Congfeng Jiang

A Survey of Virtual Machine System: Current Technology and Future Trends

With the development of the computer technology, the virtual machine has been become the main research topic. Understanding of the current technology and future trends of virtual machine system greatly help to improve the service performance of system. Therefore, we describe the current technology and present the future trends of virtual machine system in the paper. In the current technology of virtual machine system, we mainly describe the virtualization technology, the resource scheduling technology, the migration technology, the security technology and the performance evaluation technology. In the future trends of virtual machine system, we mainly present an overview of the future CPU architecture, the management mode of future memory and resource, the future maintaining method of system security and the performance evaluation method of future multiple virtual machine system.

ARAS-M: Automatic Resource Allocation Strategy based on Market Mechanism in Cloud Computing

Resource management is one of the main issues in Cloud Computing. In order to improve resource utilization of large Data Centers while delivering services with higher QoS to Cloud Clients, an automatic resource allocation strategy based on market Mechanism ( A RAS-M) is proposed. Firstly, the architecture and the market model of A RAS-M are constructed, in which a QoS-refectitive utility function is designed according to different resource requirements of Cloud Client . T he equilibrium state of A RAS-M is defined and the proof of its optimality is given. Secondly, A Genetic Algorithm (GA) -based automatic price adjusting algorithm is introduced to deal with the problem of achieving the equilibrium state of A RAS-M. Finally, A RAS-M is implemented on Xen. Experiment results show that A RAS-M can approximately achieve the equilibrium state, that is, demand and supply is nearly balanced, which validate s that A RAS-M is effective and practicable , and is capable of achieving resource balance in cloud computing .

A Method for Measuring Semantic Similarity of Concepts in the Same Ontology

The present methods for measuring concepts semantic similarity only focus on certain influencing factors, have poor convergence performances and canpsilat calculate accurately. This paper compares three kinds of ontology-based semantic similarity calculation models. On this basis, an improved algorithm that inherits the distance-based calculation model is proposed. In this approach, node depth, local density and node attributes are newly quantified and the granularity degree of clusters is firstly combined with other five factors: local density, node depth, link type, link strength, node attribute. The experimental results show that this method provides an effective quantification for the semantic relationships, and can calculate semantic similarity more precisely.

VRAA: virtualized resource auction and allocation based on incentive and penalty

Virtualization is widely used in cloud computing environments to efficiently manage resources, but it also raises several challenges. One of them is the fairness issue of resource allocation among virtual machines. Traditional virtualized resource allocation approaches distribute physical resources equally without taking into account the actual workload of each virtual machine and thus often leads to wasting. In this paper, we propose a virtualized resource auction and allocation model (VRAA) based on incentive and penalty to correct this wasting problem. In our approach, we use Nash equilibrium of cooperative games to fairly allocate resources among multiple virtual machines to maximize revenue of the system. To illustrate the effectiveness of the proposed approach, we then apply the basic laws of auction gaming to investigate how CPU allocation and contention can affect applications’ performance (i.e., response time), and its effect on CPU utilization. We find that in our VRAA model, the fairness index is high, and the resource allocation is closely proportional to the actual workloads of the virtual machines, so the wasting of resources is reduced. Experiment results show that our model is general, and can be applied to other virtualized non-CPU resources.

Resource Allocation in Contending Virtualized Environments through VM Performance Modeling and Feedback

With active deployment of virtualization in large scale data centers and cloud computing environments, allocation and scheduling of virtual and physical resources raise more challenges and may have negative impacts on system performance due to: (1) the isolation between the guest Virtual Machines (VMs) and the Virtual Machines Monitor (VMM), and (2) the independent and even conflicting operations between multiple VMs. In this paper a stochastic model of resources in virtualized environments is proposed and resource allocation and scheduling algorithm are proposed to provide performance guarantees and service differentiation in contending conditions. In the proposed algorithm user behavior and workloads are characterized through the historical and real time performance profiling and estimation from hosted agents within individual VMs. The resources are allocated according to the demand as well as the performance of the targeted VMs based on the Suffer age aggregation and performance feedback. Experiments on a real Xen based virtualization environment with 3 VMs are conducted and evaluated for accuracy, efficiency, sensitivity, and overhead. The results show that the performance feedback based allocation can achieve a higher SLA satisfaction rate as 97.5%, a lower load imbalance index as 17.6%. The results also show that this algorithm is valid, effective and scalable for implementation in real virtualized environments.

