Automated Resource Selection for Software-Defined Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional storage approaches require tying up inventory resources for server and disk selection, preventing their usage for other deployments and workloads, and face inefficiencies in calculating storage capacity for software-defined storage needs.
Innovation Solution
An automated method using an in-memory data structure to organize inventory by disk size and count per server, calculating total storage capacity for combinations of disk sizes and counts, and determining the optimal combination to satisfy storage deployment requests, allowing for dynamic resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional storage approaches are used to select servers and disks for software-defined storage deployment, then the storage needs can be accommodated, but inventory resources are tied up and cannot be used for other deployments and workloads
Solution Approach 1:
The patent implements dynamic resource selection by organizing inventory data in an in-memory data structure that enables real-time calculation of storage capacity for multiple server and disk combinations. This allows the system to dynamically determine which resources to allocate based on current deployment requests while maintaining flexibility for other workloads, resolving the contradiction between satisfying storage needs and maintaining resource utilization flexibility.
2Measurement precision
If comprehensive inventory analysis is performed to find optimal server and disk combinations, then accurate storage capacity calculation is achieved, but the process becomes complex and time-consuming
Solution Approach 1:
The patent segments the inventory analysis process by organizing the in-memory data structure to separately track disk sizes, disk counts, and server associations. This segmentation allows the system to efficiently calculate storage capacity for different combinations by independently evaluating disk configurations across servers, achieving accurate measurement without overwhelming complexity.
Solution Approach 2:
The patent performs preliminary organization of inventory data into the in-memory data structure before deployment requests are processed. This preliminary action pre-establishes the relationships between servers, disks, and their capacities, enabling rapid and accurate storage capacity calculations when deployment requests arrive, thereby reducing the complexity of the actual resource selection process.
3Reliability
If servers and disks are reserved for future storage needs, then storage capacity is guaranteed, but the resources cannot be utilized for current workloads and other deployments
Solution Approach 1:
The patent implements dynamic resource allocation that evaluates multiple server and disk combinations in real-time based on actual deployment requests. Rather than statically reserving resources, the system dynamically determines the minimum resources needed to satisfy each request, ensuring storage capacity guarantees while maximizing resource productivity by making resources available for other workloads when not allocated.
Data Source
AI summary
Methods, apparatus, and processor-readable storage media for automated resource selection for software-defined storage deployment are provided herein. An example computer-implemented method includes obtaining a software-defined storage deployment request from a user, the request comprising an amount of total storage capacity and a minimum number of servers; generating an in-memory data structure of at least a portion of an inventory associated with the user, wherein the inventory comprises servers and disks associated with the servers, and wherein the data structure organizes the inventory based on disk size and disk count per server; calculating, using at least the data structure, total storage capacity for each of multiple combinations of disk sizes and disk counts across the servers; determining at least one of the combinations, based on the total storage calculations, that satisfies the request; and performing at least one automated action based on the combination(s) determined to satisfy the request.