the capacity of the cloud is allowing users to deploy complex infrastructure on the cloud due to huge pool of the resources offered by the cloud. In the infrastructure as service virtual machines are provided to the end users as the mean of the resource from the cloud. Complex application such as load balancing system or cluster system requires multiple virtual machines and their special connectivity for system realization. Current cloud providers do not allow users to deploy their complex workload. Users will need to deploy these virtual machines manually by selecting one at a time. In this paper we are proposing hint based virtual placement which accepts hints from the users and select the nodes from cloud. This system act as middleware between cloud provider and users and automate the complex workload deployment process. This system mutually understands the constraints placed by both users and cloud admins while deploying virtual machines. Such complex virtual machine deployment can be passed to the system using constraint like AssociateVM, DistantVM, BackupVM and PrivateNetVM. Cloud admins can also place constraints like out of service, backup node etc to influence the virtual machine placement. Our virtual machine scheduler considers these constraints from both the ends and selects the appropriate nodes on the basis of the constraints and scores assigned to the nodes. We have compared this strategy with available known cloud scheduler algorithms and found that available strategies do not allow any user passed hints while our algorithm allow users to pass such hints while deploying virtual machines.

Public Cloud; Virtual Machine Placement; Resource allocation; User Hints

[1] Konstantinos Tsakalozos, Mema Roussopoulos, and Alex Delis "Hint-Based Execution of Workloads in Clouds with Nefeli" IEEE transactions on parallel and ditributed systems, vol. 24, no. 7, July 2013
[2] Bobroff, N. ; T.J. Watson Res. Center, IBM, Hawthorne, NY ; Kochut, A. ; Beaty, K. "Dynamic Placement of Virtual Machines for Managing SLA Violations" Integrated Network Management, 2007. IM '07. 10th IFIP/IEEE
[3] Chaisiri, S. ; Sch. of Comput. Eng., Nanyang Technol. Univ., Singapore, Singapore ; Bu-Sung Lee ; Niyato, D. "Optimal virtual machine placement across multiple cloud providers" Services Computing Conference, 2009
[4] Remesh Babu, K.R. ; Dept. of Inf. Technol., Gov. Eng. Coll., Idukki, India ; Samuel, P. "Virtual Machine Placement for Improved Quality in IaaS Cloud" Advances in Computing and Communications (ICACC), 2014
[5] Jamali, S. ; Dept. of Electr. & Comput. Eng., Univ. of Mohaghegh Ardebili, Ardebil, Iran ; Malektaji, S. "Improving grouping genetic algorithm for virtual machine placement in cloud data centers" Computer and Knowledge Engineering (ICCKE), 2014
[6] Ricardo Stegh Camati, Alcides Calsavara, Luiz Lima Jr. "Solving the Virtual Machine Placement Problem as a Multiple Multidimensional Knapsack Problem" ICN 2014 : The Thirteenth International Conference on Networks
[7] Nskinc study on cloud computing, Nskinc white paper 2012
[8] Linlin Wu ; Dept. of Comput. & Inf. Syst., Univ. of Melbourne, Melbourne, VIC, Australia ; Garg, S.K. ; Versteeg, S. ; Buyya, R. "SLA-Based Resource Provisioning for Hosted Software-as-a-Service Applications in Cloud Computing Environments" Services Computing, IEEE Transactions on (Volume:7 , Issue: 3 )
[9] Dong Huang ; Inst. for Infocomm Res., Agency for Sci. Technol. & Res, Singapore, Singapore ; Bingsheng He ; Chunyan Miao "A Survey of Resource Management in Multi-Tier Web Applications
[10] Pankaj R Kadam, Nilesh V Alone "Review on KVM Hypervisor" International Journal of Recent Technology and Engineering (IJRTE) ISSN: 2277-3878, Volume-3 Issue-4 September, 2014
[11] Mell, P. and Grance, T. 2011. "The NIST Definition of Cloud Computing" NIST Special Publication, 800-145