Fast Wide Area Live Migration with a Low Overhead through Page Cache Teleportation

Akiyama, Soramichi; Hirofuchi, Takahiro; Takano, Ryousei; Honiden, Shinichi

doi:10.1109/ccgrid.2013.57

Cited by 12 publications

(4 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2 The problems that did turn out to have a meaningful effect and that we were able to address are listed next.…”

Section: Problems In Baseline Swappingmentioning

confidence: 97%

“…VSWAPPER techniques may be used to enhance live migration of guests and reduce the migration time and network traffic by avoiding the transfer of free and clean guest pages. Previous research suggested to achieve this goal via guest cooperation by inflating the balloon prior to live migration [26], or by migrating non-page-cache pages first [2]. The Mapper and the Preventer techniques can achieve the same goal without guest cooperation.…”

Section: Future Workmentioning

confidence: 98%

“…Hosts cannot rely on guest cooperation, because: (1) clients may have disabled or opted not to install the balloon [10,53,74]; (2) clients may have failed to install the balloon due to technical difficulties [77,78,79,80,81,82,83]; (3) balloons could reach their upper bound, set by the hypervisor (and optionally adjusted by clients) to enhance stability and to accommodate various guest limitations [14,19,59,69,75,84]; (4) balloons might be unable to reclaim memory fast enough to accommodate the demand that the host must satisfy, notably since guest memory swapping involves slow disk activity [31,46,84]; and (5) balloons could be temporarily unavailable due to inner guest activity such as booting [84] or running high priority processes that starve guest kernel services. In all these cases, the host must resort to uncooperative swapping, which is notorious for its poor performance (and which has motivated ballooning in the first place).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

VSwapper

Amit

Tsafrir

Schuster

2014

Proceedings of the 19th International Conference on Architectural Support for Programming Languages and Operating Systems

View full text Add to dashboard Cite

The number of guest virtual machines that can be consolidated on one physical host is typically limited by the memory size, motivating memory overcommitment. Guests are given a choice to either install a "balloon" driver to coordinate the overcommitment activity, or to experience degraded performance due to uncooperative swapping. Ballooning, however, is not a complete solution, as hosts must still fall back on uncooperative swapping in various circumstances. Additionally, ballooning takes time to accommodate change, and so guests might experience degraded performance under changing conditions.Our goal is to improve the performance of hosts when they fall back on uncooperative swapping and/or operate under changing load conditions. We carefully isolate and characterize the causes for the associated poor performance, which include various types of superfluous swap operations, decayed swap file sequentiality, and ineffective prefetch decisions upon page faults. We address these problems by implementing VSWAPPER, a guest-agnostic memory swapper for virtual environments that allows efficient, uncooperative overcommitment. With inactive ballooning, VSWAPPER yields up to an order of magnitude performance improvement. Combined with ballooning, VSWAPPER can achieve up to double the performance under changing load conditions.

show abstract

“…2 The problems that did turn out to have a meaningful effect and that we were able to address are listed next.…”

Section: Problems In Baseline Swappingmentioning

confidence: 97%

Section: Future Workmentioning

confidence: 98%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

VSwapper

Amit

Tsafrir

Schuster

2014

Proceedings of the 19th International Conference on Architectural Support for Programming Languages and Operating Systems

View full text Add to dashboard Cite

show abstract

“…Zongjian He et.al [19] evaluated the security and performance, quality of service of popular virtual switch software Open vSwitch. In this paper, the evaluation environment has been implemented and conducted different tests on Open vSwitch including VLAN test, Qos test and overhead evaluation.…”

Section: Literature Reviewmentioning

confidence: 99%

Advanced Memory Reusing Mechanism for Virtual Machines in Cloud Computing

Singh

Behal

Taneja

2015

Procedia Computer Science

View full text Add to dashboard Cite

Cloud computing is an emerging computing technology for large data centers that maintains computational resources through the internet, rather than on local computers. With the increasing popularity of cloud computing, also increase the demand of cloud resources. In Infrastructure-as-a-Service datacenters, the placements of Virtual Machines (VMs) on physical hosts are dynamically optimized in response to resource utilization of the hosts. VM migration provides the capability to balance the load, system maintenance and fault tolerance etc. However, existing migration techniques, used to migrate virtual machines keeping data images of VMs in host and skipping transfer of unchanged data fields to reduce the amount of transfer data during migration, if number of migrations increases, number of data images stored on host are also increased, this causes memory starvation. In this paper, we propose a technique that reduces the size of data image stored on source host before migration. When a Virtual Machine migrates to another host, the data image for that VM is kept in the source host after removing unwanted data according to the probability factor. When the VM migrates back to the original host later, the kept memory image will be "reused", i.e. data which are identical to the kept data will not be transferred and comparative to existing system the size of memory image is small. To validate this approach, results evaluated using different threshold levels and probability factor of change in data. Proposed system required less memory to store the memory image and allow more VMs to be hosted. Specifically, our proposed work is used to improve resource efficiency throughout by reducing the size of memory image that is stored on source host. Evaluations show that size of memory image reduced 33% (approx) of unnecessary memory consumption.

show abstract