Power aware job scheduling with QoS guarantees based on feedback control

With the scale of computing system increases, power consumption has become the major challenge to system performance, reliability and IT management costs. Specifically, system performance and reliability, described by various Quality of Service(QoS) metrics, cannot be guaranteed if the objective is to minimize the total power consumption solely, despite of the violations of QoS. Various methods have been developed to control power consumption to avoid system failures and thermal emergencies through coarse-grained designs. However, the existing methods can be improved and more power can be saved if fine-grained job level adaptation is integrated into them. In this paper a feedback control based power aware job scheduling algorithm is proposed to minimize power consumption in computing system and to provide QoS guarantees. In the proposed algorithm, jobs are scheduled according to the realtime and historical power consumption as well as the QoS requirements. Simulations and experiments on real multi core computing system show that the power potential of the system can be deeply explored while still providing QoS guarantees and the performance degradation is acceptable. The experiment results also show that fine-grained job-level power aware scheduling can achieve better power/performance balancing between multiple processors or cores than coarse-grained methods.

Dynamic voltage/frequency scaling for power reduction in data centers: Enough or not?

Power consumption is becoming a critical and annoying problem to Data Centers (DCs). Higher power consumption results in more heat dissipation, cooling costs and makes servers more prone to failures. Various excellent workload and application-specific dynamic voltage/frequency scaling (DVS/DFS) algorithms have been proposed and deployed in many computing systems for power reduction and system reliabilities. However, there also exist some disadvantages associated with individual DVS/DFS algorithms when they are ported to DCs, where virtualization technologies and emerging multi-core processors are widely deployed. In this paper, we discuss several aspects and characteristics of DVS/DFS algorithms, e.g. performance, overheads, feasibility and usability in Data Centers. This paper also presents a discussion on the challenges of power reduction in large Data Centers.

Analysis and Evaluation of the Scheduling Algorithms in Virtual Environment

Virtual machine technologies currently receive great interest both in industry and research communities. And it is one of the most important technologies for the coming Cloud Computing. We surveyed the CPU scheduling algorithms in Xen and VMWare systems, and found that both of them use a distinctive VCPUs running queue for each physical CPU, which is referred to Partition Queue Model(PQM) in this paper. As a contrast, a Sharing Queue Model (SQM) of CPU scheduling algorithm is proposed. The simulation experiments results show that the Sharing Queue Model of CPU scheduling achieves better performance than the Partition Queue Model and the deduction from Queue Theory also confirms the results. Moreover, with the CPU utilization increased, the advantage of Sharing Queue Model over Partition Queue Mode is more evident in response time.

Efficient parallel implementation of incompressible pipe flow algorithm based on SIMPLE

Parallel semi‐implicit method for pressure‐linked equations(SIMPLE) algorithm is used to solve the 3‐D incompressible pipe flow problem. In this paper, we proposed a novel parallel SIMPLE algorithm that uses the alternate tiling technique. Firstly, a parallel SIMPLE algorithm based on domain decomposition method was established, and the implementation of domain partition and data exchange was presented. Then, we presented serial finite difference stencil algorithm based on alternate tiling. Furthermore, an iteration space parallel two‐way finite difference stencil algorithm based on alternate tiling was proposed, introducing the sequence of iterative space tiles as the sequence of execution and using time skewing technique to partition the iteration space, thus to improve the data locality of algorithm. The cache misses and the cost of communication and synchronization are reduced by reordering the tiles of iteration space. Finally, the effectiveness of the two parallel SIMPLE algorithms were compared. The results showed that the parallel SIMPLE algorithm that uses the two‐way finite difference stencil algorithm based on alternate tiling has good data locality, performance, and scalability in the Deepcomp7000 cluster computing environment. Copyright

Power Aware Resource Allocation in Virtualized Environments through VM Behavior Identification

Virtualization technology enables server and service consolidation to save more power and operational costs of large scale computing systems. However, the consolidation nature of virtualization intensifies the power densities in a rack, thus resulting in higher probability of failures and Service Level Agreements violations under constrained power budget. In this paper, we proposed a power aware resource allocation algorithm, named PaRA, to coordinate and tradeoff the power and performance. In order to provide enough information about power characteristics of individual VM and applications, we use agent for workload characterization and estimation. The real actuation of resource allocation is realized within the Virtual Machines Monitor (VMM). We evaluated our algorithm on a real Xen based virtualization environments with 3 VMs. The results show that the proposed approach can achieve a higher SLA satisfaction rate as 99.6 percent in RUBiS web application, and power consumption savings up to 17.6 percent compared to default Xen resource allocator. The results also show the potential of our algorithm to be used in real virtualization environments.