Showing papers by "Rajkumar Buyya published in 2020"
TL;DR: A novel framework called HealthFog is proposed for integrating ensemble deep learning in Edge computing devices and deployed it for a real-life application of automatic Heart Disease analysis.
387 citations
TL;DR: A mobility-driven cloud-fog-edge collaborative real-time framework, Mobi-IoST, has been proposed, which has IoT, Edge, Fog and Cloud layers and exploits the mobility dynamics of the moving agent.
Abstract: The design of mobility-aware framework for edge/fog computing for IoT systems with back-end cloud is gaining research interest. In this paper, a mobility-driven cloud-fog-edge collaborative real-time framework, Mobi-IoST, has been proposed, which has IoT, Edge, Fog and Cloud layers and exploits the mobility dynamics of the moving agent. The IoT and edge devices are considered to be the moving agents in a 2-D space, typically over the road-network. The framework analyses the spatio-temporal mobility data (GPS logs) along with the other contextual information and employs machine learning algorithm to predict the location of the moving agents (IoT and Edge devices) in real-time. The accumulated spatio-temporal traces from the moving agents are modelled using probabilistic graphical model. The major features of the proposed framework are: (i) hierarchical processing of the information using IoT-Edge-Fog-Cloud architecture to provide better QoS in real-time applications, (ii) uses mobility information for predicting next location of the agents to deliver processed information, and (iii) efficiently handles delay and power consumption. The performance evaluations yield that the proposed Mobi-IoST framework has approximately 93% accuracy and reduced the delay and power by approximately 23–26% and 37–41% respectively than the existing mobility-aware task delegation system.
94 citations
TL;DR: Security issues and different attack vectors are discussed along with possible solutions for securing the SDN-enabled network architecture at different planes and their associated interconnections and the architecture of permissioned blockchain for SDN is proposed.
Abstract: Smart cities have emerged as a hub of intelligent applications (e.g., intelligent transportation systems, smart parking, smart homes, and e-healthcare) to provide ambient-assisted living and quality of experience to wide communities of users. The smooth execution of these applications depends on reliable data transmission between various smart devices and machines. However, the exponential increase in data traffic due to the growing dependency of end users on smart city applications has created various bottlenecks (e.g., channel congestion, manual flow configurations, limited scalability, and low flexibility) on the conventional network backbone, which can degrade the performance of any designed solution in this environment. To mitigate these challenges, SDN emerges as a powerful new technology that provides global visibility of the network by decoupling the control logic from the forwarding devices. The abstraction of network services in SDN architecture provides more flexibility for network administrators to execute various applications. In SDN architecture, the decision making process is handled by a logically centralized controller, which may have a single point of failure. An adversary/ attacker can compromise the controller using different types of attacks (e.g., eavesdropping, man-in-the middle attack, and distributed denial of service) in order to gain total control of the network by updating the flow table entries at the data plane or hindering control plane operations. Therefore, to cope with the aforementioned challenges, new strategies and solutions are required for securing the SDN-enabled network architecture at different planes and their associated interconnections. In this article, various security issues and different attack vectors are discussed along with possible solutions. To mitigate various attacks, BlockSDN, a blockchain as a service framework, for SDN is proposed. The architecture of permissioned blockchain is presented followed by two attack scenarios, 1) a malware compromised switch at the data plane and 2) distributed denial of service attack at the control plane, to demonstrate the applicability of the BlockSDN framework for various future applications. Finally, the open issues and challenges with respect to the design of blockchain solutions for SDN in smart city applications are also discussed.
94 citations
TL;DR: A SLA-aware autonomic resource management technique called STAR which mainly focuses on reducing SLA violation rate for the efficient delivery of cloud services and optimizing other QoS parameters which effect efficient cloud service delivery is presented.
Abstract: Cloud computing has recently emerged as an important service to manage applications efficiently over the Internet. Various cloud providers offer pay per use cloud services that requires Quality of Service (QoS) management to efficiently monitor and measure the delivered services through Internet of Things (IoT) and thus needs to follow Service Level Agreements (SLAs). However, providing dedicated cloud services that ensure user's dynamic QoS requirements by avoiding SLA violations is a big challenge in cloud computing. As dynamism, heterogeneity and complexity of cloud environment is increasing rapidly, it makes cloud systems insecure and unmanageable. To overcome these problems, cloud systems require self-management of services. Therefore, there is a need to develop a resource management technique that automatically manages QoS requirements of cloud users thus helping the cloud providers in achieving the SLAs and avoiding SLA violations. In this paper, we present SLA-aware autonomic resource management technique called STAR which mainly focuses on reducing SLA violation rate for the efficient delivery of cloud services. The performance of the proposed technique has been evaluated through cloud environment. The experimental results demonstrate that STAR is efficient in reducing SLA violation rate and in optimizing other QoS parameters which effect efficient cloud service delivery.
