Characterizing application scheduling on edge,fog, and cloud computing resources

Cloud computing has grown to become a popular distributed computing service offered by commercial providers. More recently, edge and fog computing resources have emerged on the wide-area network as part of Internet of things (IoT) deployments. These three resource abstraction layers are complementary, and offer distinctive benefits. Scheduling applications on clouds has been an active area of research, with workflow and data flow models offering a flexible abstraction to specify applications for execution. However, the application programming and scheduling models for edge and fog are still maturing, and can benefit from learnings on cloud resources. At the same time, there is also value in using these resources cohesively for application execution. In this article, we offer a taxonomy of concepts essential for specifying and solving the problem of scheduling applications on edge, fog, and cloud computing resources. We first characterize the resource capabilities and limitations of these infrastructure and offer a taxonomy of application models, quality-of-service constraints and goals, and scheduling techniques, based on a literature review. We also tabulate key research prototypes and papers using this taxonomy. This survey benefits developers and researchers on these distributed resources in designing and categorizing their applications, selecting the relevant computing abstraction(s), and developing or selecting the appropriate scheduling algorithm. It also highlights gaps in literature where open problems remain.     

网络集群计算系统中的并行任务调度

基于多处理机并行任务调度模型，探讨网络集群计算系统中的并行任务调度问题，首先证明了一般网络集群计算系统中调度算法的可近似性难度，然后提出了三种不同的启发式算法：最大长度优先调度算法、最大宽度优先调度算法和最大面积优先调度算法；然后根据大量的模拟实验对这些算法以及文献中已提出的调度算法进行了比较分析，结果表明该文的启发式算法比文献中的算法在性能上效果更好。     

Man-Computer Interaction—The Contribution of the Human Sciences

An introductory survey is presented of the possible contribution of the human sciences to man-computer interaction, based upon a full review of the relevant human factors literature which is to be published, A possible taxonomy for the field is proposed founded on broad divisions of the human sciences problem areas and of the types of computer systems and services. Using the taxonomy as a framework, some examples of relevant human sciences work and some problems and research needs are discussed.     

An extensible pattern-based library and taxonomy of security threats for distributed systems

Security is one of the most essential quality attributes of distributed systems, which often operate over untrusted networks such as the Internet. To incorporate security features during the development of a distributed system requires a sound analysis of potential attacks or threats in various contexts, a process that is often termed "threat modeling". To reduce the level of security expertise required, threat modeling can be supported by threat libraries (structured or unstructured lists of threats), which have been found particularly effective in industry scenarios; or attack taxonomies, which offer a classification scheme to help developers find relevant attacks more easily. In this paper we combine the values of threat libraries and taxonomies, and propose an extensible, two-level "pattern-based taxonomy" for (general) distributed systems. The taxonomy is based on the novel concept of a threat pattern, which can be customized and instantiated in different architectural contexts to define specific threats to a system. This allows developers to quickly consider a range of relevant threats in various architectural contexts as befits a threat library, increasing the efficacy of, and reducing the expertise required for, threat modeling. The taxonomy aims to classify a wide variety of more abstract, system- and technology-independent threats, which keeps the number of threats requiring consideration manageable, increases the taxonomy's applicability, and makes it both more practical and more useful for security novices and experts alike. After describing the taxonomy which applies to distributed systems generally, we propose a simple and effective method to construct pattern-based threat taxonomies for more specific system types and/or technology contexts by specializing one or more threat patterns. This allows for the creation of a single application-specific taxonomy. We demonstrate our approach to specialization by constructing a threat taxonomy for peer-to-peer systems.     

Distributed dynamic load balancing for pipelined computations on heterogeneous systems

