A system and method that includes receiving an API request to a type of API resource; retrieving an API concurrency value for the API request; determining a comparison status associated with a comparison of the API concurrency value to a concurrency threshold; if the comparison status is within the concurrency threshold, transmitting the API request to an API processing resource; if the comparison status indicates the concurrency threshold is not satisfied, impeding processing of the API request; accounting for an increase in the API concurrency value if the API request is transmitted to an API processing resource; and accounting for a decrease in the API concurrency value at a time associated with the API processing resource completing processing of the API request.

Technologies are disclosed for real-time workload tracking and throttling within a multi-tenant service. Multi-tenant services receive requests from computing devices associated with different tenants. While processing requests, the multi-tenant service itself sends requests to an underlying resource, such as a database. Requests from computing device associated with an overactive tenant may cause the multi-tenant service to overwhelm the underlying resource. The overwhelmed underlying resource may not know which tenant a request received by the underlying resource is associated with, and so the underlying resource is unable to only throttle requests originating from computing devices associated with the overactive tenant. Instead, the underlying resource throttles all requests from the multi-tenant service. To avoid this result, the multi-tenant service tracks utilization of the underling resource associated with each tenant, and throttles requests received from overactive tenants before the underlying resource becomes overwhelmed and throttles all requests from the multi-tenant service.

An apparatus includes multiple processors, a classifier and queue management logic. The classifier is configured to classify tasks, which are received for execution by the processors, into multiple processor queues, each processor queue associated with a single processor or thread, and configured to temporarily store task entries that represent the tasks, and to send the tasks for execution by the associated processors. The queue management logic is configured to set, based on queue-lengths of the queues, an affinity strictness measure that quantifies a strictness with which the tasks of a same classified queue are to be processed by a same processor, and to assign the task entries to the queues while complying with the affinity strictness measure.

An apparatus system is provided which comprises: a first component and a second component; a first circuitry to assign the first component to a first group of components, and to assign the second component to a second group of components; and a second circuitry to assign a first maximum frequency limit to the first group of components, and to assign a second maximum frequency limit to the second group of components, wherein the first component and the second component are to respectively operate in accordance with the first maximum frequency limit and the second maximum frequency limit.

Systems and methods for allocating processes to queues are provided, which provides more efficient execution of batch jobs in various embodiments. Queue priorities are assigned while process priorities and queue limits are assigned to processes. A set of queues is determined by matching the queue priority to the process priority of a process. Batch numbers for the set of queues are determined, each batch number indicating groups of messages to be processed. First queues and second queues from the set of queues are determined, the first queues having higher batch numbers than the second queues and a number of queues up to a queue limit of the process. The first queues are processed using the process. The queue priority of the second queues is decremented and the second queues are processed by another process with the process priority that matches the decremented queue priority of the second queues.

The disclosure relates to systems, methods and computer-readable media for dynamically provisioning resources within a compute environment. The method aspect of the disclosure comprises A method of dynamically provisioning resources within a compute environment, the method comprises analyzing a queue of jobs to determine an availability of compute resources for each job. determining an availability of a scheduler of the compute environment to satisfy all service level agreements (SLAs) and target service levels within a current configuration of the compute resources, determining possible resource provisioning changes to improve SLA fulfillment, determining a cost of provisioning; and if provisioning changes improve overall SLA delivery, then re-provisioning at least one compute resource.

A system and method that includes receiving an API request to a type of API resource; retrieving an API concurrency value for the API request; determining a comparison status associated with a comparison of the API concurrency value to a concurrency threshold; if the comparison status is within the concurrency threshold, transmitting the API request to an API processing resource; if the comparison status indicates the concurrency threshold is not satisfied, impeding processing of the API request; accounting for an increase in the API concurrency value if the API request is transmitted to an API processing resource; and accounting for a decrease in the API concurrency value at a time associated with the API processing resource completing processing of the API request.

Embodiments include methods performed by a processor of a vehicle. The processor may determine a priority to safe vehicle operations of each of a plurality of vehicle applications based on relative impacts on driver performance of safety-related tasks under one or more current vehicle conditions. The processor may determine a driver-performance-safety factor for each of the plurality of vehicle applications. The processor may allocate computing resources to each of the plurality of vehicle applications based on the determined driver-performance-safety factor of each vehicle application.

Resource use efficiency is improved while realizing quality guarantee of an application. A resource allocation. device 10 includes a storage unit 13 that stores resource capacity information 120 indicating a capacity of each of server resources 200, an SLI information collection unit 11 that acquires information regarding an SLI at a predetermined time interval with regard to each of a plurality of applications 2, and a resource allocation determination unit 12 that calculates an allocation resource amount of each application 2 using a moving average and a standard deviation of the acquired informs information regarding the SLI during a predetermined period, and determines server resources 200 which are allocation. destinations of the applications 2 by sorting the applications 2 in descending order of the allocation resource amounts and sequentially adding the allocation resource amounts of the sorted. applications 2 within a range which does not exceed a capacity of each server resource 200 in descending order of the allocation resource amounts.

Embodiments are described for an autonomously and dynamically allocating resources in a distributed network based on forecasted a-priori CPU resource utilization, rather than a manual throttle setting. A multivariate (CPU idle %, disk I/O, network and memory) rather than single variable approach for Probabilistic Weighted Fuzzy Time Series (PWFTS) is used for forecasting compute resources. The dynamic throttling is combined with an adaptive compute change rate detection and correction. A single spike detection and removal mechanism is used to prevent the application of too many frequent throttling changes. Such a method can be implemented for several use cases including, but not limited to: cloud data migration, replication to a storage server, system upgrades, bandwidth throttling in storage networks, and garbage collection.