Storage Workload Analysis for Automated Performance Service Levels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing performance of workloads in large and complex storage systems is challenging due to the difficulty in assigning appropriate service levels that match workload profiles and overall system goals, leading to resource overloading and performance loss.
Innovation Solution
A performance manager automates the monitoring and management of workloads by assigning performance service levels (PSLs) based on historical data, overprovisioning resources, and adjusting PSLs to ensure consistent performance across the storage system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If manual service level assignment is used to manage workloads, then flexibility in customizing service levels is maintained, but the complexity and difficulty of assigning appropriate service levels increases prohibitively with deployment size
Solution Approach 1:
The system automatically assigns service levels to workloads based on their characteristics and performance requirements, eliminating the need for manual intervention. The performance manager autonomously monitors workload behavior, analyzes resource consumption patterns, and dynamically adjusts service level assignments, allowing the system to serve itself rather than requiring human operators to manage each workload individually
Solution Approach 2:
The system changes the parameters of service level assignment from static manual configuration to dynamic automated adjustment. By continuously monitoring workload parameters such as I/O intensity, access patterns, and resource consumption, the system adapts service level parameters automatically, transforming the management approach from fixed to flexible and responsive
2Reliability
If resources are overprovisioned to ensure consistent performance, then performance reliability is improved, but resource utilization efficiency decreases
Solution Approach 1:
The system implements dynamic resource provisioning that adjusts service level allocations in real-time based on actual workload demands. Rather than statically overprovisioning resources, the performance manager continuously monitors workload behavior and dynamically scales resource allocation up or down, ensuring consistent performance during peak demands while maximizing resource utilization during lower activity periods
Solution Approach 2:
The system establishes a feedback loop where the performance manager continuously monitors workload performance metrics and resource consumption, then uses this feedback to automatically adjust service level assignments. This closed-loop control ensures that resources are allocated appropriately based on actual needs, maintaining performance reliability while preventing unnecessary resource allocation
3Productivity
If service levels are manually assigned to match workload profiles, then performance optimization is achieved, but the time and effort required for management increases significantly
Solution Approach 1:
The performance manager autonomously performs workload analysis, service level selection, and assignment without requiring human intervention. The system automatically evaluates workload characteristics, compares them against service level criteria, and assigns appropriate service levels, eliminating the time-consuming manual process while maintaining performance optimization
Solution Approach 2:
The system pre-configures multiple service level templates with different performance characteristics and resource allocations before workloads need to be assigned. When a workload is introduced or modified, the performance manager quickly matches it against the pre-defined templates and assigns the most appropriate service level, significantly reducing the time required compared to creating and configuring service levels manually from scratch
Data Source
AI summary
Systems, methods, and machine-readable media for monitoring a storage system and assigning performance service levels to workloads running on nodes within a cluster are disclosed. A performance manager may estimate the performance demands of each workload within the cluster and assign a performance service level to each workload according to the performance requirements of the workload, and further taking into account an overall budgeting framework. The estimates are performed using historical performance data for each workload. A performance service level may include a service level object, a service level agreement, and latency parameters. These parameters may provide a ceiling to the number of operations per second that a workload may use without guaranteeing the use of the operations per second, a guaranteed number of operations per second that a workload may use before being throttled, and define the permitted delay in completing a request to the workload.