Server Performance Research Articles

The multiple roles that IPv6 addresses can play in today's internet

The Internet use IP addresses to identify and locate network interfaces of connected devices. IPv4 was introduced more than 40 years ago and specifies 32-bit addresses. As the Internet grew, available IPv4 addresses eventually became exhausted more than ten years ago. The IETF designed IPv6 with a much larger addressing space consisting of 128-bit addresses, pushing back the exhaustion problem much further in the future. In this paper, we argue that this large addressing space allows reconsidering how IP addresses are used and enables improving, simplifying and scaling the Internet. By revisiting the IPv6 addressing paradigm, we demonstrate that it opens up several research opportunities that can be investigated today. Hosts can benefit from several IPv6 addresses to improve their privacy, defeat network scanning, improve the use of several mobile access network and their mobility as well as to increase the performance of multicore servers. Network operators can solve the multihoming problem more efficiently and without putting a burden on the BGP RIB, implement Function Chaining with Segment Routing, differentiate routing inside and outside a domain given particular network metrics and offer more fine-grained multicast services.

Data Management Platform of Forest Ecological Station Based on Internet of Things and Big Data Sensor

In order to solve the problems of low data storage efficiency and poor retrieval performance in forest ecological station, a method of a forest ecological station data management platform based on Internet of Things and big data sensor is proposed. The framework method designs the prepartition algorithm to ensure the uniform distribution of data in the cluster. According to the characteristics of ecological data, the RowKey is scientifically designed to realize the rapid retrieval of ecological data. The Elasticsearch index fragment placement strategy based on index data and server performance evaluation is designed, and the packaging and merging strategy based on data site and time correlation is proposed to improve the storage efficiency. The results are as follows: when the scale of structured data is 108, the retrieval time of the system is 1.045 s, which is 3.99 times faster than that of the original HBase. When the scale of unstructured data is 107, the packaging small image strategy based on data site and time correlation is 1.15 times higher than that of the sequence file and 1.79 times higher than that of the original HBase. In the case of 104 concurrent users, the number of queries per second after optimization is 1.88 times higher than the original, the throughput per second is 1.74 times higher than that before optimization, and the system response time is 69.5% lower than that before optimization. The results show that the proposed scheme has significantly improved the performance in the aspects of cluster load balancing and massive structured and unstructured data retrieval efficiency and system throughput, and provide the necessary theoretical basis and technical implementation for the storage and management of forest ecological data.

Retraction Note to: Improved performance of cloud servers using LBSDD factors of private cloud

Retraction Note to: Improved performance of cloud servers using LBSDD factors of private cloud

Network Performance of Proxy-Enabled Server Using Three Configurations

Background: This study aimed to test the mean network performance of the proxy server from the Out-Of-The-Box (OOTB), caching, filtering, and bandwidth allocation configurations when categorized according to the time of access and type of websites. More so, determine significant differences in the network performance of the proxy server with the OOTB and compared it with caching, blocking, and bandwidth allocation configuration when categorized according to the time of access and type of websites. Methods: Experimental design was employed to determine the difference in the network performance of proxy servers using three configurations namely, caching, filtering, and bandwidth allocation as compared to the network performance of proxyserver with the OOTB configuration. These three configurations were made as independent variables that determined the network performance of a proxyenabled server. The researchers used the Squid Proxy version 3.4. The data were obtained by setting up a computer system running Ubuntu Linux version 19.10. Four sets of experiments were conducted to test the performance of the OOTB, caching, filtering and bandwidth allocation configurations simultaneously. The time of access was defined as A.M. session and P.M. session while the type of websites was defined as static webpages and dynamic webpages. The mean and t-Test were the statistical tools used in the study where t-Test was set at 0.05 alpha level. Findings: It was found out that the caching configuration, in both time of access and type of websites, was the most efficient among the three configurations. The proxy-enabled server with blocking configuration was found to have the least performance efficiency for both times of access and types of websites among the three configurations. Novelty : With the results of this study, it is highly recommended to network administrators to employ caching configuration on proxy servers to improve the performance efficiency of computer networks. Keywords: ProxyServer; OutOfTheBox; Caching; Filtering; Bandwidth Allocation

Live Migration of Virtualized Carrier Grade SIP Server

The concept of network virtualization, such as network functions virtualization, has attracted considerable attention from telecom carriers and a live migration technique is a key feature of virtualization technology. However, there are some challenges associated with applying server virtualization technology including live migration to a SIP server. Previous work has not dealt with the performance or behavior of a SIP server during live migration. Neither has it targeted a carrier grade SIP server for live migration. In this paper we present a virtualized carrier grade SIP server running on a virtual machine, which is configured with Carrier Grade Linux, HA middleware and SIP-AS application. We also assess its performance to investigate the impact of throughput degradation and suspension on a SIP layer and HA cluster configuration.

