This paper presents a comprehensive survey of the state-of-the-art work on fault tolerance methods proposed for cloud computing. The survey classifies fault-tolerance methods into three categories: 1) ReActive Methods (RAMs); 2) PRoactive Methods (PRMs); and 3) ReSilient Methods (RSMs). RAMs allow the system to enter into a fault status and then try to recover the system. PRMs tend to prevent the system from entering a fault status by implementing mechanisms that enable them to avoid errors before they affect the system. On the other hand, recently emerging RSMs aim to minimize the amount of time it takes for a system to recover from a fault. Machine Learning and Artificial Intelligence have played an active role in RSM domain in such a way that the recovery time is mapped to a function to be optimized (i.e., by converging the recovery time to a fraction of milliseconds). As the system learns to deal with new faults, the recovery time will become shorter. In addition, current issues and challenges in cloud fault tolerance are also discussed to identify promising areas for future research.

Toward a Smart Cloud: A Review of Fault-Tolerance Methods in Cloud Systems

Recurrent deep learning-based feature fusion ensemble meta-classifier approach for intelligent network intrusion detection system

An optimized weighted voting based ensemble model for DDoS attack detection and mitigation in SDN environment

Large-scale MapReduce clusters that routinely process petabytes of unstructured and semi-structured data represent a new entity in the changing landscape of clouds. A key challenge is to increase the utilization of these MapReduce clusters. In this work, we consider a subset of the production workload that consists of MapReduce jobs with no dependencies. We observe that the order in which these jobs are executed can have a significant impact on their overall completion time and the cluster resource utilization. Our goal is to automate the design of a job schedule that minimizes the completion time (makespan) of such a set of MapReduce jobs. We offer a novel abstraction framework and a heuristic, called BalancedPools, that efficiently utilizes performance properties of MapReduce jobs in a given workload for constructing an optimized job schedule. Simulations performed over a realistic workload demonstrate that 15%-38% makespan improvements are achievable by simply processing the jobs in the right order.

Two sides of a coin: Optimizing the schedule of map reduce jobs to minimize their makespan and improve cluster performance

Software-defined networking (SDN) emerges as an innovative network paradigm, which separates the control plane from the data plane to improve the network programmability and flexibility. It is widely applied in the Internet of Things (IoT). However, SDN is vulnerable to DDoS attacks, which can cause network disasters. In order to protect SDN security, a DDoS detection method using cloud-edge collaboration based on Entropy-Measuring Self-organizing Maps and KD-tree (EMSOM-KD) is designed for SDN. Entropy measurement is utilized to select the ideal SOM map and classify SOM neurons considering the limitation of dead and suspicious neurons. EMSOM can detect most flows directly and filter out a few doubtable flows. Then these flows are fine-grained, identified by KD-tree. Due to the limited and precious resources of the controller, parameter computation is performed in the cloud. The edge controller implements DDoS detection by EMSOM-KD. The experiments are conducted to evaluate the performance of the proposed method. The results show that EMSOM-KD has better detection accuracy; moreover, it improves the KD-tree detection efficiency.

/pdf/ddos-detection-using-a-cloud-edge-collaboration-method-based-1pi0c1xema.pdf

DDoS Detection Using a Cloud-Edge Collaboration Method Based on Entropy-Measuring SOM and KD-Tree in SDN

Software defined networking (SDN) is a networking paradigm that allows for the easy programmability of network devices by decoupling the data plane and the control plane. On the other hand, Distributed Denial of Service (DDoS) attacks remains one of the major concerns for organizational network infrastructures and Cloud providers. In this article, the authors propose a Neural Network based Traffic Flow Classifier (TFC-NN) for live DDoS detection in SDN environments. This study provides a live traffic analysis method with a neural network. The training of the TFC-NN model is performed by a labelled dataset constructed from SDN normal traffic and an-under DDoS traffic. The study also provides a live mitigation process combined with the live TFC-NN-based DDoS detection. The approach is deployed and evaluated on an SDN architecture based on different performance metrics with different under-DDoS attack scenarios.

Neural Network-Based Approach for Detection and Mitigation of DDoS Attacks in SDN Environments

In the Big Data era, MapReduce has been the most utilized model in academia and industry. The main objective of MapReduce and its well-known implementation Hadoop is to run distributed applications to analyze huge amount of datasets, on very large clusters of commodity machines. This can be very time consuming. Also, the cluster that runs MapReduce applications has to be very scalable. In order to improve the performance of the MapReduce model, efficient scheduling algorithms have to be developed. Now day, Scalability is a major concern for applications that works on datasets measured by petabytes. Schedulers for such applications must allow a high scalability and an optimal utilization of resources in order to minimize the execution time. For this end we propose MR-SPS, a scalable parallel scheduling algorithm that takes care of scalability of the cluster and its performance by managing workload and data locality. Our scheduler works in a parallel manner, which allows a higher use of the capacities that the Resource-manager can eventually provide, this will increase the number of node that it can manage. For experiments, we have used CloudSim to simulate MR-SPS and other scheduling algorithms designed for MapReduce model. The results show the superiority of our implementation.

MR-SPS: Scalable parallel scheduler for YARN/MapReduce platform

Cloud computing is an emerging paradigm where computing services are provided across the web. Virtualization powers the cloud by mutualizing physical resources thus ensuring flexibility and high availability of the cloud. Certainly fault tolerance like load balancing or advancement programming security aim to foster availability but classic reactive fault tolerance techniques prove to be greedy in terms of memory and recovery time. Elsewhere, proactive fault tolerance is possible by preemptive virtual machine migration requiring a strong and accurate failure predictor. In quest of an effective approach for proactive fault tolerance we introduce in this paper a probabilistic model of the cloud with a failure generator for evaluating a proposed approach based on three scenarios of virtual machine migration.

Probabilistic model for evaluating a proactive fault tolerance approach in the cloud

Cloud computing is an emerging paradigm that consists of hosting and delivering computing services across the web. The availability is one of the security features such as integrity and confidentiality. Certainly endorsing high availability by the improvement of fault tolerance techniques is one of the major concerns of the cloud. Elsewhere we cannot afford to directly evaluate new approaches for cost reason. For this reason we introduce in this paper a probabilistic model for simulation based on the principle of “Large Numbers Law”. The idea is to simulate a scenarios of cloud virtual environment in which faults can occur in a random way following failure occurrence probabilities. The global unavailability measured is faithful to unavailability average known of Cloud providers. The model allows live virtual machine migration in order to evaluate proactive fault tolerance approaches based on preemptive migration.

Oussama Hannache

Papers

Neural Network-Based Approach for Detection and Mitigation of DDoS Attacks in SDN Environments

MR-SPS: Scalable parallel scheduler for YARN/MapReduce platform

Probabilistic model for evaluating a proactive fault tolerance approach in the cloud

Cloud Simulation Model Based on Large Numbers Law for Evaluating Fault Tolerance Approaches