted resource utilization phase of a second workload on the same managed node to cause the first predicted resource utilization phase and the second predicted resource utilization phase to occur concurrently and wherein the first predicted resource utilization phase and the second resource utilization phase are predicted to utilize complementary amounts of the same resource.

An orchestrator server to assign workloads among a set of managed nodes based on resource utilization phases, the orchestrator server comprising:

• circuitry for assigning a set of workloads to the managed nodes;
• circuitry for receiving telemetry data from the managed nodes, wherein the telemetry data is indicative of resource utilization by each of the managed nodes as the workloads are performed and wherein each managed node includes a set of sleds that define pools of different types of disaggregated resources;
• means for identifying, as a function of the telemetry data, historical resource utilization phases of the workloads, wherein each historical resource utilization phase is indicative of a utilization of a particular type of managed node component that satisfies a predefined threshold amount over a time period;
• means for determining, as a function of the historical resource utilization phases and as the workloads are performed, predicted resource utilization phases for the workloads, wherein each predicted resource utilization phase is indicative of a predicted utilization of a particular type of managed node component that satisfies the predefined threshold amount over a second time period;
• and
• means for applying, as a function of the predicted resources utilization phases, an adjustment to the assignments of the workloads among the managed nodes to delay an execution of a first workload to temporally align a first predicted resource utilization phase of the first workload with a second predicted resource utilization phase of a second workload on the same managed node to cause the first predicted resource utilization phase and the second predicted resource utilization phase to occur concurrently and wherein the first predicted resource utilization phase and the second resource utilization phase are predicted to utilize complementary amounts of the same resource.

A method for assigning workloads among a set of managed nodes based on resource utilization phases, the method comprising:

• assigning, by an orchestrator server, a set of workloads to the managed nodes;
• receiving, by the orchestrator server, telemetry data from the managed nodes, wherein the telemetry data is indicative of resource utilization by each of the managed nodes as the workloads are performed and wherein each managed node includes a set of sleds that define pools of different types of disaggregated resources;
• identifying, by the orchestrator server and as a function of the telemetry data, historical resource utilization phases of the workloads, wherein each historical resource utilization phase is indicative of a utilization of a particular type of managed node component that satisfies a predefined threshold amount over a time period;
• determining, by the orchestrator server and as a function of the historical resource utilization phases and as the workloads are performed, predicted resource utilization phases for the workloads, wherein each predicted resource utilization phase is indicative of a predicted utilization of a particular type of managed node component that satisfies the predefined threshold amount over a second time period;
• and
• applying, by the orchestrator server and as a function of the predicted resources utilization phases, an adjustment to the assignments of the workloads among the managed nodes to delay an execution of a first workload to temporally align a first predicted resource utilization phase of the first workload with a second predicted resource utilization phase of a second workload on the same managed node to cause the first predicted resource utilization phase and the second predicted resource utilization phase to occur concurrently and wherein the first predicted resource utilization phase and the second resource utilization phase are predicted to utilize complementary amounts of the same resource.