No abstract

No abstract

The lightweight nature of Operating System virtualization give birth to the containers, and their high efficiency for application deployment prefers its usage in Cloud computing. Containers encapsulate and bundle the required dependencies for application development and deployment in libs packages as a single entity. The nature of containers eases application migration within the High-Performance Computing environment to handle the scientific workloads smoothly. Singularity containers are aimed to be only for High-Performance Computing (HPC) applications. The existing HPC workload managers are good with container monitoring, scheduling, and resource management, but container orchestration is always a concern. This study proposes a framework that will ease container orchestration using Kubernetes to handle the scientific applications workload. HPC cluster is built using interconnected containers in a private cloud environment. The architecture is derived based on the required configuration for containers to deploy the application and scheduling jobs.

No abstract

Virtualization is playing as a core component in cloud computing but performance overhead impact in its one of layer is forcing the researchers to think its usage in research areas with cloud for high performance computing. Where containers usage based on operating system virtualization widespread now a days is handling the performance impact via facilitating one of the lightweight layers of virtualization in terms of computing resources such as network bandwidth, throughput, CPU, memory at a large extent. Docker is one of the popular containerized platforms to be used in cloud computing and high-performance computing research areas. In this paper, we are proposing a basic strategy to avoid the performance impact and feasibility using Docker engine along with its evaluation in terms of containers vs virtual machine. Later, we have described the containerized placement now a days and performance specific problems resolution using container images for the development and deployment of HPC applications. The results and related work are focusing on a scalable approach in terms of availability and portability.