DocumentCode
167072
Title
Considering VM migration between IaaS Clouds and mobile Clients: Challenges and Potentials
Author
Katzmarski, Bernhard ; Herrholz, Andreas ; Paolino, Michele ; Rigo, Alvise ; Nebel, Wolfgang
Author_Institution
R&D Div. Energy, OFFIS, Oldenburg, Germany
fYear
2014
fDate
8-10 Oct. 2014
Firstpage
327
Lastpage
332
Abstract
With the increasing distribution and integration of mobile devices into today´s information systems, there is a growing demand for interoperability between heterogeneous computation systems. Cloud systems already serve as an integration point between heterogeneous devices using classic client/server concepts such as Web services. However, this technology significantly limits possible applications. In this paper we consider virtual machines as a possible solution for extending the applicability for integration of mobile devices into cloud systems. We investigate the concept of migrating VM instances between clouds and local devices also considering heterogeneous computing architectures like ARM and x86. We are using an off-line approach, where a VM is transferred without being responsive. A stop, move and resume process is already sufficient for certain use cases and could increase interoperability and reduce vendor lock-in effects. We present an approach for off-line VM migration between OpenStack and local devices and evaluate our proof of concept implementation. Our results for migration costs and performance show several challenges and the need for further optimization. However, even the unoptimized migration costs might already be acceptable in certain scenarios.
Keywords
application program interfaces; cloud computing; open systems; virtual machines; IaaS clouds; OpenStack; VM instance migration; cloud systems; heterogeneous computation systems; heterogeneous computing architectures; information systems; integration point; interoperability improvement; local devices; mobile clients; mobile device distribution; mobile device integration; move process; offline VM migration; performance analysis; resume process; stop process; unoptimized migration costs; vendor lock-in effect reduction; Cloud computing; Computer architecture; Image restoration; Operating systems; Random access memory; Runtime; Virtual machining; Advanced RISC Machines (ARM); Kernel Virtual Machine (KVM); Open-Stack; Virtual machine (VM); application offloading; clouds; hybrid migration; mobile clients;
fLanguage
English
Publisher
ieee
Conference_Titel
Cloud Networking (CloudNet), 2014 IEEE 3rd International Conference on
Conference_Location
Luxembourg
Type
conf
DOI
10.1109/CloudNet.2014.6969016
Filename
6969016
Link To Document