Server Failures Research Articles

An End-Process Blockchain-Based Secure Aggregation Mechanism Using Federated Machine Learning

Federated Learning (FL) is a distributed Deep Learning (DL) technique that creates a global model through the local training of multiple edge devices. It uses a central server for model communication and the aggregation of post-trained models. The central server orchestrates the training process by sending each participating device an initial or pre-trained model for training. To achieve the learning objective, focused updates from edge devices are sent back to the central server for aggregation. While such an architecture and information flows can support the preservation of the privacy of participating device data, the strong dependence on the central server is a significant drawback of this framework. Having a central server could potentially lead to a single point of failure. Further, a malicious server may be able to successfully reconstruct the original data, which could impact on trust, transparency, fairness, privacy, and security. Decentralizing the FL process can successfully address these issues. Integrating a decentralized protocol such as Blockchain technology into Federated Learning techniques will help to address these issues and ensure secure aggregation. This paper proposes a Blockchain-based secure aggregation strategy for FL. Blockchain is implemented as a channel of communication between the central server and edge devices. It provides a mechanism of masking device local data for secure aggregation to prevent compromise and reconstruction of the training data by a malicious server. It enhances the scalability of the system, eliminates the threat of a single point of failure of the central server, reduces vulnerability in the system, ensures security, and transparent communication. Furthermore, our framework utilizes a fault-tolerant server to assist in handling dropouts and stragglers which can occur in federated environments. To reduce the training time, we synchronously implemented a callback or end-process mechanism once sufficient post-trained models have been returned for aggregation (threshold accuracy achieved). This mechanism resynchronizes clients with a stale and outdated model, minimizes the wastage of resources, and increases the rate of convergence of the global model.

An Efficient E-Voting System for Business Intelligence Innovation Based on Blockchain

AbstractBusiness intelligence (BI) is driven by data and provides valuable business insights and decision support through data analysis, mining, and visualization. The application of blockchain technology in electronic voting can make the voting process more fair and transparent. This is because the decentralized nature of blockchain technology ensures that voting data is not lost due to a single central server failure and also mitigates the risk of data tampering. However, electronic voting still faces security issues that are not easily resistant to quantum attacks. To address these challenges, we propose post-quantum cryptography and verifiable random functions for secure and efficient business intelligence electronic voting. Firstly, we propose a post-quantum verifiable random function algorithm that can resist quantum computer attacks. Secondly, we introduce a consensus algorithm based on random functions to achieve fast and efficient consensus. Thirdly, we propose a blockchain architecture based on the consensus algorithm to achieve secure and efficient blockchain applications. By integrating post-quantum verifiable random functions, consensus algorithms, and blockchain technology, we present an efficient business intelligence electronic voting system. Implementation and comparison with relevant designs demonstrate that this system provides efficient and secure electronic voting services for business intelligence users. Furthermore, the efficient consensus algorithm can be utilized to improve other blockchain applications or decentralized applications.

Workflow Scheduling Scheme for Optimized Reliability and End-to-End Delay Control in Cloud Computing Using AI-Based Modeling

In the context of cloud systems, the effectiveness of placing modules for optimal reliability and end-to-end delay (EED) is directly linked to the success of scheduling distributed scientific workflows. However, the measures used to evaluate these aspects (reliability and EED) are in conflict with each other, making it impossible to optimize both simultaneously. Thus, we introduce a scheduling algorithm for distributed scientific workflows that focuses on enhancing reliability while maintaining specific EED limits. This is particularly important given the inevitable failures of processing servers and communication links. To achieve our objective, we first develop an artificial intelligence-based model that merges an improved version of the wild horse optimization technique with a levy flight approach. This hybrid approach enhances the ability to explore new possibilities effectively. Additionally, we establish a viable strategy for sharing mapping decisions and stored information among processing servers, promoting scalability and robustness—essential qualities for large-scale distributed systems. This strategy not only boosts local search capabilities but also prevents premature convergence of the algorithm. The primary goal of this study is to pinpoint resource placements that strike a balance between global exploration and local exploitation. This entails effectively harnessing the search space and minimizing the inclination toward resources with a high likelihood of failures. Through experimentation in various system configurations, our proposed method consistently outperformed competing workflow scheduling algorithms. It achieved notably higher levels of reliability while adhering to the same EED constraints.