87 citations
TL;DR: In this article, the authors proposed an asynchronous advantage actor-critic (A3C) based real-time scheduler for stochastic edge-cloud environments allowing decentralized learning, concurrently across multiple agents.
Abstract: The ubiquitous adoption of Internet-of-Things (IoT) based applications has resulted in the emergence of the Fog computing paradigm, which allows seamlessly harnessing both mobile-edge and cloud resources. Efficient scheduling of application tasks in such environments is challenging due to constrained resource capabilities, mobility factors in IoT, resource heterogeneity, network hierarchy, and stochastic behaviors. Existing heuristic-based and Reinforcement Learning approaches lack generalizability and quick adaptability, thus failing to tackle this problem optimally. They are also unable to utilize the temporal workload patterns and are suitable only for centralized setups. Thus, we propose an Asynchronous-Advantage-Actor-Critic (A3C) based real-time scheduler for stochastic Edge-Cloud environments allowing decentralized learning, concurrently across multiple agents. We use the Residual Recurrent Neural Network (R2N2) architecture to capture a large number of host and task parameters together with temporal patterns to provide efficient scheduling decisions. The proposed model is adaptive and able to tune different hyper-parameters based on the application requirements. We explicate our choice of hyper-parameters through sensitivity analysis. The experiments conducted on real-world data set show a significant improvement in terms of energy consumption, response time, Service-Level-Agreement and running cost by 14.4%, 7.74%, 31.9%, and 4.64%, respectively when compared to the state-of-the-art algorithms.
87 citations
Posted Content•
TL;DR: A comprehensive review of QC literature is presented and a proposed taxonomy of QC is proposed to map various related studies to identify the research gaps and identify various open challenges and promising future directions for research and innovation in QC.
Abstract: Quantum computing is an emerging paradigm with the potential to offer significant computational advantage over conventional classical computing by exploiting quantum-mechanical principles such as entanglement and superposition. It is anticipated that this computational advantage of quantum computing will help to solve many complex and computationally intractable problems in several areas such as drug design, data science, clean energy, finance, industrial chemical development, secure communications, and quantum chemistry. In recent years, tremendous progress in both quantum hardware development and quantum software/algorithm have brought quantum computing much closer to reality. Indeed, the demonstration of quantum supremacy marks a significant milestone in the Noisy Intermediate Scale Quantum (NISQ) era - the next logical step being the quantum advantage whereby quantum computers solve a real-world problem much more efficiently than classical computing. As the quantum devices are expected to steadily scale up in the next few years, quantum decoherence and qubit interconnectivity are two of the major challenges to achieve quantum advantage in the NISQ era. Quantum computing is a highly topical and fast-moving field of research with significant ongoing progress in all facets. This article presents a comprehensive review of quantum computing literature, and taxonomy of quantum computing. Further, the proposed taxonomy is used to map various related studies to identify the research gaps. A detailed overview of quantum software tools and technologies, post-quantum cryptography and quantum computer hardware development to document the current state-of-the-art in the respective areas. We finish the article by highlighting various open challenges and promising future directions for research.
85 citations
TL;DR: A profit-aware application placement policy is proposed using constraint Integer Linear Programming model that simultaneously enhances profit and ensures QoS during application placement on computing instances and provides compensation to users for any violation of Service Level Agreement.
81 citations
TL;DR: Experimental results demonstrate that the proposed HYBRID algorithm outperforms peer research and benchmark algorithms in terms of efficient utilization of the cloud resources, improved reliability and reduced average response time.