Queueing Subject to Activity-Dependent Server Performance: Task-Assignment Control Policies for Utilization Rate Reduction

We consider a discrete-time system comprising a first-come-first-served queue, a nonpreemptive server, and a scheduler that assigns tasks from the queue to the server. New tasks enter the queue according to a Bernoulli process with a prespecified arrival rate. At each instant, the server is either working on a task or is available. The scheduler implements a task-assignment control policy that we seek to design. When the server is available and the queue is nonempty, the policy either assigns a new task to the server or allows it to remain available (to <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">rest</i> ). In addition to the aforementioned <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">availability</i> state, we assume that the server has an integer-valued <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">activity state</i> . The activity state is nondecreasing during work periods, and is nonincreasing otherwise. In a typical application of our framework, the server performance (understood as task completion probability) worsens as the activity state increases. In this article, we build on and transcend recent stabilizability results obtained for the same framework. Specifically, we establish methods to design task-assignment control policies, which we call scheduler policies for short, which not only stabilize the queue but also reduce the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">utilization rate</i> —understood as the infinite-horizon time-averaged portion of time the server is working. This article has a main theorem leading to two key results: 1) We put forth a tractable method to determine, using a finite-dimensional linear program (LP), the infimum of all utilization rates that can be achieved by scheduler policies that are stabilizing, for a given arrival rate. 2) We propose a design method, also based on finite-dimensional LPs, to obtain stabilizing scheduler policies that can attain a utilization rate arbitrarily close to the aforementioned infimum. We also establish structural and distributional convergence properties, which are used throughout this article, and are significant in their own right.

A general method for evaluating the overhead when consolidating servers: performance degradation in virtual machines and containers

Server consolidation is one of the most commonly used techniques for reducing energy consumption in datacenters; however, this results in inherent performance degradation due to the coallocation of virtual servers, i.e., virtual machines (VMs) and containers, in physical ones. Given the widespread use of containers and their combination with VMs, it is necessary to quantify the performance degradation in these new consolidation scenarios, as this information will help system administrators make decisions based on server performance management. In this paper, a general method for quantifying performance degradation, that is, server overhead, is proposed for arbitrary consolidation scenarios. To demonstrate the applicability of the method, we develop a set of experiments with varying combinations of VMs, containers, and workload demands. From the results, we can obtain a suitable method for quantifying performance degradation that can be implemented as a recursive algorithm. From the set of experiments addressing the hypothetical consolidation scenarios, we show that the overhead depends not only on the type of hypervisor and the workload distribution but also on the combination of VMs and containers and their nesting, if feasible.

Monitoring of MQTT-based Messaging Server

In the messaging server where many devices send and receive messages, the ability to monitor the performance of the server and characteristics of the messaging is essential. This paper decides MQTT as the messaging protocol, and analyzes and defines key parameters that show the performance of the MQTT server and the characteristics of MQTT communication like MQTT message topics. In addition, this paper built an MQTT messaging server for testing by using Mosquitto as a MQTT broker and a separate monitoring system to monitor defined key parameters. In this paper, the system for monitoring the performance of the MQTT messaging server was built into three parts: a dashboard server, a monitoring application, and a test load generator. The test load generator is configured to generate a large amount of MQTT message load using 11 Raspberry PIs. The monitoring application was developed and installed on the MQTT server computer and periodically stored the server's performance and the MQTT message-related parameters processed by Mosquitto in the dashboard server's DB. The dashboard server was developed as a web server and implemented so that the administrator can view the data stored in the DB through a web browser in real time. Through experiments that generate various loads in the test load generator, it has been confirmed that the monitoring system operates normally. The monitoring system built in this paper is expected to be a good model of the monitoring system to be built together when developing an MQTT-based messaging server.