The Use of Web-Based Tests in an English Classroom: EFL Students’ Perceptions

One of the practical applications of technology integration in the educational field is a web-based test. Since students need to be involved in the classroom assessment decision, knowing their perceptions toward the use of evaluation media is deemed important. Thus, this study was aimed to identify the students’ perceptions on the use of web-based tests in an English class and to examine the problems experienced by students when using them in the classrooms. One hundred fourteen high school students participated in this study. This study utilized a survey which consisted of both open and close ended questions. A google form-based questionnaire was used to collect the data from the selected participants. The data were analyzed by using descriptive statistical analysis. The result of this study indicated that the students had positive perceptions of the use of web-based tests. However, they experienced some problems during the implementation of the web-based tests. The problems include the mobile device breakdown, lack of security, unfamiliarity, internet connection issue, server failure, browser incompatibility, mobile device unavailability, mobile device incompatibility, and time limitation. The study also found that the most frequent problem experienced by students was the server failure problem. Some pedagogical implications of this study were discussed.

Optimalisasi Overload Traffic Dan Request Cloud Environment Menggunakan Metode Content Delivery Network Dan Private Zone Di RCTI+

Since 2019 Indonesia has had rapid development, especially in the digital industry, recorded at the 2020 World Economic Forum. Indonesia has 2,193 startup companies engaged in the digital field. It can also affect the ability of these companies to address existing or future digital problems. As is known, down or system failure in a server can cause losses in the company's economy. For a company the size of Facebook alone, it could suffer a loss of 99.75 million US Dollars. One of the causes of server failure is Overload Traffic and Request Cloud Environment caused by the large number of requests submitted to the server. Requests submitted to the server are recorded into 2 data, namely, CCU (Count Current User) and Traffic. In this study, the author uses the CDN and Private zone methods on Alibaba Cloud to share the server load so that there is no downtime or server failure caused by Overload Traffic and Request Cloud. It is hoped that this research can be helpful to networking engineers in overcoming Traffic Overload and Request Cloud, especially on Alibaba Cloud.

Fault-tolerant unicast using conditional local safe model in the data center network BCube

As an essential infrastructure to support cloud computing services, data center networks (DCNs) are primarily used for transmitting and storing data. BCube is a large-scale DCN architecture that can effectively handle the massive data generated by network end devices due to the explosive growth of the Internet. However, as server failures in DCNs become more frequent, ensuring reliable data communication in BCube is critical. In this paper, we first establish a conditional local safe model in BCube, which divides the fault-free nodes in sub-BCube by adding some restrictions, and effectively avoids communication obstacles that may arise from faulty nodes. Then, based on the proposed model, a fault-tolerant unicast path algorithm is designed, which can construct a reliable data transmission path in the local safe sub-BCube of BCube and can tolerate more faulty nodes than existing algorithms. Finally, we conduct simulation experiments to verify that our algorithm can construct the shortest path with a high probability and achieve almost complete success in data transmission when the number of faulty nodes does not exceed half of the total nodes in BCube.

Analysis of the Efficiency of the Simulation Model of Cloud Computing in Case of Multiple Server Failures

Currently, the use of distributed (cloud) resources is becoming more and more widespread. Despite the high reliability of equipment used in call centers, failures do occur. And if long-term storage systems have high reliability due to multiple duplication, then the failure of the server on which the task is performed in specialized computing systems (SCS) can correspond to the failure of the entire system. In order to eliminate it, a simulation model of a specialized computing system built in the Simulink (SimEvent) environment, classified in the queuing theory as G/G/n/, is investigated. It is characterized by many input streams with an infinite queue, has two feedbacks, reflecting reprocessing situations in case of a server failure or no solution at the first service attempt. The system architecture is focused on the parallel execution of a certain class of tasks, divided into data-independent sub-tasks. The main criterion for the functioning of such a system is the completion of all received tasks in a time not exceeding the deadline for their implementation. The purpose of the work, the results of which are presented in the article, is to study the reliability of the functioning of such a system for various values of the probability of server failure. Since the performance and efficiency of such a system has already been studied in previous works, the main direction of research will be focused on the impact of server failures on overall productivity and efficiency under various laws of input flows.