Abstract: In this paper, we propose a novel HYBRID Bio-Inspired algorithm for task scheduling and resource management, since it plays an important role in the cloud computing environment. Conventional scheduling algorithms such as Round Robin, First Come First Serve, Ant Colony Optimization etc. have been widely used in many cloud computing systems. Cloud receives clients tasks in a rapid rate and allocation of resources to these tasks should be handled in an intelligent manner. In this proposed work, we allocate the tasks to the virtual machines in an efficient manner using Modified Particle Swarm Optimization algorithm and then allocation / management of resources (CPU and Memory), as demanded by the tasks, is handled by proposed HYBRID Bio-Inspired algorithm (Modified PSO + Modified CSO). Experimental results demonstrate that our proposed HYBRID algorithm outperforms peer research and benchmark algorithms (ACO, MPSO, CSO, RR and Exact algorithm based on branch-and-bound technique) in terms of efficient utilization of the cloud resources, improved reliability and reduced average response time.
77 citations
01 Jan 2020
TL;DR: In this article, the authors provide a comprehensive review of SDN security domain while focusing on its data plane, which is one of the least explored but most critical aspects in securing this technology.
Abstract: Software-defined network (SDN) radically changes the network architecture by decoupling the network logic from the underlying forwarding devices. This architectural change rejuvenates the network-layer granting centralized management and reprogrammability of the networks. From a security perspective, SDN separates security concerns into control and data plane, and this architectural recomposition brings up exciting opportunities and challenges. The overall perception is that SDN capabilities will ultimately result in improved security. However, in its raw form, SDN could potentially make networks more vulnerable to attacks and harder to protect. In this paper, we provide a comprehensive review of SDN security domain while focusing on its data plane, which is one of the least explored but most critical aspects in securing this technology. We review the most recent enhancements in SDNs, identify the main vulnerabilities of SDNs, and provide a novel attack taxonomy for SDNs. Thereafter, we provide a comprehensive analysis of challenges involved in protecting SDN data plane and control plane and provide an in-depth look into available solutions with respect to the identified threats and identify their limitations. To highlight the importance of securing the SDN platform, we also review the numerous security services built on top of this technology. We conclude the paper by offering future research directions.
68 citations
TL;DR: This article proposes a context-aware application placement policy for Fog environments that coordinates the IoT device-level contexts with the capacity of Fog nodes and minimizes the service delivery time of various I4OAs such as image processing and robot navigation applications.
Abstract: The fourth industrial revolution, widely known as Industry 4.0, is realizable through widespread deployment of Internet of Things (IoT) devices across the industrial ambiance. Due to communication latency and geographical distribution, Cloud-centric IoT models often fail to satisfy the Quality of Service requirements of different IoT applications assisting Industry 4.0 in real time. Therefore, Fog computing focuses on harnessing edge resources to place and execute these applications in the proximity of data sources. Since most of the Fog nodes are heterogeneous, distributed, and resource-constrained, it is challenging to place Industry 4.0-oriented applications (I4OAs) over them ensuring time-optimized service delivery. Diversified data sensing frequency of different industrial IoT devices and their data size further intensify the application placement problem. To address this issue, in this article we propose a context-aware application placement policy for Fog environments. Our policy coordinates the IoT device-level contexts with the capacity of Fog nodes and minimizes the service delivery time of various I4OAs such as image processing and robot navigation applications. It also ensures that the streams of input data flowing toward the placed applications neither congest the network nor increase the computing overhead of host Fog nodes significantly. Performance of the proposed policy is evaluated in both real-world and simulated Fog environments and compared with the existing placement policies. The experiment results show that our policy offers overall 16% improvement in service latency, network relaxation, and computing overhead management compared to other placement policies.
67 citations
TL;DR: Simulation results show that the proposed approaches significantly reduce the number of VM migrations and energy consumption while maintaining the QoS guarantee.
TL;DR: In this article, the authors investigate the existing application management strategies in fog computing and review them in terms of architecture, placement and maintenance, and highlight the research gaps in fog-based application management.
Abstract: The Internet of Things (IoT) paradigm is being rapidly adopted for the creation of smart environments in various domains. The IoT-enabled cyber-physical systems associated with smart city, healthcare, Industry 4.0 and Agtech handle a huge volume of data and require data processing services from different types of applications in real time. The Cloud-centric execution of IoT applications barely meets such requirements as the Cloud datacentres reside at a multi-hop distance from the IoT devices. Fog computing, an extension of Cloud at the edge network, can execute these applications closer to data sources. Thus, Fog computing can improve application service delivery time and resist network congestion. However, the Fog nodes are highly distributed and heterogeneous, and most of them are constrained in resources and spatial sharing. Therefore, efficient management of applications is necessary to fully exploit the capabilities of Fog nodes. In this work, we investigate the existing application management strategies in Fog computing and review them in terms of architecture, placement and maintenance. Additionally, we propose a comprehensive taxonomy and highlight the research gaps in Fog-based application management. We also discuss a perspective model and provide future research directions for further improvement of application management in Fog computing.