Efficient Utilization of Mirroring Servers Using Artificial Neural Networks

Due to the fact that an enormous amount of data is available to any user on the Internet, this leads to increased requests of Internet users to download or process their data among different servers. In fact, the need for increasing the reliability and performance of such servers has become an importance subject to tackle. Replicating servers is a way of reducing the overhead of Internet users’ requests as well as increasing the reliability and performance of servers. However, there is still a need to redirect users’ requests to a single server from those replications so these servers get unknown by Internet users. This technique is called mirroring servers. In this study, a new model that uses the Artificial Neural Networks (ANN) is proposed to select the appropriate server for any new user’s requests. In particular, this method considers the current features of servers with the new user’s requests as an input and provides the selected server as an output. The effectiveness of the proposed method is compared with two different techniques: Human’s selection after eliminating the required time from a user to make the selection and Round Robin selection. This comparison shows that the proposed method takes the advantages of these two techniques, which are based on the speed of selection for the Round Robin selection method and the selection of the best server according to the mirroring server features that are derived from the manual selection method. The results of this study indicate that the proposed model can improve the use of mirroring servers by 10% better than the Round Robin selection method since in this selection method, most of servers are idle in more than 25% of the time and do not have any more requests to serve.

A Perfect Knob to Scale Thread Pool on Runtime

Scalability is one of the utmost nonfunctional requirement of server applications, because it maintains an effective performance parallel to the large fluctuating and sometimes unpredictable workload. In order to achieve scalability, thread pool system (TPS) has been used extensively as a middleware service in server applications. The size of thread pool is the most significant factor, that affects the overall performance of servers. Determining the optimal size of thread pool dynamically on runtime is a challenging problem. The most widely used and simple method to tackle this problem is to keep the size of thread pool equal to the request rate, i.e., the frequency-oriented thread pool (FOTP). The FOTPs are the most widely used TPSs in the industry, because of the implementation simplicity, the negligible overhead and the capability to use in any system. However, the frequency-based schemes only focused on one aspect of changes in the load, and that is the fluctuations in request rate. The request rate alone is an imperfect knob to scale thread pool. Thus, this paper presents a workload profiling based FOTP, that focuses on request size (service time of request) besides the request rate as a knob to scale thread pool on runtime, because we argue that the combination of both truly represents the load fluctuation in server-side applications. We evaluated the results of the proposed system against state of the art TPS of Oracle Corporation (by a client-server-based simulator) and concluded that our system outperformed in terms of both; the response times and throughput.

Fault-Tolerant Load Balancing in Cloud Computing: A Systematic Literature Review

Nowadays, cloud computing is growing daily and has been developed as an effective and flexible paradigm in solving large-scale problems. It has been known as an Internet-based computing model in which computing and virtual resources, such as services, applications, storage, servers, and networks, are shared among numerous cloud users. Since the number of cloud users and their requests are increasing rapidly, the loads on the cloud systems may be underloaded or overloaded. These situations cause different problems, such as high response time and power consumption. To handle the mentioned problems and improve the performance of cloud servers, load balancing methods have a significant impact. Generally, a load balancing method aims to identify under-loaded and overloaded nodes and balance the load among them. In the recent decade, this problem has attracted a lot of interest among researchers, and several solutions have been proposed. Considering the important role of fault-tolerant in load balancing algorithms, there is a lack of an organized and in-depth study in this field yet. This gap prompted us to provide the current study aimed to collect and review the available papers in the field of fault tolerance load balancing methods in cloud computing. The existing algorithms are divided into two categories, namely, centralized and distributed, and reviewed based on vital qualitative parameters, such as scalability, makespan, reliability, resource utilization, throughput, and overhead. In this regard, other criteria such as the type of detected faults and adopted simulation tools are taken into account.

