Shrinker: improving live migration of virtual clusters over WANs with distributed data deduplication and content-based addressing

摘要: Live virtual machine migration is a powerful feature of virtualization technologies. It enables efficient load balancing, reduces energy consumption through dynamic consolidation, and makes infrastructure maintenance transparent to users.While live available across wide area networks with state the art systems, it remains expensive use because large amounts data transfer, especially when migrating clusters rather than single instances. As evidenced by previous research, machines running identical or similar operating systems have significant portions their memory storage containing data. We propose Shrinker, system leveraging this common improve cluster between centers interconnected networks. Shrinker detects pages disk blocks duplicated in avoid sending multiple times same content over WAN links. Virtual retrieved destination site distributed content-based addressing. implemented prototype KVM hypervisor present performance evaluation environment. Experiments show that both total transferred time.