Applications In Cloud Data Centers Research Articles

A priority, power and traffic-aware virtual machine placement of IoT applications in cloud data centers

Recent telecommunication paradigms, such as big data, Internet of Things (IoT), ubiquitous edge computing (UEC), and machine learning, are encountering with a tremendous number of complex applications that require different priorities and resource demands. These applications usually consist of a set of virtual machines (VMs) with some predefined traffic load between them. The efficiency of a cloud data center (CDC) as prominent component in UEC significantly depends on the efficiency of its VM placement algorithm applied. However, VM placement is an NP-hard problem and thus there exist practically no optimal solution for this problem. In this paper, motivated by this, we propose a priority, power and traffic-aware approach for efficiently solving the VM placement problem in a CDC. Our approach aims to jointly minimize power consumption, network consumption and resource wastage in a multi-dimensional and heterogeneous CDC. To evaluate the performance of the proposed method, we compared it to the state-of-the-art on a fat-tree topology under various experiments. Results demonstrate that the proposed method is capable of reducing the total network consumption up to 29%, the consumption of power up to 18%, and the wastage of resources up to 68%, compared to the second-best results.

An intelligent energy efficient storage system for cloud based big data applications

Storage technology has emerged as an indispensable paradigm for processing various applications in cloud data centers. The storage infrastructure consisting of Hard Disk Drives (HDDs) and Solid-State Drives (SSDs) accounts for high energy consumption. Also, the trade-offs between HDDs and SSDs in terms of cost and energy consumption are extremely high. Therefore, disk-based storage subsystems need to be more energy efficient. This paper proposes an intelligent energy-efficient hybrid disk storage system. The proposed system recognizes the frequently used data from traces of applications. Replica management along with data layout is used to allocate frequently used files in hot disks and the other files in cold disks. The request is executed using an intelligent scheduling technique that searches and selects the disk based on its states. The scheduling technique selects either an idle or active disk without spinning the standby disks. The selection procedure also considers minimum waiting time for the active disk and maximum remaining idle time for the idle disk. The proposed system has been implemented by adding disk management in the cloud environment, which proved to be highly effective in achieving a 39% power savings with an 18.26% decrease in execution time.

CECT: computationally efficient congestion-avoidance and traffic engineering in software-defined cloud data centers

The proliferation of cloud data center applications and network function virtualization (NFV) boosts dynamic and QoS dependent traffic into the data centers network. Currently, lots of network routing protocols are requirement agnostic, while other QoS-aware protocols are computationally complex and inefficient for small flows. In this paper, a computationally efficient congestion avoidance scheme, called CECT, for software-defined cloud data centers is proposed. The proposed algorithm, CECT, not only minimizes network congestion but also reallocates the resources based on the flow requirements. To this end, we use a routing architecture to reconfigure the network resources triggered by two events: 1) the elapsing of a predefined time interval, or, 2) the occurrence of congestion. Moreover, a forwarding table entries compression technique is used to reduce the computational complexity of CECT. In this way, we mathematically formulate an optimization problem and define a genetic algorithm to solve the proposed optimization problem. We test the proposed algorithm on real-world network traffic. Our results show that CECT is computationally fast and the solution is feasible in all cases. In order to evaluate our algorithm in term of throughput, CECT is compared with ECMP (where the shortest path algorithm is used as the cost function). Simulation results confirm that the throughput obtained by running CECT is improved up to 3x compared to ECMP while packet loss is decreased up to 2x.

Multicast Scheduling of Wavelength-Tunable, Multiqueue Optical Data Center Switches

The all-optical switching of multicast flows using star couplers and tunable transceivers is a promising solution for emerging cloud data center applications. However, the limited tuning range of optical components on one hand and the buffer management challenges for multicast traffic delivery on the other pose a significant impact on the performance of optical multicast scheduling algorithms. Using only one queue per input port results in head-of-line (HOL) blocking and limits the throughput, especially for bursty traffic patterns. As the number of possible multicast destinations grows exponentially with the switch size, allocating one queue per destination is not a feasible solution. To resolve HOL blocking, in this paper we consider only a handful of queues per switch input port and devise scalable scheduling algorithms that take into account transceiver tunability constraints. According to our Monte Carlo analysis of a switch with 64 ports and operating under bursty traffic, it is possible to improve the maximum achievable throughput by 44% when the number of queues per port is increased from one to eight. We show that the performance gains due to an increase in the queue count depend on the availability of the spectral resources. With the scarcity of wavelengths, an increase in the number of queues leads to diminishing returns.

