Optimized Security-aware VM placement algorithm

The rapidly increasing dependency on use of clouds results in Cloud Service Providers (CSPs) having to deal with high cloud services demands. To meet these demands, CSPs take advantage of virtualization technology to provide a seemingly unlimited pool of computing resources. This technology consolidates multiple instances of Virtual Machines (VMs) into the same Physical Machines (PMs) and share physical computing resources. To guarantee customer satisfaction, CSPs need to ensure optimized cloud environment that provides good Quality of Services (QoS) which conform to the performance levels stipulated in Service Level Agreements (SLAs). However, vulnerabilities associated with virtualization make it difficult to ensure optimization, more especially in multi-tenant clouds. In multi-tenant clouds, there are possibilities of consolidating VMs belonging to adversary users into the same PMs. This promotes inter-VM attacks that take advantage of shared resources to either spy, disrupt or corrupt co-located VMs. With this regard, it is important to consider placement of VMs in a manner that minimizes inter-VM attacks. This placement must, however, ensure initial objectives of providing good QoS. The aim of this study is to implement a VM placement algorithm that reduces architectural vulnerabilities brought by multi-tenancy while observing optimization objectives. It focuses on currently available VM placement algorithms and evaluates them to identify the algorithm that assumes highest optimization objectives. The identified VM placement algorithm is further augmented with security features to implement Optimized Security-aware (O-Sec) VM Placement algorithms. CloudSim Plus is used to evaluate and validate the implemented O-Sec VM placement algorithms. The evaluations in this study show that O-sec VM placement algorithm retains optimization objectives inherited from the identified VM placement algorithm. This is an algorithm that is augmented towards O-sec VM placement algorithm.dissertation (MSc)--University of Pretoria, 2019.Computer ScienceMScUnrestricte