Joint Deployment Optimization of Parallelized SFCs and BVNFs in Multi-Access Edge Computing

In multi-access edge computing (MEC) networks, parallelized service function chains (P-SFCs) can provide low-delay network services for mobile users by deploying virtualized network functions (VNFs) to process user requests in parallel. These VNFs are unreliable due to software faults and server failures. A practical way to address this is to deploy idle backup VNFs (BVNFs) near these active VNFs and activate them when active VNFs fail. However, deploying BVNFs preempts server resources and decreases the number of accepted user requests. Thus, this paper proposes a reliability enhancement approach that uses BVNFs satisfying the delay requirement as active VNFs to form P-SFCs, which contributes to the delay reduction and reliability enhancement. Since the resource capacities of edge servers can only deploy a certain number of P-SFCs and BVNFs, establishing how to deploy the minimum number of P-SFCs and BVNFs to satisfy the delay and reliability requirements of mobile users and maximize the number of accepted user requests is a challenging problem. In this paper, we first model the dynamics of delay and reliability caused by VNF parallelization and BVNFs deployment, then formulate the joint deployment problem of P-SFCs and BVNFs. Next, we design an approximation algorithm to deploy critical VNFs and BVNFs on a target edge server and schedule the data traffic of user requests processed by P-SFCs. Experimental results based on real-world datasets show that our proposed algorithm outperforms two benchmark algorithms in terms of throughput, delay, reliability, and resource utilization.

A Blockchain-Based Federated-Learning Framework for Defense against Backdoor Attacks

Federated learning (FL) is a technique that involves multiple participants who update their local models with private data and aggregate these models using a central server. Unfortunately, central servers are prone to single-point failures during the aggregation process, which leads to data leakage and other problems. Although many studies have shown that a blockchain can solve the single-point failure of servers, blockchains cannot identify or mitigate the effect of backdoor attacks. Therefore, this paper proposes a blockchain-based FL framework for defense against backdoor attacks. The framework utilizes blockchains to record transactions in an immutable distributed ledger network and enables decentralized FL. Furthermore, by incorporating the reverse layer-wise relevance (RLR) aggregation strategy into the participant’s aggregation algorithm and adding gradient noise to limit the effectiveness of backdoor attacks, the accuracy of backdoor attacks is substantially reduced. Furthermore, we designed a new proof-of-stake mechanism that considers the historical stakes of participants and the accuracy for selecting the miners of the local model, thereby reducing the stake rewards of malicious participants and motivating them to upload honest model parameters. Our simulation results confirm that, for 10% of malicious participants, the success rate of backdoor injection is reduced by nearly 90% compared to Vanilla FL, and the stake income of malicious devices is the lowest.

Blockchain-Based Data Access Control and Key Agreement System in IoT Environment.

Recently, with the increasing application of the Internet of Things (IoT), various IoT environments such as smart factories, smart homes, and smart grids are being generated. In the IoT environment, a lot of data are generated in real time, and the generated IoT data can be used as source data for various services such as artificial intelligence, remote medical care, and finance, and can also be used for purposes such as electricity bill generation. Therefore, data access control is required to grant access rights to various data users in the IoT environment who need such IoT data. In addition, IoT data contain sensitive information such as personal information, so privacy protection is also essential. Ciphertext-policy attribute-based encryption (CP-ABE) technology has been utilized to address these requirements. Furthermore, system structures applying blockchains with CP-ABE are being studied to prevent bottlenecks and single failures of cloud servers, as well as to support data auditing. However, these systems do not stipulate authentication and key agreement to ensure the security of the data transmission process and data outsourcing. Accordingly, we propose a data access control and key agreement scheme using CP-ABE to ensure data security in a blockchain-based system. In addition, we propose a system that can provide data nonrepudiation, data accountability, and data verification functions by utilizing blockchains. Both formal and informal security verifications are performed to demonstrate the security of the proposed system. We also compare the security, functional aspects, and computational and communication costs of previous systems. Furthermore, we perform cryptographic calculations to analyze the system in practical terms. As a result, our proposed protocol is safer against attacks such as guessing attacks and tracing attacks than other protocols, and can provide mutual authentication and key agreement functions. In addition, the proposed protocol is more efficient than other protocols, so it can be applied to practical IoT environments.

