Main Article Content
Abstract
The cloud resource provisioning is carried out as services. The resources are provided with user requests. The enterprise applications are build with the support of the service components. The Virtual Machine (VM) allocation for the service components are carried out with resource capacity levels. The process of Virtual Machines to the service components is called as service consolidation or Virtual Machine consolidation process. The dependant service components are services that are used in the same application. The network delay and communication cost is reduced with the support of the intelligent service management models and service dependencies. The enterprise applications are build with services under many hosts and devices. The resource, services and network constraints are considered in the service consolidation process. The non intrusive dependant service component discovery is achieved with CloudScout model. The service component relationship is analyzed with the time series details collected from the historical data maintained under the logs. The privacy preserved service dependencies are identified with log mining models. The service distance estimation is carried out with the weight of resource metric derived from the iEntropy method. The dependant service clustering process is carried out with the hierarchical and iterative k-means (HiKM) algorithm. The virtual machine consolidation operations are performed with queue network and network latency optimization process. The optimal service dependency discovery and service consolidation operations are integrated with the CloudScout scheme. The sequential dependencies and complimentary dependencies are combined in the service dependency analysis. The service components are assigned to the virtual machines using the Multi Objective Service
Consolidation (MOSC) scheme. The resource level constraints are also analyzed in the service consolidation process.