Agile Live Migration of Virtual Machines

Deshpande, Umesh; Chan, Danny; Guh, Ten-Young; Edouard, James; Gopalan, Kartik; Bila, Nilton

doi:10.1109/ipdps.2016.120

Cited by 25 publications

(7 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The memory migration techniques aim at continuing the current processing of those VMs which have to be migrated, and to achieve this goal, the memory contents of these VMs are also transferred; along with the memory content, CPU current state is also migrated. Many variations of memory migration techniques have been presented in the literature recently: hybrid memory copyoriented technique [3]; efficiency-oriented technique [4]; multiple application-oriented technique [5]; workload prediction-oriented technique [6]; NFV featured network-oriented technique [7]; Traffic sensitive technique [8]; Agile live migration technique [9]; Hash-Table oriented technique [10]; VNF live migration technique [11]. However, these techniques do not specifically address the issue of VM migration in DCCs by considering VM workload characteristics.…”

Section: Related Workmentioning

confidence: 99%

Meta-heuristic Techniques for Placement of Virtual Machines in Distributed Cloud Centers

N¹,

Nair²

2019

IJIES

View full text Add to dashboard Cite

The Virtual Machine Placement techniques provide an attractive opportunity to achieve the goal of energy conservation in Cloud Centers. The Virtual Machines can be broadly classified into data intensive and CPU intensive based on their workload characteristics. Placing or migrating large number of data intensive Virtual Machines can lead to degradation of task execution efficiency in Cloud Centers. Hence, the Virtual Machine Placement techniques need to consider the Virtual Machine workload characteristics to prevent associated negative consequences. Even though workload characteristics aware Virtual Machine migration technique has been presented in the literature for single-location Cloud Centers, it does not cater to Distributed Cloud Centers, where the cloud resources are distributed in different geographical locations, and in such setting, new performance issues arise which have to be effectively addressed. In this work, the NP-Complete Bin Packing Problem framework is extended to model the Workload Characteristic aware Virtual Machine Placement problem in Distributed Cloud Centers. The proposed solution is based on two meta-heuristic techniques having polynomial complexity: Particle Swarm Optimization and Ant Colony Optimization technique. The proposed techniques are compared against the contemporary technique in simulated environment. In the simulated study of the proposed techniques, both these techniques discover approximate solutions to the Bin Packing Problem which are extremely close to the optimal solution, and thereby directly contribute in efficient load distribution in Distributed Cloud Centers.

show abstract

Section: Related Workmentioning

confidence: 99%

Meta-heuristic Techniques for Placement of Virtual Machines in Distributed Cloud Centers

N¹,

Nair²

2019

IJIES

View full text Add to dashboard Cite

show abstract

“…In [31], the basic memory to properly run a VM is estimated to provide fast VM migration. To do so, the VM memory size is constrained with cgroups, then swapped-in pages are marked as hot pages and will build the minimum working set memory of a VM.…”

Section: Related Workmentioning

confidence: 99%

“…SEaMLESS, however, provides a fast estimation of the hot pages by enabling a dummy restore Gateway Process procedure. [31] will also suffer from slow service restoration if the required service hits several cold pages, which is not the case of SEaMLESS where the idle VM and its original memory environment is restored in case of user activity.…”

Section: Related Workmentioning

confidence: 99%

Towards Massive Consolidation in Data Centers with SEaMLESS

Segalini

Pacheco

Jacquemart

et al. 2018

2018 18th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID)

View full text Add to dashboard Cite

In Data Centers (DCs), an abundance of virtual machines (VMs) remain idle due to network services awaiting for incoming connections, or due to established-and-idling sessions. These VMs lead to wastage of RAM-the scarcest resource in DCs-as they lock their allocated memory. In this paper, we introduce SEaMLESS, a solution designed to (i) transform fully-fledged idle VMs into lightweight and resourceless virtual network functions (VNFs), then (ii) reduces the allocated memory to those idle VMs. By replacing idle VMs with VNFs, SEaMLESS provides fast VM restoration upon user activity detection, thereby introducing limited impact on the Quality of Experience (QoE). Our results show that SEaMLESS can consolidate hundreds of VMs as VNFs onto one single machine. SEaMLESS is thus able to release the majority of the memory allocated to idle VMs. This freed memory can then be reassigned to new VMs, or lead to massive consolidation, to enable a better utilization of DC resources.

show abstract

“…Jettison [2] proposes partial VM migration in which only the working set of an idle VM is transferred to the destination, while the remaining memory pages are slowly demand-paged from the source. Agile [7] migration reduces the migration time by swapping the non-working set pages of a VM to a per-VM swap device and migrating only the working set pages using a hybrid pre/post-copy technique. Vecycle [18] stores a checkpoint at each machine which a VM has visited in the past.…”

Section: Related Workmentioning

confidence: 99%

Quick eviction of virtual machines through proactive live snapshots

Fernando

Bagdi

Yao-hui

et al. 2016

Proceedings of the 9th International Conference on Utility and Cloud Computing

Self Cite

View full text Add to dashboard Cite

Cloud computing platforms routinely use virtualization to improve service availability, resiliency, and flexibility. Live migration of Virtual Machines (VM) is a key technique to quickly migrate workloads in response to events such as impending failure or load changes. Despite extensive research, state-of-the-art live migration approaches take a long time to migrate a VM (in the order of tens of seconds for moderately sized VMs) which in turn negatively impacts the application performance during migration. We present Quick Eviction, a new approach to significantly speed up the eviction of a VM from the source host with low impact on VM's performance during migration. Our approach can also improve the effectiveness of VM resilience tools that back up a VM's state in anticipation of system failures. The insight behind Quick Eviction is that the majority of time to evict a VM during migration is due to the transfer of memory contents over the network, sometimes repeatedly. Before migration, Quick Eviction regularly snapshots the VM's memory to a destination or a failover node. During the actual migration, Quick Eviction has to transfer only a small amount of dirtied memory resulting in a very short time to completely evict the VM out of the source. The key challenge is to dynamically adapt the snapshot intervals so as to have minimal impact on application performance during migration. We show that our implementation of Quick Eviction in the *

show abstract

Agile Live Migration of Virtual Machines

Cited by 25 publications

References 17 publications

Meta-heuristic Techniques for Placement of Virtual Machines in Distributed Cloud Centers

Meta-heuristic Techniques for Placement of Virtual Machines in Distributed Cloud Centers

Towards Massive Consolidation in Data Centers with SEaMLESS

Quick eviction of virtual machines through proactive live snapshots

Contact Info

Product

Resources

About