Systems and methods for proactively and reactively allocating resources in cloud-based networks

A resource allocation system comprising:

• at least one processor;
• at least one memory;
• and
• one or more applications stored in the at least one memory and including instructions, which are executable by the at least one processor and are configured to:
• receive forecast data from a forecast server computer including a plurality of performance count values, wherein the forecast data indicates predicted metric data corresponding to a cloud-based service for a first geographical region,
• determine an expected usage amount for the first geographical region based on the forecast data, a confidence value, and an overload value,
• determine a failover resource amount to cover a failure in a second geographical region, wherein the expected usage amount does not include the failover resource amount and the failover resource amount does not include the expected usage amount,
• determine a first resource allocation value based on the expected usage amount and the failover resource amount,
• determine a second resource allocation value based on the plurality of performance count values,
• prioritize the first resource allocation value and the second resource allocation value by at least selecting one of the first resource allocation value or the second resource allocation value based at least in part on timing conditions, and
• adjust a resource allocation amount for the first geographical region based on a result of the prioritization.

A resource allocation system comprising:

• at least one processor;
• at least one memory;
• and
• one or more applications stored in the at least one memory and including instructions, which are executable by the at least one processor and are configured to:
• load a time series of metric values and scaled resource allocation values corresponding to providing a cloud-based service for a first geographical region,
• calculate a workload time series based on the time series of metric values and the scaled resource allocation values,
• analyze the workload time series to detect a pattern of workloads, wherein a workload refers to an amount of resource allocation,
• based on the detected pattern of workloads, generate a predicted schedule of resource allocation amounts for future predetermined periods of time for the first geographical region, wherein the resource allocation amounts of the predicted schedule are for the cloud-based service and are determined according to a forecasting algorithm or an autoregressive integrated moving average,
• allocate cloud-based resources based on the predicted schedule,
• detect a fault in a second geographical region,
• adjust the predicted schedule for the first geographical region to accommodate additional anticipated workload as a result of the fault in the second geographical region, and
• reallocate the cloud-based resources based on the adjusted predicted schedule.