Design Of Server Research Articles

Dense Server Design for Immersion Cooling

The growing demands for computational power in cloud computing have led to a significant increase in the deployment of high-performance servers. The growing power consumption of servers and the heat they produce is on track to outpace the capacity of conventional air cooling systems, necessitating more efficient cooling solutions such as liquid immersion cooling. The superior heat exchange capabilities of immersion cooling both eliminates the need for bulky heat sinks, fans, and air flow channels while also unlocking the potential go beyond conventional 2D blade servers to three-dimensional designs. In this work, we present a computational framework to explore designs of servers in three-dimensional space, specifically targeting the maximization of server density within immersion cooling tanks. Our tool is designed to handle a variety of physical and electrical server design constraints. We demonstrate our optimized designs can reduce server volume by 25--52% compared to traditional flat server designs. This increased density reduces land usage as well as the amount of liquid used for immersion, with significant reduction in the carbon emissions embodied in datacenter buildings. We further create physical prototypes to simulate dense server designs and perform real-world experiments in an immersion cooling tank demonstrating they operate at safe temperatures. This approach marks a critical step forward in sustainable and efficient datacenter management.

Numerical investigation of the influence of heat-generating components on the heat dissipation in a tower server

Many of today’s servers rely on high-power electronic components. During continuous operation, elevated temperatures can lead to sluggish performance and system instability. Therefore, servers urgently require safe and reliable heat dissipation systems. In this work, we focus on a highly scalable tower server as our research subject, exploring its maximum configuration within the designated range. We aim to enhance the server’s thermal performance through a parametric investigation of factors such as the server air outlet, the FAR in air outlet, and the layout of GPUs. The results demonstrate that narrowing the air outlet can effectively reduce the temperature of GPU-2 by 2.83 °C. Further analysis reveals that an optimal FAR of 0.74 for the air outlet leads to a temperature decrease of 1.67 °C for GPU-2 compared to a FAR of 0.24. Moreover, adjusting the position of GPU-2, specifically employing the VLO-L71.16-1 structure with the air outlet positioned on the GPU-1 side, yields optimal heat dissipation performance, resulting in a remarkable temperature decrease of 16.63 °C for GPU-2. Additionally, it was observed that GPU-2 in the base case approaches its limiting temperature. By optimizing the structure to VLO-L71.16-1, the study managed to reduce fan airflow while maintaining GPU-2 within safe operating temperatures. Specifically, the optimal structure achieves approximately 35 % airflow savings when GPU-2 reaches its limiting temperature. This research provides valuable insights for the exploration and design of novel server cooling systems.

A Design of an SMTP Email Server

This study developed a mail server program using Socket API and Python. The program uses the Hypertext Transfer Protocol (HTTP) to receive emails from browser clients and forward them to actual email service providers via the Simple Mail Transfer Protocol (SMTP). As a web server, it handles Transmission Control Protocol (TCP) connection requests from browsers, receives HTTP commands and email data, and temporarily stores the emails in a file. Simultaneously, as an SMTP client, the program establishes a TCP connection with the actual mail server, sends SMTP commands, and transmits the previously saved emails. In addition, we also analyzed security issues and the efficiency and availability of this server, providing insights into the design of SMTP mail servers.

Micro-FL: A Fault-Tolerant Scalable Microservice-Based Platform for Federated Learning

As the number of machine learning applications increases, growing concerns about data privacy expose the limitations of traditional cloud-based machine learning methods that rely on centralized data collection and processing. Federated learning emerges as a promising alternative, offering a novel approach to training machine learning models that safeguards data privacy. Federated learning facilitates collaborative model training across various entities. In this approach, each user trains models locally and shares only the local model parameters with a central server, which then generates a global model based on these individual updates. This approach ensures data privacy since the training data itself is never directly shared with a central entity. However, existing federated machine learning frameworks are not without challenges. In terms of server design, these frameworks exhibit limited scalability with an increasing number of clients and are highly vulnerable to system faults, particularly as the central server becomes a single point of failure. This paper introduces Micro-FL, a federated learning framework that uses a microservices architecture to implement the federated learning system. It demonstrates that the framework is fault-tolerant and scalable, showing its ability to handle an increasing number of clients. A comprehensive performance evaluation confirms that Micro-FL proficiently handles component faults, enabling a smooth and uninterrupted operation.

