Virtual Machine Scheduling for Architectural Shared Resources in Open Stack Based Cloud
Jayalakshmi N1, Sagar B M2
1Jayalakshmi N, MTECH [SE], 4th Semester, Department of Information Science and Engineering, R V College Of Engineering, India, Bangalore–560059,India.
2Sagar B M, Associate Professor Department of Information Science and Engineering, RV College Of Engineering ,India, Bangalore–560059, India.
Manuscript received on May 11, 2013. | Revised Manuscript received on May 15, 2013. | Manuscript published on May 25, 2013. | PP: 85-87 | Volume-1 Issue-7, May 2013. | Retrieval Number: G0313051713/2013©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In IAAS cloud virtual machine (VM) migration between socket and nodes has been used to avoid conflicts among VM on system resources such as CPU and memory. Micro-architectural resources such as caches, memory controllers and non-uniform memory access (NUMA) affinity, have relied on intra socket scheduling to reduce cache contention. This paper proposes algorithm for cluster-level virtual machine scheduling based on cache sharing on same socket and these are considered in dynamic environment which do not require any prior knowledge on nature of VMs.The paper would present algorithm for the scheduler into a real cloud system (Open Stack), and the performance of the scheduler will then be investigated on various scientific and commercial workloads.
Keywords: architecture, NUMA, OpenStack , CPU bound, memory bound.