TL;DR: This work proposes a heterogeneous task allocation strategy for cost-efficient container orchestration through resource utilization optimization and elastic instance pricing with three main features to support heterogeneous job configurations to optimize the initial placement of containers into existing resources by task packing.
Abstract: Containers, as a lightweight application virtualization technology, have recently gained immense popularity in mainstream cluster management systems like Google Borg and Kubernetes. Prevalently adopted by these systems for task deployments of diverse workloads such as big data, web services, and IoT, they support agile application deployment, environmental consistency, OS distribution portability, application-centric management, and resource isolation. Although most of these systems are mature with advanced features, their optimization strategies are still tailored to the assumption of a static cluster. Elastic compute resources would enable heterogeneous resource management strategies in response to the dynamic business volume for various types of workloads. Hence, we propose a heterogeneous task allocation strategy for cost-efficient container orchestration through resource utilization optimization and elastic instance pricing with three main features. The first one is to support heterogeneous job configurations to optimize the initial placement of containers into existing resources by task packing. The second one is cluster size adjustment to meet the changing workload through autoscaling algorithms. The third one is a rescheduling mechanism to shut down underutilized VM instances for cost saving and reallocate the relevant jobs without losing task progress. We evaluate our approach in terms of cost and performance on the Australian National Cloud Infrastructure (Nectar). Our experiments demonstrate that the proposed strategy could reduce the overall cost by 23% to 32% for different types of cloud workload patterns when compared to the default Kubernetes framework.
TL;DR: This work investigates the existing application management strategies in Fog computing and review them in terms of architecture, placement and maintenance, and proposes a comprehensive taxonomy and highlights the research gaps in Fog-based application management.
Abstract: The Internet of Things (IoT) paradigm is being rapidly adopted for the creation of smart environments in various domains. The IoT-enabled Cyber-Physical Systems (CPSs) associated with smart city, healthcare, Industry 4.0 and Agtech handle a huge volume of data and require data processing services from different types of applications in real-time. The Cloud-centric execution of IoT applications barely meets such requirements as the Cloud datacentres reside at a multi-hop distance from the IoT devices. \textit{Fog computing}, an extension of Cloud at the edge network, can execute these applications closer to data sources. Thus, Fog computing can improve application service delivery time and resist network congestion. However, the Fog nodes are highly distributed, heterogeneous and most of them are constrained in resources and spatial sharing. Therefore, efficient management of applications is necessary to fully exploit the capabilities of Fog nodes. In this work, we investigate the existing application management strategies in Fog computing and review them in terms of architecture, placement and maintenance. Additionally, we propose a comprehensive taxonomy and highlight the research gaps in Fog-based application management. We also discuss a perspective model and provide future research directions for further improvement of application management in Fog computing.
TL;DR: A new cloud resource management procedure based on a multi-criteria decision-making method that takes advantage of a joint virtual machine and container migration approach concurrently is proposed which shows notable reductions in energy consumption, SLA violation, and number of migrations in comparison with the state-of-the-art algorithms.
TL;DR: Wang et al. as discussed by the authors developed an e-government privacy-preserving mobile, and fog computing framework entitled PPMF that can trace infected, and suspected cases nationwide, using personal mobile devices with contact tracing app, and two types of stationary fog nodes, named Automatic Risk Checkers (ARC), and Suspected User Data Uploader Node (SUDUN), to trace community transmission alongside maintaining user data privacy.
