Storage Capacity Planning in Hyper-Converged Infrastructure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In hyper-converged infrastructure (HCI) environments, storage capacity planning is challenging due to the complexity of storage activities and the time-consuming procurement process for adding or removing storage resources, leading to potential performance downgrades, upgrade failures, or service interruptions if capacity thresholds are exceeded.
Innovation Solution
A system and process for performing storage capacity planning in HCI environments using a machine learning model trained on historical storage capacity usage data, which includes data preprocessing to remove invalid and spike data, and normalization, to generate accurate predictions for storage capacity usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional manual storage capacity planning is used, then procurement process can be completed, but it takes months to add new storage resources and may exceed storage capacity threshold before new resources are obtained
Solution Approach 1:
The system performs preliminary storage capacity planning by analyzing historical usage data and predicting future storage needs before the procurement process completes. This allows the organization to prepare approval workflows, budget allocations, and resource provisioning in advance, so that when new storage resources are needed, they can be deployed immediately rather than waiting months for the entire procurement cycle.
Solution Approach 2:
The system implements continuous monitoring of storage capacity usage and provides real-time feedback to stakeholders. By tracking actual usage patterns against predictions and alerts, the system enables dynamic adjustment of procurement timing and resource allocation, ensuring capacity thresholds are maintained while optimizing the procurement cycle duration.
2Ease of operation
If storage capacity planning is performed manually in HCI environment, then procurement can be managed, but storage capacity planning becomes challenging due to complicated storage activities
Solution Approach 1:
The system automates storage capacity planning by having it self-service through machine learning models that automatically analyze complex storage activities, apply storage policies, and generate predictions without manual intervention. The system autonomously processes complicated storage patterns, workload variations, and policy applications, transforming an operationally complex task into a simple, automated service.
Solution Approach 2:
The patent replaces manual mechanical planning processes with an automated machine learning-based system. Instead of human analysts manually analyzing complex storage activities and making planning decisions, the system uses algorithms to automatically process storage data, predict capacity needs, and generate recommendations, thereby simplifying the operational complexity.
3Reliability
If storage resources are added quickly to prevent exceeding capacity threshold, then service continuity is maintained, but procurement process complexity increases
Solution Approach 1:
The system performs preliminary capacity planning and prepares procurement approvals in advance, so that when storage resources need to be added, the administrative and procurement processes are already authorized and ready to execute quickly. This maintains service continuity by having pre-approved resource allocation plans that can be implemented immediately when needed.
Solution Approach 2:
The system provides continuous feedback on storage capacity usage and predicts when thresholds will be approached, enabling proactive procurement planning. This feedback mechanism allows the organization to time resource additions optimally, maintaining service continuity while smoothing out procurement processes rather than creating urgent, complex last-minute procurement situations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
One example method to perform storage capacity planning in a hyper-converged infrastructure (HCl) environment is disclosed. The method includes obtaining historical storage capacity usage data of a set of virtual storage area network (vSAN) clusters, processing the historical storage capacity usage data to generate processed historical storage capacity usage data, training a machine learning model with the processed historical storage capacity usage data to generate a first trained machine learning model, and in response to a first vSAN cluster being newly deployed in the HCl environment, dispatching the first trained machine learning model to the first vSAN cluster.