Machine Learning-based Optimal Framework for Internet of Things Networks

Deep neural networks (DNN) are widely employed in a wide range of intelligent applications, including image and video recognition. However, due to the enormous amount of computations required by DNN. Therefore, performing DNN inference tasks locally is problematic for resource-constrained Internet of Things (IoT) devices. Existing cloud approaches are sensitive to problems like erratic communication delays and unreliable remote server performance. The utilization of IoT device collaboration to create distributed and scalable DNN task inference is a very promising strategy. The existing research, on the other hand, exclusively looks at the static split method in the scenario of homogeneous IoT devices. As a result, there is a pressing need to investigate how to divide DNN tasks adaptively among IoT devices with varying capabilities and resource constraints, and execute the task inference cooperatively. Two major obstacles confront the aforementioned research problems: 1) In a heterogeneous dynamic multi-device environment, it is difficult to estimate the multi-layer inference delay of DNN tasks; 2) It is difficult to intelligently adapt the collaborative inference approach in real time. As a result, a multi-layer delay prediction model with fine-grained interpretability is proposed initially. Furthermore, for DNN inference tasks, evolutionary reinforcement learning (ERL) is employed to adaptively discover the approximate best split strategy. Experiments show that, in a heterogeneous dynamic environment, the proposed framework can provide considerable DNN inference acceleration. When the number of devices is 2, 3, and 4, the delay acceleration of the proposed algorithm is 1.81 times, 1.98 times and 5.28 times that of the EE algorithm, respectively.

Comparison of Performance Analysis of Comodo SSL and Let's Encypt SSL (Case Study: Bukalapak and Tokopedia)

The world is becoming connected with the internet, the arrival of the internet has greatly influenced new network technology. The number of internet users is increasing day by day, there is an increase in online transactions. As a growing number of websites have sprung up with easy information transmission facilities, it has been seen that it has caused a huge increase in fraud. Broad e-commerce spectrum (B2B / B2C / C2C), banking, finicial trading and others. Business applications require very high data exchange for security. Secure Sockets Layer (SSL), and its influence on web server performance. Secure SocketLayer (SSL) is the most popular protocol used on the internet to facilitate communication security through authentication, encryption, and decryption. Although the use of SSL provides adequate security, this can cause a decrease in performance compared to insecure protocols as well as a decrease in the percentage of downloads compared to insecure protocols. In this paper, we analyze the performance and impact of SSL on web performance and the type of SSL used by Bukalapak e-commerce and Tokopedia

OSov: An Interactive Web Server to Evaluate Prognostic Biomarkers for Ovarian Cancer.

Simple SummaryThe OSov web server incorporates gene expression profiles with clinical risk factors to estimate the ovarian cancers patients’ survival, and provides a tool for multiple analysis, such as forest-plot, uni/multi-variate survival analysis, Kaplan-Meier plot and nomogram construction.Ovarian cancer is one of the most aggressive and highly lethal gynecological cancers. The purpose of our study is to build a free prognostic web server to help researchers discover potential prognostic biomarkers by integrating gene expression profiling data and clinical follow-up information of ovarian cancer. We construct a prognostic web server OSov (Online consensus Survival analysis for Ovarian cancer) based on RNA expression profiles. OSov is a user-friendly web server which could present a Kaplan–Meier plot, forest plot, nomogram and survival summary table of queried genes in each individual cohort to evaluate the prognostic potency of each queried gene. To assess the performance of OSov web server, 163 previously published prognostic biomarkers of ovarian cancer were tested and 72% of them had their prognostic values confirmed in OSov. It is a free and valuable prognostic web server to screen and assess survival-associated biomarkers for ovarian cancer.

HEPiX Benchmarking Solution for WLCG Computing Resources

The HEPiX Benchmarking Working Group has developed a framework to benchmark the performance of a computational server using the software applications of the High Energy Physics (HEP) community. This framework consists of two main components, named HEP-Workloads and HEPscore. HEP-Workloads is a collection of standalone production applications provided by a number of HEP experiments. HEPscore is designed to run HEP-Workloads and provide an overall measurement that is representative of the computing power of a system. HEPscore is able to measure the performance of systems with different processor architectures and accelerators. The framework is completed by the HEP Benchmark Suite that simplifies the process of executing HEPscore and other benchmarks such as HEP-SPEC06, SPEC CPU 2017, and DB12. This paper describes the motivation, the design choices, and the results achieved by the HEPiX Benchmarking Working group. A perspective on future plans is also presented.

