Server QoS Management via Segmented Power Policies
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Server computing systems face challenges in balancing energy efficiency with Quality of Service (QoS) assurance, particularly in managing diverse processing demands from devices with varying QoS requirements, such as instantaneous responses for online gaming and delayed processing for file downloads, without compromising performance or increasing energy consumption.
Innovation Solution
A method that categorizes incoming requests into service classes based on QoS criteria and assigns them to specific service queues, with each queue group having a distinct power management policy to ensure energy efficiency while meeting QoS standards, utilizing dynamic voltage and frequency scaling, core folding, and low power modes to optimize resource utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the server processes all requests at high performance mode, then QoS requirements are met, but energy consumption increases
Solution Approach 1:
The patent segments computing resources into multiple groups (first group, second group, third group) with different power management policies. Each group handles specific service queues (real-time, deadline, available resource), allowing the system to apply different performance levels to different request types, thereby meeting QoS requirements only where necessary while reducing overall energy consumption.
Solution Approach 2:
The patent implements dynamic power management policies that adjust the operational state of computing resource groups based on real-time conditions. The policies can transition groups between active, idle, and sleep states, and dynamically adjust voltage and frequency settings, enabling the system to adapt performance levels to actual demand and minimize energy consumption while maintaining QoS assurance.
2Use of energy by moving object
If the server consolidates work onto fewer cores, then energy efficiency improves, but processing capability decreases
Solution Approach 1:
The patent segments computing resources into multiple groups with different operational characteristics. By consolidating work onto fewer cores within specific groups while maintaining other groups in active or idle states, the system achieves energy efficiency without sacrificing overall processing capability, as the segmented architecture allows flexible resource allocation based on workload demands.
Solution Approach 2:
The patent creates a multi-functional computing resource architecture where each group can serve different service queues and workload types. This universality allows the system to consolidate work onto fewer cores for energy efficiency when appropriate, while still maintaining the capability to process diverse request types across multiple groups when high processing power is needed.
3Use of energy by moving object
If the server uses dynamic voltage and frequency scaling, then power consumption adjusts to workload, but system complexity increases
Solution Approach 1:
The patent applies DVFS within segmented computing resource groups, where each group can independently adjust its voltage and frequency settings. This segmentation allows DVFS to be implemented in a controlled manner across different groups with different power management policies, managing the complexity through modular organization rather than system-wide complexity.
Solution Approach 2:
The patent implements dynamic voltage and frequency scaling as part of its dynamic power management policies. The system dynamically adjusts voltage and frequency settings based on workload conditions and service queue requirements, allowing power consumption to adapt to actual demand while the policy-based framework manages the complexity of these dynamic adjustments.
Data Source
AI summary
Aspects of an embodiment of the invention disclose a method, computer program product, and system for managing the energy efficiency of servers providing multi-class computing services with Quality of Service (QoS) assurance. Computing resources are clustered into at least three groups, where each group has a separate power management policy (PMP). A plurality of requests are received from a plurality of devices, and are sorted into at least three service classes based on the requests' QoS criteria. Each request is assigned to one of at least three service queues based on the request's service class, and each service group is processed by a group of computing resources. The power management policies are configured such that each group of computing resources may service requests at an energy efficient point while meeting the QoS criteria of the service class.