Implements Optimal Combinations of New Techniques, Lowers Costs in Heterogeneous Storage Environments
Virtual machine live storage migration is essential in the era of big data and cloud computing. Major providers in cloud computing – a market estimated to hit $241 billion by 2020 – offer virtual machines. The cornerstone for virtual machine moving is live storage migration, which transfers virtual disk volumes of virtual machines between physical storage media. Live storage migration techniques encounter several problems in heterogeneous storage environments, including speed discrepancies and shortening the lifespan of solid-state drives (SSD). This software developed by University of Florida researchers adaptively mixes storage migration strategies to improve the efficiency of live data migration. By implementing optimal combinations of new live storage migration techniques, the algorithm lowers the cost of massive migration while enhancing performance, user experience, and reliability of virtual machine migrations in heterogeneous storage environments.
Data migration algorithm improves efficiency, mobility and manageability of big data cloud computing in heterogeneous storage environments.
- Preserve lifetime of solid-state drives, decreasing costs of massive migration
- Leverages speed discrepancies, increasing efficiency
- Enhances performance of systems, improving user experience
This algorithm addresses the efficiency of adaptive storage migration strategies in heterogeneous storage by combining the migration techniques of Low Redundancy (generates the least amount of redundant writes), Source-based Low Redundancy (keeps the balances between input/output performance and write redundancy), and Asynchronous IO Mirroring (seeks the highest input/output performance). Adaptively combining these migration techniques improves user experience, solid-state drive wearing, and manageability of massive migration. This algorithm lowers the cost of massive virtual machine live storage migration as compared to using the best of an individual mechanism.