Virtual Machine Migration via Hash-Based Data Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtual machine migration techniques are inefficient due to high network resource usage and time consumption, particularly when transferring virtual machine images between different cloud deployments such as public, private, and hybrid clouds.
Innovation Solution
The technology employs data suppression techniques by dividing virtual machine images into data chunks, computing hash identifiers for each chunk, and using prepopulated hash databases in both source and destination data centers to transfer only hash identifiers instead of actual data, reducing network resource usage and transfer time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If virtual machine images are transferred directly between data centers, then complete migration is achieved, but network resource usage and migration time increase significantly
Solution Approach 1:
The virtual machine image is divided into multiple data chunks, each with its own hash identifier. This segmentation allows the system to process and transfer only the necessary chunks rather than the entire image, reducing migration time while maintaining completeness.
Solution Approach 2:
Hash databases are prepopulated in both source and destination data centers with hash identifiers of existing data chunks. This preliminary action enables quick comparison and identification of matching chunks, allowing the system to skip transferring duplicate data and significantly reduce migration time.
2Reliability
If virtual machine images are transferred directly between data centers, then complete migration is achieved, but network resource usage increases
Solution Approach 1:
The patent extracts only the hash identifiers from the virtual machine image data for transfer, rather than transferring the actual image data. This extraction approach dramatically reduces network resource usage while still enabling complete migration through hash matching and data reconstruction.
Solution Approach 2:
Instead of copying actual image data, the system copies hash identifiers which serve as references. The destination data center uses these hash identifiers to locate and retrieve existing data chunks from its prepopulated hash database, eliminating the need to transfer duplicate data over the network.
3Productivity
If hash databases are prepopulated in both source and destination data centers, then transfer efficiency improves, but device complexity increases
Solution Approach 1:
The hash database serves multiple functions: it stores data chunk references for migration, enables quick data matching, and facilitates efficient data retrieval. This multi-functionality justifies the additional device complexity by providing significant productivity improvements across multiple operations.
Data Source
AI summary
The subject technology addresses the need in the art for improving intra-cloud migration of virtual machines in a cloud computing environment. A hash database may be prepopulated with key-value pairs corresponding to hash IDs and associated data chunks of a virtual machine image. In this regard, the virtual machine image may be divided into chunks using boundaries chosen by a Rabin fingerprinting technique. A hash (e.g., MD5 or SHA-1) may be computed over each chunk and act as a unique identifier for the data contained in each chunk. At appropriate times, one or more hash IDs are sent instead of the actual data chunks between clouds when performing the inter-cloud migration of a virtual machine.