Assessment methodology for security of an automated control system of critical information infrastructure against DDoS attacks based on Monte Carlo simulation

Objective. The purpose of the study is to develop a methodology for assessing the security of an automated control system of critical information infrastructure from DDoS attacks. The purpose of the methodology development is to provide the decision–maker with a scientifically sound tool for assessing the risk of implementing a DDoS attack.Method. To achieve the stated goal of the study, simulation modeling based on the Monte Carlo method was used.Result. The expediency of using Monte Carlo simulation to assess the probability of server failure under DDoS attacks is confirmed. It was concluded that the server can be considered as a queuing system, however, the flow of incoming applications under DDoS attacks is not Poisson, so the use of analytical expressions to assess the probability of failure is considered incorrect. The simulation results allow the decision-maker to assess the probability of server failure and make organizational and technical decisions to increase the level of security. Analysis of the simulation results showed the effectiveness of improving server performance by increasing service channels.Conclusion. Thus, the developed methodology will be useful in conducting an information security audit of an organization to justify the amount of its insurance premium in the framework of cyber risk insurance. A possible direction for further research is to study the issue of computer network security, taking into account the features of a specific topology.

Investigating the effect of virtual machine migration accounting on reliability using a cluster model

The purpose of the article is to develop and verify with the help of mathematical modeling a software method of deploying a fault-tolerant computing cluster with a virtual machine, which consists of two physical servers (main and backup), on which a distributed data storage system with synchronous data replication from the source server to the backup server is deployed. For this purpose, the task is to conduct a computational experiment on a model of a fault-tolerant cluster, which neglects costs during recovery for the migration of virtual machines by means of the mathematical application Mathcad. Combining computing resources into clusters is a way to ensure high reliability, fault tolerance, and continuity of the computing process of computer systems. This is achieved through virtualization, which enables the movement of virtual resources, services, or applications between physical servers while maintaining the continuity of computing processes. The focus of this study is on a failover cluster, which is composed of two physical servers (primary and backup) connected through a switch, and each server has a local hard disk. A distributed storage system with synchronous data replication from the source server to the backup server is deployed on the local disks of the servers, and a virtual machine is running on the cluster. Markovian processes, flows of podias, and Kolmogorov's systems of differential equations are built into the mathematical tools of the model of a water cluster. To ensure the continuity of the computing process in case of a failure of the main server, a shadow copy of the virtual machine is launched on the backup server. The reliability of the failover cluster is measured by the coefficient of non-stationary readiness. A Markov model is proposed to assess the reliability of the failover cluster, taking into account the costs of migrating virtual machines and mechanisms that ensure the continuity of the computing process in the cluster in case of a failure of one physical server. The memory migration process maintains two copies of the virtual machine on different physical servers, enabling them to continue working on the other in the event of failure. A simplified model of the failover cluster neglects the cost of migrating virtual machines and provides an upper estimate of reliability. The study shows that the reliability of a failover cluster, as measured by the non-stationary availability factor, is significantly impacted by the virtual machine migration process. The findings of this study can be used to inform decisions about the technology chosen to ensure the failure stability and continuity of the computing process of computer systems with cluster architecture. The calculations allow us to draw a conclusion about the significant impact of virtual machine migration accounting on reliability. The calculations allow us to draw a conclusion about the significant impact of virtual machine migration accounting on reliability. The calculation was performed under the following failure rates of the server, disk, and switch: λ0 = 1,115×10-5 1/h, λ1 = 3,425×10-6 1/h, λ2 = 2,3×10-6 1/h recovery respectively: μ0 = 0,33 1/h, μ1 = 0,171/h, μ2 = 0,33 1/h. The intensity of synchronization of the distributed storage system: μ3 = 1 1/h, μ4 = 2 1/h. The difference of non-stationary cluster availability coefficients is d = К2(t) – К1(t) = 2.7×10-10

Website Traffic Forecasting Using Deep Learning

Abstract: Nowadays, there is an increasing emphasis on how to predict traffic on web pages, and there is a need to explore different methods for effectively predicting future values of multiple time series. Evaluating website traffic on a web server is crucial for web service providers because, without proper demand forecasting, customers might face long waiting times and abandon the website. However, this is a difficult task because it requires reliable predictions based on the arbitrariness of human behavior. The most effective way of transmitting information would be to predict network traffic and display it visually. Nowadays we depend too much on Google's servers, but if we wanted to host a server for many people, we might have predicted in advance the number of users to prevent server failure. Time series prediction is important in many different areas. Although there are already many systems and models for predicting Internet traffic flow, most of them use shallow traffic models and are still somewhat unsatisfactory. Therefore, we will use deep learning techniques based on current and past data to predict future traffic.