One of the most significant causes for performance degradation of scientific and engineering applications on high performance computing systems is the uneven distribution of the computational work to the resources of the system. This effect, which is known as load imbalance, is even more noticeable in the case of irregular applications and heterogeneous distributed systems. This motivated the parallel and distributed computing research community to focus on methods that provide good load balancing for scientific and engineering applications running on (heterogeneous) distributed systems. Efficient load balancing and scheduling methods are employed for scientific applications from various fields, such as mechanics, materials, physics, chemistry, biology, applied mathematics, etc. Such applications typically employ a large number of computational methods in order to simulate complex phenomena, on very large scales of time and magnitude. These simulations consist of routines that perform repetitive computations (in the form of DO/FOR loops) over very large data sets, which, if not properly implemented and executed, may suffer from poor performance. The number of repetitive computations in the simulation codes is not always constant. Moreover, the computational nature of these simulations may be in fact irregular, leading to the case when one computation takes (unpredictably) more time than others. For successful and timely results, large scale simulations require the use of large scale computing systems, which often are widely distributed and highly heterogeneous. Moreover, large scale computing systems are usually shared among multiple users, which causes the quality and quantity of the available resources to be highly unpredictable. There are numerous load balancing methods in the literature for different parallel architectures. The most recent of these methods typically follow the master-worker paradigm, where a single coordinator (master) is responsible for making all the scheduling decisions based on information provided by the workers. Depending on the application requirements, the scheduling policy and the computational environment, the benefits of this paradigm may be limited as follows: (1) its efficiency may not scale as the number of processors increases, and (2) it is quite probable that the scheduling decisions are made based on outdated information, especially on systems where the workload changes rapidly. In an effort to address these limitations, we propose a distributed (master-less) load balancing scheme, in which the scheduling decisions are made by the workers in a distributed fashion. We implemented this method along with other two master-worker schemes (a previously existing one and a recently modified one) for three different scientific computational kernels. In order to validate the usefulness and efficiency of the proposed scheme, we conducted a series of comparative performance tests with the two master-worker schemes for each computational kernel. The target system is an SMP cluster, on which we simulated three different patterns of system load fluctuation. The experiments strongly support the belief that the distributed approach offers greater performance and better scalability on such systems, showing an overall improvement ranging from 13% to 24% over the master-worker approaches.     

Context‐aware scheduling in MapReduce: a compact review

It is a fact that the attention of research community in computer science, business executives, and decision makers is drastically drawn by big data. As the volume of data becomes bigger, it needs performance‐oriented data‐intensive processing frameworks such as MapReduce, which can scale computation on large commodity clusters. Hadoop MapReduce processes data in Hadoop Distributed File System as jobs scheduled according to YARN fair scheduler and capacity scheduler. However, with advancement and dynamic changes in hardware and operating environments, the performance of clusters is greatly affected. Various efforts in literature have been made to address the issues of heterogeneity (i.e., clusters consisting of virtual machines and machines with different hardware), network communication, data locality, better resource utilization, and run‐time scheduling. In this paper, we present a survey to discuss various research efforts made so far to improve Hadoop MapReduce scheduling. We classify scheduling algorithms and techniques proposed in the literature so far based on their addressing areas and present a taxonomy. Furthermore, we also discuss various aspects of open issues and challenges in the scheduling of MapReduce to improve its performance. Copyright © 2015 John Wiley & Sons, Ltd.     

A fuzzy rule-based meta-scheduler with evolutionary learning for grid computing

Grid computing is increasingly emerging as a promising platform for large-scale problems solving in science, engineering and technology. Nevertheless, a major effort is still required to harness the high potential performance of such computational framework and in this sense, an important challenge is to develop new strategies that efficiently address scheduling on the distributed, heterogeneous and shared environment of grids. Fuzzy rule-based systems (FRBSs) models are dynamic and are currently attracting the interest of scheduling research community to obtain near-optimal solutions on grids. However, FRBSs performance is strongly related to the quality of their knowledge bases and thus, with the knowledge acquisition process. Due to the inherent dynamic nature and the typical complex search spaces of grids, automatically finding a high-quality knowledge base that accurately describes the fuzzy system is extremely relevant. In this work, we propose a scheduling system for grids considering a novel learning strategy inspired by Michigan and Pittsburgh approaches that applies genetic algorithms (GAs) to evolve the fuzzy rule bases and improves the classical learning strategies in terms of computational effort and convergence behaviour. In addition, experimental results show that the proposed schema significantly outperforms other extensively used scheduling strategies.     

Communication infrastructure in distributed scheduling

The emergence of distributed artificial intelligent (DAI) introduced a new approach to solve scheduling problems by a set of scheduling systems that interact with each other in the problem-solving process. In this paper, we describe a communication infrastructure to handle connection and communication between distributed Internet scheduling systems for distributed applications. First, we present an agent model of distributed scheduling systems where agents can communicate and coordinate activities with each other via an agent communication language. Then, we define the syntax and semantics for the agent communication languages, and negotiation mechanism. Following that, we discuss the design and development of the prototype for the multi-agent scheduling systems. We conclude with a discussion of communication issues for heterogeneous agent-based scheduling systems to solve distributed scheduling problems.     

Enhancing rule-based scheduling in wafer fabrication facilities by evolutionary algorithms: Review and opportunity

Scheduling is a critical and challenging task in manufacturing systems, especially in large-scale complex systems like wafer fabrication facilities. Although evolutionary algorithms (EAs) have demonstrated many successful applications in the field of manufacturing scheduling, there are very few studies on scheduling of wafer fabs using EAs. Dispatching rules are one of the most common techniques for fab scheduling. In this paper, we present six ways of applying EAs for enhancing the rule-based scheduling system. We provide potential EA-based solutions and review relevant literature. Many of the mentioned viewpoints can serve as new research topics for both researchers in the fields of scheduling and evolutionary computation (EC). Several general EC techniques including multiobjective optimization, expensive optimization, and parallelization are also introduced and shown to be helpful to fab scheduling.     

