Workload Storage Allocation Engine for HCI Performance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hyper-Converged Infrastructure (HCI) systems face challenges in guaranteeing workload performance requirements due to unpredictable workload provisioning and performance variations among storage devices, leading to potential session drops or failure in meeting storage allocations.
Innovation Solution
An Information Handling System (IHS) with a workload/storage allocation engine that identifies workloads requiring storage resources, retrieves performance requirements, and allocates storage devices based on their attributes to ensure performance capabilities meet the workload demands, using a storage device attribute structure to determine suitable devices for allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If HCI systems allocate storage to multiple workloads using best-effort allocation, then storage utilization is improved, but workload performance requirements cannot be guaranteed
Solution Approach 1:
The storage system is segmented into multiple storage pools with different performance characteristics (e.g., high-performance SSD pool, standard HDD pool). Each workload is allocated to appropriate storage pools based on its performance requirements, allowing the system to maintain both high utilization and performance guarantees by dividing the storage resource management into distinct segments.
Solution Approach 2:
Different storage pools are assigned different quality levels based on performance characteristics. Critical workloads requiring guaranteed performance are allocated to high-performance storage pools with stricter quality controls, while non-critical workloads can use standard storage pools. This local quality differentiation enables the system to satisfy performance requirements for specific workloads while maintaining overall high utilization.
2Adaptability or versatility
If HCI systems allocate storage dynamically to unpredictable workloads, then system flexibility is improved, but storage allocation reliability deteriorates
Solution Approach 1:
The system implements dynamic workload/storage allocation where storage assignments can be adjusted based on changing workload requirements and storage pool availability. The allocation engine continuously monitors workload performance requirements and storage pool status, dynamically reassigning workloads to appropriate storage pools while maintaining performance guarantees through real-time validation against stored performance requirement data.
Solution Approach 2:
The system incorporates feedback mechanisms where workload performance monitoring and storage pool status information are continuously fed back to the allocation engine. This feedback loop enables the system to maintain reliability by detecting when performance requirements are at risk and taking corrective actions such as reallocating workloads to different storage pools, while simultaneously maintaining flexibility to adapt to changing conditions.
3Productivity
If storage devices are pooled together for shared access, then resource efficiency is improved, but performance variations among devices cause allocation failures
Solution Approach 1:
The system changes the parameter of storage device characterization from generic capacity-only metrics to multi-dimensional performance parameters including IOPS, throughput, latency, and other workload-specific metrics. By storing and utilizing these detailed performance parameters in the allocation engine, the system can make informed allocation decisions that match workload requirements with appropriate device capabilities, maintaining both resource efficiency and allocation reliability despite performance variations among pooled devices.
Data Source
AI summary
A performance-based workload/storage allocation system includes a workload/storage allocation device coupled via controller device(s) to storage devices that each include a respective storage device attribute structure having storage device attributes that identify performance capabilities of that storage device. The workload storage/allocation device identifies a first workload that requires storage resources, and retrieves first workload performance requirement(s) associated with the first workload. The workload storage/allocation device then retrieves the storage device attributes that identify the performance capabilities of each of the storage devices via the controller device(s) and from the respective storage device attribute structure included in each of the storage devices, and uses them to determine that at least one of the plurality of storage devices includes performance capabilities that satisfy the first workload performance requirement(s). The workload/storage allocation device then allocates the at least one of the plurality of storage devices for use with the first workload.