Cross-Node AEP Storage Cloning via Snapshot Slicing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Apache Persistent Memory (AEP) storage volumes, used in cloud platforms, cannot perform cross-node cloning due to their local storage nature, limiting application migration and data backup capabilities.
Innovation Solution
A method involving pmem-csi, which creates AEP storage in a cluster node, monitors for clone volumes, stops and freezes the storage, creates a snapshot, and transfers the snapshot data to a clone volume, dividing data into slices for efficient transmission, and notifies administrators of any errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If AEP storage is used as local storage in Kubernetes, then storage performance is improved, but cross-node cloning capability deteriorates
Solution Approach 1:
The patent creates a snapshot copy of the AEP storage volume and transfers this copy to the target node for cloning. This allows cross-node cloning by copying data through the network rather than moving physical storage, resolving the contradiction between local storage performance and cross-node cloning capability
Solution Approach 2:
The patent introduces a storage management system as an intermediary that coordinates the cloning process between source and target nodes. This mediator manages snapshot creation, data slicing, and transmission, enabling cross-node cloning while maintaining AEP storage performance characteristics
2Adaptability or versatility
If snapshot data is transmitted across the network, then cross-node cloning is enabled, but network transmission pressure increases
Solution Approach 1:
The patent divides the snapshot data into multiple data slices before transmission. This segmentation reduces the burden on network transmission by breaking large data transfers into smaller, manageable chunks that can be transmitted efficiently and reassembled at the destination
Solution Approach 2:
The patent creates snapshots of the source storage volume before initiating data transmission. This preliminary action captures the data state in advance, allowing efficient subsequent transmission without requiring continuous synchronization during the cloning process
3Productivity
If data is divided into slices for transmission, then transmission efficiency is improved, but system complexity increases
Solution Approach 1:
The patent implements automatic data slicing and reassembly mechanisms that operate autonomously without requiring manual intervention. The system automatically divides data into slices, transmits them, and reassembles them at the destination, reducing operational complexity despite the technical sophistication of the process
Data Source
AI summary
The present application provides a method for cross-node cloning of a storage volume, a device, apparatus and a storage medium. The method includes: creating AEP storage in a node of a cluster by using pmem-csi, and monitoring whether a clone volume of the AEP storage exist in other nodes of the cluster, in response to that a clone volume of the AEP storage is monitored in the other nodes of the cluster, stopping the AEP storage and the clone volume; creating a snapshot of the AEP storage, and recovering the AEP storage; and starting the clone volume after transmitting snapshot data of the AEP storage to the clone volume. By using the solution of the present application, the problem of performing application migration and data backup of an AEP storage volume used in a cloud platform may be solved.