Rancang Bangun Sistem Monitoring Suhu dan Kelembaban Udara Pada Ruang Server Berbasis Wireless Sensor Network

A server has a pivotal role that needs certain attention when the server starts up. The server has data such as important documents stored on the server computer database. One of the factors that obstacle the server performance is a room that has a high temperature or humidity. In addition, the administrator is not able to check the state of the server room frequently. The purpose of this research is to design a temperature and humidity monitoring system in a server room based wireless sensor network. The method used is wireless sensor network. This method pursues the network administrators to monitor the conditions of the temperature and humidity in the server room. The research results show that the error generated by the sensor is relatively small. The average error generated by the temperature sensor is 0.12985% and the error generated by the air humidity sensor is 0.611538%. The collected temperature and humidity data can be displayed and stored on the web application and mobile view

Web Service Performance Analysis and Research of Distributed LAMP Architecture

There are many websites on the Internet in the early stage of development, its Web service performance is not strong enough, resulting in the failure to run the system and finally the system crash. Based on this situation, in order to ensure the accurate and efficient service of the Internet website, it is necessary to analyze and study the service performance of the Web. Distributed LAMP architecture studies and analyzes the performance of Web services, which can accurately discover the performance defects and problems of Web, and research a more appropriate configuration of service performance, and the final results presented are more accurate and effective. With the deepening development of the Internet application, the website has become a key window for many people’s information sources. Through the Internet website, we can not only understand the world more comprehensively, but also learn more knowledge. Therefore, building Web services with better performance is something we should consider. In this paper, the performance of the Web server was optimized and improved through the analysis and research of the load of the Web server, and the tuning test was carried out through distributed LAMP architecture by observing the Web service’s access to users, data information maintenance and monitoring, etc.

Advances in Microprocessor Cache Architectures Over the Last 25 Years

Over the last 25 years, the use of caches has advanced significantly in mainstream microprocessors to address the memory wall challenge. As we transformed microprocessors from single-core to multicore to manycore, innovations in the architecture, design, and management of on-die cache hierarchy were critical to enabling scaling in performance and efficiency. In addition, at the system level, as input/output (I/O) devices (e.g., networking) and accelerators (domain-specific) started to interact with general-purpose cores across shared memory, advancements in caching became important as a way of minimizing data movement and enabling faster communication. In this article, we cover some of the major advancements in cache research and development that have improved the performance and efficiency of microprocessor servers over the last 25 years. We will reflect upon several techniques including shared and distributed last-level caches (including data placement and coherence), cache Quality of Service (addressing interference between workloads), direct cache access (placing I/O data directly into CPU caches), and extending caching to off-die accelerators (CXL.cache). We will also outline potential future directions for cache research and development over the next 25 years.

Cluster load balancing algorithm based on dynamic consistent hash

In server clusters, the static scheduling algorithm has superior performance when the user visits are relatively stable. In the face of sudden increase in user traffic, the dynamic scheduling algorithm has a better load balancing effect than the static scheduling algorithm. However, in the face of complex network environments, the static scheduling algorithm cannot adjust the load according to the performance of the server in real time. The dynamic scheduling algorithm using a single weight to evaluate server performance is unreliable, and load scheduling with reference to the number of connections has uncertainty. In view of this problem, this paper proposes a cluster load balancing algorithm based on dynamic consistent hash based on the study of load balancing based on LVS clusters. By analyzing the performance and load parameters, we divide the request process into in-cycle and out-of-cycle. By setting up the LVS cluster system, the performance weights, load parameters, number of virtual nodes and cycles of the algorithm in this paper are determined experimentally. Finally, the response time and throughput of the algorithm in this paper are compared with the WRR algorithm and WLC algorithm. The results show that the time and throughput of this algorithm are better than WRR algorithm and WLC algorithm.

Deep Learning-Assisted Energy-Efficient Task Offloading in Vehicular Edge Computing Systems

In this paper, we study an energy-efficient computation offloading for vehicular edge computing systems, where multiple roadside units assist vehicular users to offload computation tasks to edge servers. Our goal is to minimize the users’ energy consumption by optimizing user association, data partition, transmit power, and computation resources, subject to the constraints of partial tasks offloading, user latency, maximum transmit power, outage performance, and computation capacity of edge servers. We utilize deep learning for user association to avoid combinatorial complexity, and develop an efficient optimization algorithm to optimize other variables. The resulting algorithm has scalable complexity with convergence guarantee, as confirmed by our theoretical analysis. Simulation results demonstrate that the introduced resource allocation algorithm can significantly reduce the total energy consumption of users.