Abstract: To slow down the spread of COVID-19, governments worldwide are trying to identify infected people, and contain the virus by enforcing isolation, and quarantine. However, it is difficult to trace people who came into contact with an infected person, which causes widespread community transmission, and mass infection. To address this problem, we develop an e-government Privacy-Preserving Mobile, and Fog computing framework entitled PPMF that can trace infected, and suspected cases nationwide. We use personal mobile devices with contact tracing app, and two types of stationary fog nodes, named Automatic Risk Checkers (ARC), and Suspected User Data Uploader Node (SUDUN), to trace community transmission alongside maintaining user data privacy. Each user's mobile device receives a Unique Encrypted Reference Code (UERC) when registering on the central application. The mobile device, and the central application both generate Rotational Unique Encrypted Reference Code (RUERC), which broadcasted using the Bluetooth Low Energy (BLE) technology. The ARCs are placed at the entry points of buildings, which can immediately detect if there are positive or suspected cases nearby. If any confirmed case is found, the ARCs broadcast pre-cautionary messages to nearby people without revealing the identity of the infected person. The SUDUNs are placed at the health centers that report test results to the central cloud application. The reported data is later used to map between infected, and suspected cases. Therefore, using our proposed PPMF framework, governments can let organizations continue their economic activities without complete lockdown.
TL;DR: A conceptual model for reliable cloud computing has been proposed, along with a discussion on future research directions, and a case study of astronomy workflow is presented for reliable execution in the cloud environment.
Abstract: The next generation of cloud computing must be reliable to fulfil the end-user requirements, which are changing dynamically. Presently, cloud providers are facing challenges to ensure the reliability of their services. In this paper, we propose a comprehensive taxonomy of failure management in cloud computing. The taxonomy is used to investigate the existing techniques for reliability that need careful attention and investigation, as proposed by several academic and industry groups. Further, the existing techniques have been compared based on the common characteristics and properties of failure management as implemented in commercial and open-source solutions. A conceptual model for reliable cloud computing has been proposed, along with a discussion on future research directions. Moreover, a case study of astronomy workflow is presented for reliable execution in the cloud environment.
TL;DR: This article designs a data controller based on fuzzy logic to calculate the Rating of Allocation (RoA) value of each data request considering multiple context parameters, i.e., data, network, and quality and decides its on-chain allocation in IoT-blockchain systems.
Abstract: The use of Internet of Things (IoT) has introduced genuine concerns regarding data security and its privacy when data are in collection, exchange, and use. Meanwhile, blockchain offers a distributed and encrypted ledger designed to allow the creation of immutable and tamper-proof records of data at different locations. While blockchain may enhance IoT with innate security, data integrity, and autonomous governance, IoT data management and its allocation in blockchain still remain an architectural concern. In this article, we propose a novel context-aware mechanism for on-chain data allocation in IoT-blockchain systems. Specifically, we design a data controller based on fuzzy logic to calculate the Rating of Allocation (RoA) value of each data request considering multiple context parameters, i.e., data, network, and quality and decide its on-chain allocation. Furthermore, we illustrate how the design and realization of the mechanism lead to refinements of two commonly used IoT-blockchain architectural styles (i.e., blockchain-based cloud and fog). To demonstrate the effectiveness of our approach, we instantiate the data allocation mechanism in the blockchain-based cloud and fog architectures and evaluate their performance using FogBus. We also compare the efficacy of our approach to the existing decision-making mechanisms through the deployment of a real-world healthcare application. The experimental results suggest that the realization of the data allocation mechanism improves network usage, latency, and blockchain storage and reduces energy consumption.
TL;DR: To address the carbon emission problem of data centers, this paper considers shifting the workloads among multi-cloud located in different time zones and formulate the energy usage and carbon emission of data center and model the solar power corresponding to the locations.
01 Mar 2020
TL;DR: A QoS-aware Cloud Based Autonomic Information System for delivering agriculture related information as a service through the use of latest Cloud technologies which manage various types of agriculture related data based on different domains is presented.
Abstract: The Internet of Things (IoT) and cloud computing paradigms offer enhanced services for agricultural applications to manage the data efficiently. To provide an effective and reliable agriculture as a service, there is a need to manage Quality of Service (QoS) parameters to efficiently monitor and measure the delivered services dynamically. This paper presents a QoS-aware cloud based autonomic information system called Agri-Info for delivering agriculture related information as a service through the use of latest Internet-based technologies such as cloud computing and IoT which manage various types of agriculture related data based on different domains of agricultural industry. Proposed system gathers information from various users through preconfigured IoT devices (mobiles, laptops or iPads). It further manages and delivers the required information to users and diagnoses the agriculture status automatically. We have developed the web and mobile-based application and evaluated the performance of the proposed system in cloud environment using CloudSim toolkit based small scale environment. Results demonstrate our system yields in a reduction on 12.46% cost, on 15.52% network bandwidth, on 10.18% execution time and 13.32% in latency. Furthermore, a case study of an Indian village is presented to identify the customer satisfaction of farmers.