Lightweight Federated Learning for Rice Leaf Disease Classification Using Non Independent and Identically Distributed Images

Rice (Oryza sativa L.) is a vital food source all over the world, contributing 15% of the protein and 21% of the energy intake per person in Asia, where most rice is produced and consumed. However, bacterial, fungal, and other microbial diseases that have a negative effect on the health of plants and crop yield are a major problem for rice farmers. It is challenging to diagnose these diseases manually, especially in areas with a shortage of crop protection experts. Automating disease identification and providing readily available decision-support tools are essential for enabling effective rice leaf protection measures and minimising rice crop losses. Although there are numerous classification systems for the diagnosis of rice leaf disease, no reliable, secure method has been identified that meets these needs. This paper proposes a lightweight federated deep learning architecture while maintaining data privacy constraints for rice leaf disease classification. The distributed client–server design of this framework protects the data privacy of all clients, and by using independent and identically distributed (IID) and non-IID data, the validity of the federated deep learning models was examined. To validate the framework’s efficacy, the researchers conducted experiments in a variety of settings, including conventional learning, federated learning via a single client, as well as federated learning via multiple clients. The study began by extracting features from various pre-trained models, ultimately selecting EfficientNetB3 with an impressive 99% accuracy as the baseline model. Subsequently, experimental results were conducted using the federated learning (FL) approach with both IID and non-IID datasets. The FL approach, along with a dense neural network trained and evaluated on an IID dataset, achieved outstanding training and evaluated accuracies of 99% with minimal losses of 0.006 and 0.03, respectively. Similarly, on a non-IID dataset, the FL approach maintained a high training accuracy of 99% with a loss of 0.04 and an evaluation accuracy of 95% with a loss of 0.08. These results indicate that the FL approach performs nearly as well as the base model, EfficientNetB3, highlighting its effectiveness in handling both IID and non-IID data. It was found that federated deep learning models with multiple clients outperformed conventional pre-trained models. The unique characteristics of the proposed framework, such as its data privacy for edge devices with limited resources, set it apart from the existing classification schemes for rice leaf diseases. The framework is the best alternative solution for the early classification of rice leaf disease because of these additional features.

OWASP Framework-based Network Forensics to Analyze the SQLi Attacks on Web Servers

One of dangerous vulnerabilities that attack the web is SQLi. With this vulnerability, someone can obtain user data information, then change and delete that data. The solution to this attack problem is that the design website must improve security by paying attention to input validation and installing a firewall. This study's objective is to use network forensic tools to examine the designlink website's security against SQLi attacks, namely Whois, SSL Scan, Nmap, OWASP Zap, and SQL Map. OWASP is the framework that is employed; it is utilized for web security testing. According to the research findings, there are 14 vulnerabilities in the design website, with five medium level, seven low level, and two informational level. When using SQL commands with the SQL Map tool to get username and password information on its web server design. The OWASP framework may be used to verify the security of websites against SQLi attacks using network forensic tools, according to the study's findings. So that information about the vulnerabilities found on the website can be provided. The results of this study contribute to forensic network knowledge against SQLi attacks using the OWASP framework as well as for parties involved in website security.

A Traffic Model Based Approach to Parameter Server Design in Federated Learning Processes

A Traffic Model Based Approach to Parameter Server Design in Federated Learning Processes

Cloud server design for heavy workload gaming computing with Google cloud platform

Cloud servers are generally used for data storage and remote office activities, but it can be applied for gaming purposes, where cloud servers can be paired with virtual machines and gaming platform that can be accessed by users via an internet connection. This makes the device used by user no longer needs to process resources because the workload is carried out by virtual machines on cloud server. The author designs a cloud gaming system using Google cloud platform as a cloud server and parsec as an optimizer that is attached to a virtual machine for game computing purposes. Author takes measurements of the cloud gaming system using 2 test games varying from low to middle specifications. Resource testing on central processing unit (CPU) and random-access memory (RAM) usage on the user side is below 40% when running game 1 and below 44% when running game 2, while on the system it reaches a capacity above 40% for CPU and RAM and 99% maximum on graphics processing unit (GPU). Quality of service testing of the system is carried out at bandwidths of 5, 10, and 30 Mbps with a minimum bandwidth of 10 Mbps. In general, there are a little difference that occurred between test game and different bandwidths.

VIPER HTTP WEB SERVER (A Custom Web Server)

This research paper presents the design and implementation of a Web Server designed in C++ which is a Custom Web server that is it can be used for specific tasks which makes it faster than other open- source web servers it can also be used to meet the specific needs of a particular organization. The paper discusses the challenges faced during the development process and the solutions that were implemented to overcome them. Performance tests were conducted to evaluate the server's efficiency and effectiveness using different web applications that interact with Data Bases and external Third Party APIs, and the results demonstrate that it performs at a high level even under heavy loads. Overall, this paper provides valuable insights into the development of custom web servers and highlights the benefits of creating a tailored solution for specific organizational needs. Keywords: - Customized Web Server, Web Services, HTTP Protocols, Web Applications

FedDual: Pair-Wise Gossip Helps Federated Learning in Large Decentralized Networks