Exploring the Benefits of Resource Disaggregation for Service Reliability in Data Centers

By overcoming the server box barrier, resource disaggregation in data centers significantly improves resource utilization, which also provides a more cost-efficient approach for resource upgrade and expansion. The advantages of disaggregation have been explored in earlier research to improve the resource efficiency. This paper investigates the potential benefits of disaggregation from the aspect of reliability, which has not been considered before. Resource disaggregation brings a new failure pattern. For example, in a conventional server, the failure of one type of resource leads to the failure of the entire server, so that other types of resources in the same server also become unavailable. After disaggregating, the failure of different types of resources becomes more isolated, so that other resources are still available. In this paper, we model the reliability of a resource allocation request in a server-based or disaggregated DC based on whether the request is allocated with only working resources, or also provisioned with backup. We then consider a resource allocation problem with the objective of maximizing the number of requests accepted with guaranteed reliability. We provide an integer linear programming and a heuristic approach for this. Numerical studies demonstrate that resource disaggregation is possible to improve service reliability.

Commentary: Blockchain technology and its current importance in toxic anterior segment syndrome outbreak in India.

Commentary: Blockchain technology and its current importance in toxic anterior segment syndrome outbreak in India.

On the Age of Status Updates in Unreliable Multi-Source M/G/1 Queueing Systems

The timeliness of status message delivery in communications networks is subjective to time-varying wireless channel transmissions. In this letter, we investigate the age of information (AoI) of each source in a multi-source M/G/1 queueing update system with active server failures. In particular, we adopt the method of supplementary variables to derive a closed-form expression for the average AoI in terms of system parameters, where the server repair time follows a general distribution and the service time of packets generated by independent sources is a general random variable. Numerical results are provided to validate the effectiveness of the proposed packet serving policy under different parametric settings.

Analysis of an unreliable server with soft failures and working vacations and its application in a smart healthcare system

The majority of the studies on queues with failing servers focused on catastrophic failures. Catastrophic failures are actually a particular kind of server failure that causes system clearing. Routine errors, unlike catastrophic failures, do not destroy the system; instead, all users must wait for the server to be activated again. In this study, with an application to a smart healthcare system, the analytical equations for the transient state probability of an unreliable server with soft failures and a working vacation are presented for the first time. The performance metrics are also developed. Numerical visuals are developed to support theoretical results.

Blockchain-Based Decentralized Federated Learning Method in Edge Computing Environment

In recent years, federated learning has been able to provide an effective solution for data privacy protection, so it has been widely used in financial, medical, and other fields. However, traditional federated learning still suffers from single-point server failure, which is a frequent issue from the centralized server for global model aggregation. Additionally, it also lacks an incentive mechanism, which leads to the insufficient contribution of local devices to global model training. In this paper, we propose a blockchain-based decentralized federated learning method, named BD-FL, to solve these problems. BD-FL combines blockchain and edge computing techniques to build a decentralized federated learning system. An incentive mechanism is introduced to motivate local devices to actively participate in federated learning model training. In order to minimize the cost of model training, BD-FL designs a preference-based stable matching algorithm to bind local devices with appropriate edge servers, which can reduce communication overhead. In addition, we propose a reputation-based practical Byzantine fault tolerance (R-PBFT) algorithm to optimize the consensus process of global model training in the blockchain. Experiment results show that BD-FL effectively reduces the model training time by up to 34.9% compared with several baseline federated learning methods. The R-PBFT algorithm can improve the training efficiency of BD-FL by 12.2%.

A State Dependent Arrival Analysis In a Non-Markovian Bulk Queue with Server Failures

A State Dependent Arrival Analysis In a Non-Markovian Bulk Queue with Server Failures

A Study On Decentralized Web Hosting Using Peer-To-Peer Architecture

Traditionally, websites are hosted viacentralized network, where the web servers distribute the website files to the clients. However a server failure can prevent the web applications from being used until the server goes live again. Due to the fact that websites rely entirely on the integrity of a single organization, it enables larger companies or government to decide what content is shown, this put into question the freedom that the internet originally brought. The block chain hosting service, also known as decentralized web hosting, is not controlled by a single organization or a third party organization. The architecture relies on peer-to-peer communication rather than client-to-server communication. Due to the decentralized nature of this system, all clients do not rely on a single server for their data, but instead data is sent directly between the clients across the network. It is possible to achieve this with the help of IPFS and Block chain. IPFS is a protocol for distributed file storage that allows computers to store and serve files in a peer-to-peer network by distributing them across computers and with the block chain technology, it is possible to ensure both authenticity as well as confidentiality.