TL;DR: This paper entails a new approach for the prediction of water level in association with flood severity using the ensemble model that leverages the latest developments in the Internet of Things and machine learning for the automated analysis of flood data that might be useful to prevent natural disasters.
Abstract: River flooding is a natural phenomenon that can have a devastating effect on human life and economic losses. There have been various approaches in studying river flooding; however, insufficient understanding and limited knowledge about flooding conditions hinder the development of prevention and control measures for this natural phenomenon. This paper entails a new approach for the prediction of water level in association with flood severity using the ensemble model. Our approach leverages the latest developments in the Internet of Things (IoT) and machine learning for the automated analysis of flood data that might be useful to prevent natural disasters. Research outcomes indicate that ensemble learning provides a more reliable tool to predict flood severity levels. The experimental results indicate that the ensemble learning using the Long-Short Term memory model and random forest outperformed individual models with a sensitivity, specificity and accuracy of 71.4%, 85.9%, 81.13%, respectively.
TL;DR: An ensemble learning based workload forecasting method is presented that uses extreme learning machines and their corresponding forecasts are weighted by a voting engine and a metaheuristic algorithm inspired by blackhole theory is used to select the optimal weights.
TL;DR: Workflows are an application model that enables the automated execution of multiple interdependent and interconnected tasks and are widely used by the scientific community to manage the distributed execution and dataflow of complex simulations and experiments as mentioned in this paper.
Abstract: Workflows are an application model that enables the automated execution of multiple interdependent and interconnected tasks. They are widely used by the scientific community to manage the distributed execution and dataflow of complex simulations and experiments. As the popularity of scientific workflows continue to rise, and their computational requirements continue to increase, the emergence and adoption of multi-tenant computing platforms that offer the execution of these workflows as a service becomes widespread. This article discusses the scheduling and resource provisioning problems particular to this type of platform. It presents a detailed taxonomy and a comprehensive survey of the current literature and identifies future directions to foster research in the field of multiple workflow scheduling in multi-tenant distributed computing systems.
TL;DR: To better streamline time/delay-sensitive varied IoE requests, the authors contributes by introducing a smart layer between IoE devices and fog nodes to incorporate an intelligent and adaptive learning based task scheduling technique.
TL;DR: This paper proposes a game theoritic resource management technique that minimises infrastructure energy consumption and costs while ensuring applications performance and suggests that this approach could reduce up to 11.95% energy consumption, and approximately 17.86% user costs with negligible loss in performance.
Abstract: Internet of Things (IoT) is producing an extraordinary volume of data daily, and it is possible that the data may become useless while on its way to the cloud, due to long distances Fog/edge computing is a new model for analysing and acting on time-sensitive data, adjacent to where it is produced Further, cloud services provided by large companies such as Google, can also be localised to improve response time and service agility This is accomplished through deploying small-scale datacentres in various locations, where needed in proximity of users; and connected to a centralised cloud that establish a multi-access edge computing (MEC) The MEC setup involves three parties, ie service-providers (IaaS), application-providers (SaaS), network-providers (NaaS); which might have different goals, therefore, making resource management difficult Unlike existing literature, we consider resource management with-respect-to all parties; and suggest game-theoretic resource management techniques to minimise infrastructure energy consumption and costs while ensuring applications' performance Our empirical evaluation, using Google's workload traces, suggests that our approach could reduce up to 1195% energy consumption, and ~1786% user costs with negligible loss in performance Moreover, IaaS can reduce up-to 2027% energy bills and NaaS can increase their costs-savings up-to 1852% as compared to other methods
TL;DR: This work proposes an A3C based real-time scheduler for stochastic Edge-Cloud environments allowing decentralized learning, concurrently across multiple agents, and uses the R2N2 architecture to capture a large number of host and task parameters together with temporal patterns to provide efficient scheduling decisions.