Effect of Service Broker Policies and Load Balancing Algorithms on the Performance of Large Scale Internet Applications in Cloud Datacenters

Cloud computing is advancing rapidly. With such advancement, it has become possible to develop and host large scale distributed applications on the Internet more economically and more flexibly. However, the geographical distribution of user bases, the available Internet infrastructure within those geographical areas, and the dynamic nature of usage patterns of the user bases are critical factors that affect the performance of these applications. Therefore, it is necessary to compromise between datacenters, service broker policies, and load balancing algorithms to optimize the performance of the application and the cost to the owners. This paper aims at studying the effect of service broker policies and load balancing algorithms on the performance of large-scale Internet applications under different configurations of datacenters. To achieve this goal, we modeled the behavior of the popular Facebook application with the most recent worldwide users’ statistics. Then, we evaluated the performance of this application under different configurations of datacenters using: 1) two different service broker policies, namely, closest datacenter and optimum response time; and 2) three load-balancing algorithms, namely, round robin, equally spread current execution, and throttled load balancer. The overall average response time of the application and the overall average time spent for processing a user request by a datacenter are measured and the results are discussed. This study would help service providers generate valuable insights on coordination between datacenters, service policies, and load balancing algorithms when designing Cloud infrastructure services in geographically distributed areas. In addition, application designers would benefit greatly from this study in identifying the optimal arrangement for their applications.

Joint Optimization of Operational Cost and Performance Interference in Cloud Data Centers

Virtual machine (VM) scheduling is an important technique for the efficient operation of the computing resources in a data center. Previous work has mainly focused on consolidating VMs to improve resource utilization and to optimize energy consumption. However, the interference between collocated VMs is usually ignored, which can result in much worse performance degradation of the applications running on the VMs due to the contention of the shared resources. Based on this observation, we aim at designing efficient VM assignment and scheduling strategies in which we consider optimizing both the operational cost of the data center and the performance degradation of the running applications. We then propose a general model that captures the tradeoff between the two contradictory objectives. We present offline and online solutions for this problem by exploiting the spatial and temporal information of performance interference of VM collocation, where VM scheduling is performed by jointly considering the combinations and the life-cycle overlap of the VMs. Evaluation results show that the proposed methods can generate efficient schedules for VMs, achieving low operational cost while significantly reducing the performance degradation of applications in cloud data centers.

An algorithm for network and data-aware placement of multi-tier applications in cloud data centers

Today's Cloud applications are dominated by composite applications comprising multiple computing and data components with strong communication correlations among them. Although Cloud providers are deploying large number of computing and storage devices to address the ever increasing demand for computing and storage resources, network resource demands are emerging as one of the key areas of performance bottleneck. This paper addresses network-aware placement of virtual components (computing and data) of multi-tier applications in data centers and formally defines the placement as an optimization problem. The simultaneous placement of Virtual Machines and data blocks aims at reducing the network overhead of the data center network infrastructure. A greedy heuristic is proposed for the on-demand application components placement that localizes network traffic in the data center interconnect. Such optimization helps reducing communication overhead in upper layer network switches that will eventually reduce the overall traffic volume across the data center. This, in turn, will help reducing packet transmission delay, increasing network performance, and minimizing the energy consumption of network components. Experimental results demonstrate performance superiority of the proposed algorithm over other approaches where it outperforms the state-of-the-art network-aware application placement algorithm across all performance metrics by reducing the average network cost up to 67% and network usage at core switches up to 84%, as well as increasing the average number of application deployments up to 18%.

TTSA: An Effective Scheduling Approach for Delay Bounded Tasks in Hybrid Clouds.

The economy of scale provided by cloud attracts a growing number of organizations and industrial companies to deploy their applications in cloud data centers (CDCs) and to provide services to users around the world. The uncertainty of arriving tasks makes it a big challenge for private CDC to cost-effectively schedule delay bounded tasks without exceeding their delay bounds. Unlike previous studies, this paper takes into account the cost minimization problem for private CDC in hybrid clouds, where the energy price of private CDC and execution price of public clouds both show the temporal diversity. Then, this paper proposes a temporal task scheduling algorithm (TTSA) to effectively dispatch all arriving tasks to private CDC and public clouds. In each iteration of TTSA, the cost minimization problem is modeled as a mixed integer linear program and solved by a hybrid simulated-annealing particle-swarm-optimization. The experimental results demonstrate that compared with the existing methods, the optimal or suboptimal scheduling strategy produced by TTSA can efficiently increase the throughput and reduce the cost of private CDC while meeting the delay bounds of all the tasks.

