Authors: Chris Nero; Anindya Neogi; Venkata Ramesh Somisetty
Addresses: IBM/Tivoli Systems, 3901 S. Miami Blvd., Durham, NC 27703, USA ' IBM India Software Lab, ETZ, Rajiv Gandhi Infotech Park Ph 2, Pune 411057, India ' IBM India Software Lab, B176, 3rd Floor, MSP1, MINDSPACE, HI-TECH City Hyderabad, AP, 500081, India
Abstract: It has been predicted in a cloud management survey, that over 68% of production workloads in the USA will be virtualised by 2013, which is almost 25% more than in 2010 (IDC Private Cloud Management Study). However, virtualisation will not deliver its full potential unless it is carefully managed to achieve the adequate consolidation density within acceptable performance risk. In this paper, we describe an interactive, what-if analysis driven, planning tool to manage and optimise virtualised/cloud environments. The analysis is performed based on usage data, discovered configuration, policies and other meta-data. The recommended optimisation plan may suggest possible savings based on various cost factors. The paper also describes a case study of how we have applied the tool in an IBM internal cloud and present some insights and learning. Based on several what-if analysis scenarios we demonstrated, 100+ more virtual machines can be accommodated in the existing cloud, resulting in improved ROI.
Keywords: cloud computing; optimisation; virtualisation; capacity planning; infrastructure as a service; IaaS; software as a service; SaaS; platform as a service; PaaS; what-if analysis; return on investment; ROI; case study; cloud infrastructure; virtual machines.
International Journal of Cloud Computing, 2014 Vol.3 No.2, pp.191 - 208
Available online: 03 Jun 2014 *Full-text access for editors Access for subscribers Purchase this article Comment on this article