Storage Node VM Restore via Master Template Copying
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In storage systems with multiple virtual machines, restoring a failed storage node virtual machine is complex and requires significant storage capacity for backup data, especially when synchronizing across clusters, leading to inefficiencies in processing and resource utilization.
Innovation Solution
A storage system architecture that includes a physical server with a template for configuring storage node virtual machines, a collection unit for gathering configuration information, a creation unit for creating new virtual machines, and a process restore unit to replicate the failed node's processes, allowing for efficient restoration without the need for extensive backups.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If backup data of each storage node virtual machine is stored to enable restoration, then restoration capability is improved, but storage capacity requirement increases
Solution Approach 1:
The patent uses a master virtual machine template as a copyable blueprint to create storage node virtual machines. Instead of storing complete backup data for each VM, the system maintains a single master template that can be copied to restore failed nodes, dramatically reducing storage requirements while preserving restoration capability
Solution Approach 2:
The system performs preliminary configuration by creating a master virtual machine template with all necessary software, configurations, and processes pre-configured. This template is stored in advance and can be quickly copied to restore failed storage nodes without needing to maintain extensive backup data
2Manufacturing precision
If backup data is updated every time storage node virtual machine content changes to restore to immediate pre-failure state, then restoration accuracy is improved, but storage capacity requirement increases
Solution Approach 1:
The system copies the master virtual machine template to create new storage nodes rather than maintaining incremental backup data. This copying approach ensures accurate restoration to the intended state without requiring continuous updates of backup data
Solution Approach 2:
Instead of updating backup data forward in time to capture all changes, the system inverts the approach by using a master template that defines the desired state. New storage nodes are created by copying this template, ensuring consistency without requiring continuous backup updates
3Stability of the object's composition
If backup timings are synchronized across multiple storage node virtual machines to ensure cluster consistency, then data consistency is improved, but processing complexity increases
Solution Approach 1:
The patent merges the configuration management of multiple storage nodes into a single master virtual machine template. By consolidating configuration control in one template, the system ensures cluster consistency without requiring complex synchronization of backup timings across multiple nodes
Solution Approach 2:
The master virtual machine template serves as a universal blueprint for all storage nodes in the cluster. This single template can be copied to create or restore any number of storage nodes, ensuring consistency across the entire cluster without requiring node-specific backup synchronization
Data Source
AI summary
A storage system in which a cluster is configured by a plurality of storage nodes is configured to include: a management server configured to store a template including an OS of a storage node, a process program that configures a process, and an installer; a maintenance node VM configured to collect configuration information required for configuring a restore-target storage node from within the storage system; a maintenance node, a management server, and a hypervisor configured to create a VM for a new storage node on the basis of the template and the configuration information; and a cluster controller configured to create a process in the new storage node using the installer and restore the process to a state corresponding to the process of the restore-target storage node.