Heuristic virtual machine allocation for multi-tier Ambient Assisted Living applications in a cloud data center

Cloud computing provides the essential infrastructure for multi-tier Ambient Assisted Living (AAL) applications that facilitate people's lives. Resource provisioning is a critically important problem for AAL applications in cloud data centers (CDCs). This paper focuses on modeling and analysis of multi-tier AAL applications, and aims to optimize resource provisioning while meeting requests' response time constraint. This paper models a multi-tier AAL application as a hybrid multi-tier queueing model consisting of an M/M/c queueing model and multiple M/M/1 queueing models. Then, virtual machine (VM) allocation is formulated as a constrained optimization problem in a CDC, and is further solved with the proposed heuristic VM allocation algorithm (HVMA). The results demonstrate that the proposed model and algorithm can effectively achieve dynamic resource provisioning while meeting the performance constraint.

Ensemble: A Tool for Performance Modeling of Applications in Cloud Data Centers

We introduce Ensemble, a runtime framework and associated tools for building application performance models on-the-fly. These dynamic performance models can be used to support complex, highly dimensional resource allocation, and/or what-if performance inquiry in modern heterogeneous environments, such as data centers and Clouds. Ensemble combines simple, partially specified, and lower-dimensionality models to provide good initial approximations for higher dimensionality application performance models. We evaluated Ensemble on industry-standard and scientific applications. The results show that Ensemble provides accurate, fast, and flexible performance models even in the presence of significant environment variability.

Deadline and Incast Aware TCP for cloud data center networks

Nowadays, cloud data centers have become a key resource to provide a plethora of rich online services such as social networking and cloud computing. The cloud data center applications typically follow the Partition/Aggregate traffic pattern based on a tree-like logical topology, where the aggregator node may gather response data from thousands of worker nodes. One of the key challenges for such applications, however, is to meet the soft real-time constraints. In this paper, we introduce the design and implementation of DIATCP, a new transport protocol that is both deadline-aware and incast-avoidable for cloud data center applications. Prior work achieve deadline awareness by host-based or network-based approaches, but they are either imperfect in meeting their deadlines or have weakness in practical deployment. In contrast, DIATCP is deployed only at the aggregator, which directly controls the peers’ sending rate to avoid incast congestion and, more importantly, to meet the application deadline. This is under the key observation that the aggregator knows the bottleneck link status as well as its workers’ information under the Partition/Aggregate traffic pattern. Through detailed ns-3 simulations and real testbed experiments, we show that DIATCP significantly outperforms the previous protocols in the cloud data center environment.

Differentiated Bandwidth Guarantees for Cloud Data Centers

By means of virtualization, computing and storage resources are effectively multiplexed by different applications in cloud data centers. However, there lacks useful approaches to share the internal network resource of cloud data centers. Invalid network sharing not only degrade the performance of applications, but also affect the efficiency of data center operation. To guarantee network performance of applications and provide fine-grained service differentiation, in this paper, we propose a differentiated bandwidth guarantee scheme for data center networks. Utility functions are constructed according to the throughput and delay sensitive characteristics of different applications. Aiming to maximize the utility of all applications, the problem is formulated as a multi-objective optimization problem. We solve this problem using a heuristic algorithm: the elitist Non-Dominated Sorted Genetic Algorithm-II(NSGA-II), and we make a multi-attribute decision to refine the solutions. Extensive simulations are conducted to show that our scheme provides minimum band-width guarantees and achieves more fine-grained service differentiation than existing approaches. The simulation also verifies that the proposed mechanism is suitable for arbitrary data center architectures.

The Application Research of Private Cloud in the Data Centers Colleges of Universities

With the development of information technology and business model innovation, cloud computing has become a new user experience and business model. This paper first introduced the concept of cloud computing features and cloud computing system architecture, and then as the private cloud data center applications in colleges and universities an example, analysised the value of private cloud for the university data center, private cloud key technology and implementation details, and finally Indicates that a private cloud computing the future development trends and the application mode of school-enterprise cooperation private cloud.