Abstract: The ubiquitous adoption of Internet-of-Things (IoT) based applications has resulted in the emergence of the Fog computing paradigm, which allows seamlessly harnessing both mobile-edge and cloud resources. Efficient scheduling of application tasks in such environments is challenging due to constrained resource capabilities, mobility factors in IoT, resource heterogeneity, network hierarchy, and stochastic behaviors. xisting heuristics and Reinforcement Learning based approaches lack generalizability and quick adaptability, thus failing to tackle this problem optimally. They are also unable to utilize the temporal workload patterns and are suitable only for centralized setups. However, Asynchronous-Advantage-Actor-Critic (A3C) learning is known to quickly adapt to dynamic scenarios with less data and Residual Recurrent Neural Network (R2N2) to quickly update model parameters. Thus, we propose an A3C based real-time scheduler for stochastic Edge-Cloud environments allowing decentralized learning, concurrently across multiple agents. We use the R2N2 architecture to capture a large number of host and task parameters together with temporal patterns to provide efficient scheduling decisions. The proposed model is adaptive and able to tune different hyper-parameters based on the application requirements. We explicate our choice of hyper-parameters through sensitivity analysis. The experiments conducted on real-world data set show a significant improvement in terms of energy consumption, response time, Service-Level-Agreement and running cost by 14.4%, 7.74%, 31.9%, and 4.64%, respectively when compared to the state-of-the-art algorithms.
TL;DR: This article introduces the hierarchical structure of streaming systems, defines the scope of the resource management problem, and presents a comprehensive taxonomy in this context covering critical research topics such as resource provisioning, operator parallelisation, and task scheduling.
Abstract: Stream processing is an emerging paradigm to handle data streams upon arrival, powering latency-critical application such as fraud detection, algorithmic trading, and health surveillance. Though there are a variety of Distributed Stream Processing Systems (DSPSs) that facilitate the development of streaming applications, resource management and task scheduling is not automatically handled by the DSPS middleware and requires a laborious process to tune toward specific deployment targets. As the advent of cloud computing has supported renting resources on-demand, it is of great interest to review the research progress of hosting streaming systems in clouds under certain Service Level Agreements (SLA) and cost constraints. In this article, we introduce the hierarchical structure of streaming systems, define the scope of the resource management problem, and present a comprehensive taxonomy in this context covering critical research topics such as resource provisioning, operator parallelisation, and task scheduling. The literature is then reviewed following the taxonomy structure, facilitating a deeper understanding of the research landscape through classification and comparison of existing works. Finally, we discuss the open issues and future research directions toward realising an automatic, SLA-aware resource management framework.
TL;DR: In this paper, the authors proposed an online and adaptive job scheduling algorithm for Apache Mesos cluster, which can reduce resource usage by up to 34% under different workloads and improve job performance.
TL;DR: This study proposes a job scheduler based on a dynamic grouping integrated neighboring search strategy, which can balance the resource utilization and improve the performance and data locality in heterogeneous computing environments.
Abstract: MapReduce is a crucial framework in the cloud computing architecture, and is implemented by Apache Hadoop and other cloud computing platforms. The resources required for executing jobs in a large data center vary according to the job types. In general, there are two types of jobs, CPU-bound and I/O-bound, which require different resources but run simultaneously in the same cluster. The default job scheduling policy of Hadoop is first-come-first-served and therefore, may cause unbalanced resource utilization. Considering various job workloads, numerous job allocation schedulers were proposed in the literature. However, those schedulers encountered the data locality problem or unreasonable job execution performance. This study proposes a job scheduler based on a dynamic grouping integrated neighboring search strategy, which can balance the resource utilization and improve the performance and data locality in heterogeneous computing environments.
TL;DR: The proposed approaches are capable of reducing the cost by 58% when compared to the default Kubernetes scheduler and a rescheduling mechanism to further support the efficient use of resources by consolidating applications into fewer VMs when possible is proposed.
Abstract: Containers are widely used by organizations to deploy diverse workloads such as web services, big data, and IoT applications. Container orchestration platforms are designed to manage the deployment of containerized applications in large-scale clusters. The majority of these platforms optimize the scheduling of containers on a fixed-sized cluster and are not enabled to autoscale the size of the cluster nor to consider features specific to public cloud environments. This chapter presents a resource management approach with three objectives: 1) optimize the initial placement of containers by efficiently scheduling them on existing resources, 2) autoscale the number of resources at runtime based on the cluster's workload, and 3) consolidate applications into fewer VMs at runtime. The framework was implemented as a Kubernetes plugin and its efficiency was evaluated on an Australian cloud infrastructure. The experiments demonstrate that a reduction of 58% in cost can be achieved by dynamically managing the cluster size and placement of applications.