Amol Jaikar

A performance analysis of precopy, postcopy and hybrid live VM migration algorithms in scientific cloud computing environment

Virtualization technology plays a vital role in cloud computing. One of the core features of virtualization technology is live virtual machine migration. The live migration is a process of transferring the complete state of virtual machine between physical hosts without any service interruption. This capability is being widely used for the purpose of system maintenance, load balancing, energy efficiency, reconfiguration and fault tolerance. Live migration has been extensively studied for commercial workloads and the ongoing research is mainly focusing on the performance improvements of well-known live migration algorithms. Today, scientific communities are actively thinking to take the advantage of cloud computing for scientific workloads. In this paper, we analyze the performance of well-known precopy, postcopy and hybrid live migration algorithms and examine the migration times of VMs running high throughput computing (HTC) jobs in a scientific cloud computing environment. The results of our research not only show the performance comparison of live migration algorithms but also will be helpful when selecting a live migration algorithm in scientific cloud computing environment.

Power efficient virtual machine migration in a scientific federated cloud

Virtualization technology enables cloud computing to provide virtual infrastructure. This virtual infrastructure is managed by centralized entity which is responsible for creation, deletion and migration of the virtual machine. Migration module helps to reduce the energy consumption using server consolidation which means multiple underutilized virtual machines can be deployed on a single physical machine. In such an environment, migration module moves the virtual machine from one host to another host to improve utilization. Therefore, migration algorithm plays an important role in the management of virtual cluster and power consumption. In this paper, we have proposed a slot-based migration approach to manage the virtual infrastructure using space-sharing technology of virtualization with the solution of multiple hops problem. The experimental results show that the migration module reduces considerable amount of power consumption.

Priority-based virtual machine load balancing in a scientific federated cloud.

Cloud computing offers utility oriented services to users. It also reduces the maintenance cost of the infrastructure as well as initial investment cost for users or startup companies. Due to the effectiveness of these services, many companies or institutes are trying to adopt this technology for offloading scientific or industrial applications. Virtualization in cloud computing is used to build virtual infrastructure which is managed by a load balancer. Efficient design of the load balancer can reduce over or under provisioning. Load balancer contains virtual machine allocation technique which performs an important role with respect to power consumption of the data center. In this paper, we propose a system architecture and the virtual machine allocation algorithm for the load balancer in a scientific federated cloud. We have tested the proposed approach in a scientific federated cloud. Experimental results show that the proposed algorithm not only increases the utilization of resources but also reduces the energy consumption.

An adaptive load monitoring solution for logically centralized SDN controller

The recent emerging concept in the networking world is known as SDN (Software-Defined Network). SDN architecture introduced the new methods of network management and configuration. It makes underlying networking devices as simply just packet forwarding units and leaving the control logic on centralized software program to dictate the entire network behavior. In logically centralized SDN architecture the switches are statically assigned to controllers regardless of actual load of individual controllers. One of the key concern of network operators is to efficiently monitor the actual load of individual controllers in a logically centralized SDN environment. In this paper, we propose an adaptive solution to monitor the actual load of a SDN controller for efficient utilization in a logically centralized SDN architecture. We proposed a load monitoring solution which is combination of ONOS (Open Network Operating System) based collection manager and collectd based “onos write” plugin/module that act as an agent on SDN controllers. Our proposed solution helps to overcome the under-utilization and over-utilization of a controller implemented in logically centralized SDN environment. The results show that our proposed monitoring system gives a comprehensive statistical view of individual controllers load.

Cloud Computing: Read Before Use

Cloud computing is evolving as a new paradigm in service computing in order to reduce initial infrastructure investment and maintenance cost. Virtualization technology is used to create virtual infrastructure by sharing the physical resources through virtual machine. By using these virtual machines, cloud computing technology enables the effective usage of resources with economical profit for customers. Because of these advantages, scientific community is also thinking to shift from grid and cluster computing to cloud computing. However, this virtualization technology comes with significant performance penalties. Moreover, scientific jobs are different from commercial workload. In order to understand the reliability and feasibility of cloud computing for scientific workload, we have to understand the technology and its performance. In this work, we have evaluated the scientific jobs as well as standard benchmarks on private and public cloud to understand exact performance penalties involved in adoption of cloud computing. These jobs are categorized into CPU, memory, N/W and I/O intensive. We also analyzed the results and compared the private and public cloud virtual machines performance by considering execution time as well as price. Results show that the cloud computing technology faces considerable performance overhead because of virtualization technology. Therefore, cloud computing technology needs improvement to execute scientific workload.

NOVEL: NO-VowEL Technique to Search Fuzzy Keyword

Information retrieval techniques play vital role in the era of information technology. Inverted index is one of the technique to retrieve the information/data related with certain keyword. This technique gives faster results to retrieve relevant document from billions of documents, which contains specified keyword. In order to support wrongly spelled keyword, many techniques have been proposed including edit distance, wild-card and n-gram. The n-gram index has language-neutral and errortolerant advantage. However, it has a drawback of large size and less performance. In this paper, we have proposed NOVEL technique to search fuzzy keyword. We have implemented and tested the proposed technique on two datasets. The result shows that NOVEL technique supports not only wrongly spelled keywords, but also reduced gram size by 40-50% than K/n-gram technique. Therefore, the proposed technique is the most efficient technique to support fuzzy keyword search.

Performance Evaluation of Scientific Workflow on OpenStack and OpenVZ

Cloud computing is capturing attention of the market by providing infrastructure, platform and software as a services. Using virtualization technology, resources are shared among multiple users to improve the resource utilization. By leasing the infrastructure from public cloud, users can save money and time to maintain the expensive computing facility. Therefore, it gives an option for cluster and grid computing technology which is used for industrial application or scientific workflow. Virtual machine enables more flexibility for consolidation of the underutilized servers. However, containers are also competing with virtual machine to improve the resource utilization. Therefore, to adopt cloud computing for scientific workflow, scientist needs to understand the performance of virtual machine and container. We have used cloud computing with different virtualization technologies like KVM and container to test the performance of scientific workflow. In this work, we analyze the performance of scientific workflow on OpenStack’s virtual machine and OpenVZ’s container. Our result shows that container gives better and stable performance than virtual machine.