Pre-run-time scheduling in real-time systems: Current researches and Artificial Intelligence perspectives

This paper presents the taxonomy of real-time systems with special emphasize on pre-run-time scheduling problem. Firstly, we present real-time systems, real-time tasks, timing, precedence and exclusion constraints. Then, we describe the problem of pre-run-time scheduling of tasks under constraints. After that, we present the most existing efficient techniques to deal with the latter problem. We summarize the discussion of existing techniques and possible research perspectives after surveying the Artificial Intelligence’s point of view about the problem of pre-run-time scheduling of real-time tasks. The Artificial Intelligence survey includes Constraint Satisfaction Problems class since pre-run-time scheduling belongs to the latter class. The Artificial Intelligence survey includes also Path-finding Problems from which intelligent algorithms could be observed such as Learning-Real-Time-A1(LRTA1) thanks to its important properties (optimality, linear space complexity and determinism). The development of an algorithm like LRTA1 to solve Constraints Satisfaction Problems and particularly the pre-run-time scheduling of real-time tasks problem is one clear research direction to deal with large-scale real-time systems. The overall objective of this paper is to show what are the perspectives to Artificial Intelligence literature that could be beneficial firstly to Artificial Intelligence community itself and secondly to real-time systems community.     

Performance evaluation of a new scheduling algorithm for distributed systems with security heterogeneity

High quality of security service is increasingly critical for applications running on heterogeneous distributed systems. However, existing scheduling algorithms for heterogeneous distributed systems disregard security requirements of applications. To address this issue, in this paper, we introduce security heterogeneity concept for our scheduling model in the context of distributed systems. Based on the concept, we propose a novel heuristic scheduling algorithm, or SATS, which strives to maximize the probability that all tasks are executed without any risk of being attacked. Extensive experimental studies using real-world traces indicate that the scheduling performance is affected by heterogeneities of security and computational power. Additionally, empirical results demonstrate that with respect to security and performance, the proposed scheduling algorithm outperforms existing approaches under a wide spectrum of workload conditions.     

Multirobot systems: a classification focused on coordination

Multirobot systems (MRS) are, nowadays, an important research area within robotics and artificial intelligence and a growing number of systems have recently been presented in the literature. Since application domains and tasks that are faced by MRS are of increasing complexity, the ability of the robots to cooperate can be regarded as a fundamental feature. In this paper, we present a survey of the recent work in the area by specifically examining the forms of cooperation and coordination realized in the MRS. In particular, we propose a new taxonomy for classification of the approaches to coordination in MRS and we describe some systems, which we consider representative in our taxonomy. We finally discuss the outcomes of our analysis and try to highlight future trends of the research on MRS.     

分布式系统下的DAG任务调度研究综述

近年来随着网格、云计算工作流等分布式计算技术的发展,关于DAG(有向无环图)模型任务在分布式系统环境下的调度问题逐渐成为备受关注的研究热点。根据最新研究进展,对分布式系统下的DAG任务调度问题和有关技术进行了研究与讨论,主要包括四个方面:系统地描述了分布式系统和异构分布式系统的有关概念,异构分布式系统下的DAG任务调度问题、调度模型及其典型应用;对现有分布式系统下DAG任务调度的研究按照不同的方式进行了分类;探讨了多DAG共享异构分布式资源调度的研究现状;讨论了目前多DAG共享异构分布式资源调度研究存在的问题和未来可能的研究方向。     

Shop-floor scheduling at shipbuilding yards using the multiple intelligent agent system

The intelligent agent system has become a typical approach to research distributed artificial intelligence and distributed problem solving. However, despite its various technical advantages, the distribution and expansion of the intelligent agent system approach has been limited because existing methodologies rely on specialized applications and therefore require intensive investment to develop new systems. In an attempt to conquer these difficulties, a standardized methodology to construct intelligent agent systems is proposed. This approach deviates from current agent system approaches of repeatedly constructing new and customized expert systems. Specifically, it provides an economical method for developing intelligent agent systems by investigating the possibility of standardizing message communication protocols in linguistics speech-act theory and by supplementing traditional algorithmic systems with intelligent segments using, among others, expert system tools. To verify effectiveness, the shop-floor scheduling system of a large-scale shipbuilding yard has been redesigned, developed, and tested using this approach. This shop-floor scheduling system requires the scheduling of when and where to process block construction under various constraints. It is a difficult four-dimensional time and space allocation problem involving traditional NP-complete search spaces. The tested intelligent agent system proposes an innovative method for reducing the search space into three levels: the algorithm level, the agent intelligence level, and the level of cooperation among agents. It also provides methods for solving deadlock occurrence and non-uniformity problems resulting from parallel processing. Test results demonstrate applicability and economy, among other technical advantages.     

