Systems and methods for automatically scaling compute resources based on demand

A scaling system for scaling compute resources in a compute group including one or more compute resources, the system comprising:

• a processor:
• a non-transitory computer-readable storage medium storing computer executable instructions, which when executed by the processor, cause the processor to:
• determine an amount of compute capacity required to complete one or more job requests;
• determine an amount of allocable compute capacity available on the compute resources in the compute group;
• calculate a utilization of the compute group based on the amount of required compute capacity and the amount of allocable compute capacity;
• determine that the utilization of the compute group either falls above a first threshold value or below a second threshold value, the first threshold value higher than the second threshold value;
• upon determining that the utilization falls above the first threshold value:
• calculate a first scaling delta using the first threshold value and the utilization to determine a number of additional compute resources required to bring the utilization of the compute group below the first threshold value;
• and
• instantiate the number of additional compute resources into the compute resources in the compute group to bring the utilization of the compute group below the first threshold value by instructing a resource provider to start up the number of additional compute resources in the compute group;
• and
• upon determining that the utilization of the compute group falls below the second threshold value:
• calculate a second scaling delta using the second threshold value and the utilization to determine a number of excess compute resources which, when removed from the compute resources would bring the utilization of the compute group above the second threshold value;
• and
• remove the number of excess compute resources from the compute resources in the compute group to bring the utilization of the compute group above the second threshold value by instructing the resource provider to terminate the number of excess compute resources in the compute group.