Scalable group management in large-scale virtualized clusters

To save cost, more and more users choose provision resources at the granularity of virtual machines in cluster systems, especially data centres. Maintaining a consistent member view is the foundation of reliable cluster managements, and it also raises several challenge issues for large scale cluster systems deployed with virtual machines （which we call virtualized clusters）. In this paper, we introduce our experience in design and implementation of scalable member view management on large-scale virtual clusters. Our research contributions include three-aspects ： 1 ） we propose a scalable and reliable management infrastructure that combines a peer-to-peer structure and a hierarchy structure to maintain a consistent member view in virtual clusters; 2 ） we present a light-weighted group membership algorithm that can reach the consistent member view within a single round of message exchange; 3 ） we design and implement a scalable membership service that can provide virtual machines and maintain a consistent member view in virtual clusters. Our work is verified on Dawning 5000A, which ranked No. 10 of Top 500 super computers in November, 2008.

Electricity-Carbon Interactive Optimal Dispatch of Multi-Virtual Power Plant Considering Integrated Demand Response

As new power systems and dual carbon policies develop,virtual power plant cluster(VPPC)provides another reliable way to promote the efficient utilization of energy and solve environmental pollution problems.To solve the coordinated optimal operation and low-carbon economic operation problem in multi-virtual power plant,a multi-virtual power plant(VPP)electricity-carbon interaction optimal scheduling model considering integrated demand response(IDR)is proposed.Firstly,a multi-VPP electricity-carbon interaction framework is established.The interaction of electric energy and carbon quotas can realize energy complementarity,reduce energy waste and promote low-carbon operation.Secondly,in order to coordinate the multiple types of energy and load in VPPC to further achieve low-carbon operation,the IDR mechanism based on the user comprehensive satisfaction(UCS)of electricity,heat as well as hydrogen is designed,which can effectively maintain the UCS in the cluster within a relatively high range.Finally,the unit output scheme is formulated to minimize the total cost of VPPC and the model is solved using theCPLEX solver.The simulation results showthat the proposed method effectively promotes the coordinated operation among multi-VPP,increases the consumption rate of renewable energy sources and the economics of VPPC and reduces carbon emissions.

Energy-efficient low-latency virtual cluster synchronization algorithm based on SMAC protocol in wireless sensor networks

The MAC protocol design for wireless sensor networks has been researched and developed for decades. SMAC protocol is a famous energy-efficient MAC protocol. Based on SMAC protocol, we find that the boundary nodes in the cluster-shaped synchronization structure bring energy consumption seriously, and provide a virtual cluster aggregation (VCA) algorithm. Because the bounder node follows multiple schedules in one cycle, it may deplete earlier and cause segmentation in wireless sensor networks. The algorithm reduces energy consumption of boundary nodes and extends the lifetime of entire sensor network by merging different virtual clusters, but increases the data transmission delay. Because the sensor nodes have the fixed duty cycle, the larger the coverage area of network is, the greater the data transmission delay increases. We propose the dynamic duty cycle (DDC) algorithm to solve this effect. When the network load and data transmission delay increase, the DDC algorithm exponentially changes the duty cycle of the node to reduce latency. The simulation results show that the performance of SMAC with the VCA and DDC algorithm obtains improvement significantly.

REMUDA: A Practical Topology Control and Data Forwarding Mechanism for Wireless Sensor Networks

In wireless sensor networks, topology control plays an important role for data forwarding efficiency in the data gathering applications. In this paper, we present a novel topology control and data forwarding mechanism called REMUDA, which is designed for a practical indoor parking lot management system. REMUDA forms a tree-based hierarchical network topology which brings as many nodes as possible to be leaf nodes and constructs a virtual cluster structure. Meanwhile, it takes the reliability, stability and path length into account in the tree construction process. Through an experiment in a network of 30 real sensor nodes, we evaluate the performance of REMUDA and compare it with LEPS which is also a practical routing protocol in TinyOS. Experiment results show that REMUDA can achieve better performance than LEPS.

A comparative study of cluster-based methods at finite strain

Virtual clustering analysis(VCA)is a reduced-order method for numerical homogenization.We formulate VCA for finite strain problems,illustrate its implementation,and provide numerical codes for free download and perusal.Comparison for four test examples shows that both VCA and Self-consistent clustering analysis(SCA)excellently approximate FFT-based direct numerical simulation(DNS)results,yet at much reduced numerical expenses.VCA is actually even faster in the online stage.

Cluster as a Service： A Resource Sharing Approach for Private Cloud

With the rapid development of cloud computing and big data processing, an increasing number of application frameworks are being considered to run in a ＂cloud way＂. This development brings about several challenges to the enterprise private cloud computing platform, e.g., being able to run most existing heterogeneous applications, providing scalability and elasticity support for newly emerged frameworks, and most importantly,sharing cluster resources effectively. In this paper, we propose a new service model, namely, Cluster as a Service（Claa S）, which is suitable for medium- and small-sized data centers to solve these problems in a relatively easy and general way. The idea behind this model is virtualizing the cluster environment for distributed application frameworks. Most applications can directly run in the virtual cluster environment without any modification, which is a great advantage. Based on lightweight containers, we implement a real system of Claa S named Docklet to prove the feasibility of this service model. Meanwhile, we preliminarily design the definition of applications to make them easy to deploy. Finally, we present several examples and evaluate the entire system.