Workload Manager for Multi-Tier Storage Resource Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-tiered storage systems lack awareness of workload schedules and resource allocation priorities, leading to inefficient resource utilization and conflicts among tenants, as they are not coupled with workload managers and do not dynamically adjust storage tier allocation based on actual demand.
Innovation Solution
Implementing a system that interfaces between workload managers and multi-tiered storage systems, allowing the storage system to be aware of resource demands and translating resource plans into data placement and migration requirements, thereby dynamically allocating resources across storage tiers to exceed individual tier capacities and optimize resource utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a static storage tier allocation system is used, then device complexity is reduced and ease of operation is improved, but resource utilization efficiency deteriorates and tenant conflicts increase
Solution Approach 1:
The patent introduces a workload manager as an intermediary component that sits between the storage system and tenants. This workload manager receives workload information from tenants, determines resource demands, and translates them into storage tier allocation requirements. By inserting this mediator layer, the system achieves dynamic resource allocation without requiring complex direct communication between all tenants and the storage system, thus improving resource utilization while controlling system complexity.
Solution Approach 2:
The patent implements dynamic storage tier allocation where the assignment of storage tiers to tenants is not fixed but changes based on workload demands. The system continuously monitors workload information, determines current resource demands, and reallocates storage tiers accordingly. This dynamic approach allows the system to adapt to changing conditions and optimize resource utilization without requiring complete system redesign.
2Productivity
If storage tiers are allocated statically to tenants, then device complexity is reduced, but the capacity and IOPS available to each tenant are limited and resource utilization is inefficient
Solution Approach 1:
The patent implements a proactive resource allocation mechanism where the workload manager receives workload information in advance, determines resource demands before they occur, and pre-allocates storage tiers accordingly. This preliminary action allows the system to prepare storage resources ahead of time, ensuring that capacity and IOPS are available when needed, thereby improving storage throughput without requiring complex real-time allocation mechanisms.
Solution Approach 2:
The workload manager serves multiple functions: it receives workload information from tenants, determines resource demands, translates demands into storage tier allocations, and manages the allocation process. This multi-functional approach consolidates several operations into a single component, improving storage throughput through efficient resource management while avoiding the complexity of multiple separate systems working in coordination.
3Adaptability or versatility
If the storage system operates without workload manager integration, then device complexity is reduced and ease of operation is improved, but the system lacks awareness of resource demands and cannot dynamically adjust allocation
Solution Approach 1:
The patent implements a feedback mechanism where the workload manager continuously receives workload information from tenants, processes this information to determine resource demands, and uses this feedback to adjust storage tier allocations dynamically. This closed-loop feedback system enables the storage system to become aware of resource demands and adapt its allocation accordingly, improving adaptability while managing integration complexity through a centralized management approach.
Data Source
AI summary
Embodiments for managing resources in a tiered data storage system, by a processor device. In response to differing business priorities for executing workloads by a workload manager, a resource plan is generated by an administrator of the workload manager. The resource plan defines what resources of the storage system are allocated to a particular group of tenants and at what time the workloads of the particular group are executed, thereby optimizing utilization of the resources of the storage system. A time period of which a particular one of a plurality of storage tiers of the tiered data storage system is allocated for use by the respective one of the tenants is defined by an administrator, where a peak demand summed over all of tenants is allowed to exceed a total capacity of each individual one of the plurality of storage tiers.