Improving the remote scheduling of distributed production with process statistics and AI techniques

Stochastic events, such as rush orders, stock-out events, and local failures have an important impact on the performance of distributed production, but they are difficult to anticipate and account for when scheduling production activities. Process statistics and artificial intelligence techniques can provide this knowledge to effectively time synchronization events among the simulation and scheduling federates of a same distributed architecture. Measurable benefits include reduced communication delays and, thus, improved responsiveness of the system to changes in production and new scheduling needs, as they arise. Comparative results on the productivity of actual industrial systems are proposed and discussed in the paper.     

From Predefined Consistency to User-Centered Emergent Consistency in Real-Time Collaborative Editing Systems

In this paper, we introduce a basic conceptual model of real-time collaborative editing systems which differentiates the coordination aspect of actions (which do not exist in single-user systems) from the application-dependent semantic aspect of actions (i.e., those actions that happen in single-user systems). Based on this model, a new taxonomy is provided to analyze the existing consistency-maintenance mechanisms and highlight potential new mechanisms. Furthermore, a new user-centered consistency model is proposed, which focuses on dynamic consistency that is negotiable among end users rather than static consistency that is automatically determined by the system. In the mean time, some challenging research issues in real-time collaborative editing systems are outlined     

A taxonomy of network computing systems

Rapid advances in networking and microprocessor technologies have led to the emergence of Internet-wide distributed computing systems ranging from simple LAN-based clusters to planetary-scale networks. As these network computing systems evolve by combining the best features of existing systems, differences among NCs are blurring. To address this problem, researchers have proposed formal taxonomies of NC systems. We propose a new taxonomy that is both broad enough to encompass all NC systems and simple enough to be widely used.     

仿真网格中一种基于知识的动态任务调度算法

仿真网格是以通用网格技术为基础、面向仿真领域的专用网格,目前国际上对仿真网格的研究尚处于起步阶段.现有的分布式仿真HLA(high level architecture)体系结构中的仿真资源和联邦成员是静态绑定的,网格技术的引入使得仿真资源的动态分配成为可能.根据仿真网格任务调度的特点,在仿真网格中建立了一种任务调度模型,并针对该模型,提出了一种新的基于知识的动态任务调度算法KMO,该算法适用于将N个相互独立的计算需求不同的仿真任务调度到M个随时间动态变化的仿真资源上,它能对若干次调度后的结果进行统计并提炼成"知识"反馈给算法预处理部分,使得该算法在动态多变的环境中能获得比较稳定的性能.实验结果表明,在仿真网格环境中,该算法的性能优于网格中传统的任务调度算法.     

A simulation-based approach for dynamic process management at web service platforms

Web services technology is being adopted as a viable deployment approach for future distributed software systems that enable business-to-business and business-to-consumer interactions across the open and dynamic internet environment. Recent research is focused on developing support technologies for web service discovery, on-demand service composition, and robust execution to facilitate web services based deployment of business processes. Developing techniques to cope with the volatile and open nature of the web during execution of composite services at the service platform is essential for delivering reliable and acceptable performance in this new process delivery framework. In this paper, we propose a simulation-based framework to guide scheduling of composite service execution. Online simulation of the dynamics of the open environment is used for scheduling service requests at the service platform. Comparison of the look-ahead simulation for different scheduling policies with the current execution state provides guidelines for service execution in order to cope with system volatility. We have implemented a prototype of the proposed framework and illustrate the feasibility of our approach with experimental studies.     

High-throughput query scheduling with spatial clustering based on distributed exponential moving average

In distributed scientific query processing systems, leveraging distributed cached data is becoming more important. In such systems, a front-end query scheduler distributes queries among many application servers rather than processing queries in a few high-performance workstations. Although many query scheduling policies exist such as round-robin and load-monitoring, they are not sophisticated enough to exploit cached results as well as balance the workload. Efforts were made to improve the query processing performance using statistical methods such as exponential moving average. However, existing methods have limitations for certain query patterns: queries with hotspots, or dynamic query distributions. In this paper, we propose novel query scheduling policies that take into account both the contents of distributed caching infrastructure and the load balance among the servers. Our experiments show that the proposed query scheduling policies outperform existing policies by producing better query plans in terms of load balance and cache-hit ratio.