There is a significant recent interest in collaboratively training a machine learning (ML) model without collecting data to a central server. Federated learning (FL) emerges as an efficient solution mitigating systemic privacy risks and communication costs. However, conventional FL inherited from parameter server designs relies too much on a central server, which may lead to privacy risks, communication bottlenecks, or a single point of failure. In this paper, we propose an asynchronous and hierarchical local gradient aggregation and global model update algorithm, FedDual, under three different security considerations for FL in large decentralized networks. Particularly, FedDual preserves privacy by introducing local differential privacy (LDP) and aggregates local gradients asynchronously and hierarchically via a pair-wise gossip algorithm, which is more competitive than previous gossip-based decentralized FL methods in terms of privacy preservation and communication efficiency, and offers more computational efficiency compared to existing blockchain-assisted decentralized FL methods. Further, we devise a noise cutting trick based on Private Set Intersection (PSI) to mitigate the prediction performance loss of the global model caused by the leveraged LDP. Rigorous analyses show that FedDual helps decentralized FL achieve the same convergence rate of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mathcal {O}\left({\frac {1}{T}}\right) $ </tex-math></inline-formula> as centralized ML theoretically. Ingenious experiments on MNIST, CIFAR-10, and FEMNIST confirm that the model prediction performance gained from FedDual is close to centralized ML. More importantly, the proposed noise cutting trick helps FedDual to train better global models than LDP-based FL methods in terms of prediction performance and convergence rate.

Server load and network-aware adaptive deep learning inference offloading for edge platforms

This work presents DIAMOND, a deep neural network computation offloading scheme consisting of a lightweight client-to-server latency profiling component combined with a server inference time estimation module to accurately assess the expected latency of a deep learning model inference. Latency predictions for both the network and server are comprehensively used to make dynamic (partial) model offloading decisions at the client in run-time. Compared to previous work, DIAMOND targets to minimize network latency estimation overhead and considers the concurrent processing nature of state-of-the-art deep learning inference server designs. Our extensive evaluations with an NVIDIA Jetson Nano client connected to an NVIDIA Triton server shows that DIAMOND completes inference operations with noticeably reduced computational/energy overhead and latency compared to previously proposed model offloading approaches. Furthermore, our results show that DIAMOND well-adapts to practical server load and network dynamics.

Simulasi Kinerja Web Server Pada Jaringan LAN (Local Area Network) Kampus Menggunakan NS2 (Network Simulator 2)

Web server is a computer that connected to the network or internet which can give information such as text, audio and video. To reach efficiency of design and optimal reliable level, a big scale web server needs that measurement of network design performance. This final task discussed about modeling and simulation of web server design on Local Area Network University of Lampung to find out reachable optimal performance of web server. Research has been done by create modeling of client and server on software simulator. And simulation of web server model design on LAN network has been done by make implementation of simulation scenario. Modeling and simulation will give approximation on web server performance level on LAN network as the consideration of the making of principle realization of web server on LAN network. Based on the result of simulation, web server optimum design on LAN network is on the design with 100 nodes. The design gives average performance level of throughput almost reach the average maximum value of throughput with the minimum average of delay and jitter and the lowest percentage packet loss performance compared with the simulation which is used the other quantity of node.

Rancangan Virtual Private Server Pada Kantor Kelurahan Menggunakan ZeroTier

Banyuanyar Village is a guided village developed by the Boyolali Regency Government to become an internet village. The operational activities of the Banyuanyar Urban Village Office also utilize the existing internet network. However, its use is still not optimal and is at risk of being attacked because it uses the same network as general internet users and the Banyuanyar Urban Village Office does not yet have a central server as a database due to limited funds and technicians. Because of this, the activities of the staff who work become sluggish and less effective. Therefore, it takes a Virtual Private Server Design that is useful to support performance. This design uses ZeroTier which can be used as an open-source VPS and VPN and has the latest technology. This research provides effectiveness and security regarding data transfer and network in the Banyuanyar Village Office. Because ZeroTier can be controlled and supervised by admins who can later provide convenience in network management.

Web Server Analysis

Abstract: This is a design and implementation of a Web Server. “WEB SERVER ANALYSIS” is a software developed for managing log files on the server. Web log analysis is an innovative and unique field constantly formed and changed by the convergence of various emerging Web technologies. Due to its interdisciplinary character, the diversity of issues it addresses, and the variety and number of Web applications, it is the subject of many distinctive and diverse research methodologies. This project deals with the analysis of the log files of a web server. Keywords: Web server, Log files, Error, URL, Resources

IoT Network of Sensor Array for Intrusion Detection and Diagnosis of Electrical Systems

Modern buildings consist of various equipment, including heating, ventilation, air conditioning (HVAC), and lighting. All equipment can be monitored and managed by the building management system. All of these components can be damaged due to prolonged use, misconfiguration, and network connection problems. Equipment breakdown affects maintenance costs and, in particular, energy efficiency. This study aims to develop a monitoring system of the current consumption of lighting (lamps) by light detection and current consumption of air conditioning (AC) by room temperature detection using Internet of Things (IoT) implementation. Hardware design consists of a power supply circuit, installing an ACS 712 current sensor, LDR sensor, the temperature sensor of DHT22, and thermal sensor of LM35. While the software design consists of a diagram flow for the current sensor, light sensor, temperature sensor reading program, program on the display board, and a web server design. The detection of current, lamplight, room temperature, and thermal cable is carried out to determine errors that occur in electrical equipment. Monitoring the consumption of lighting flows by detecting lamp light and air conditioning current consumption by detecting room temperature is done through the Firebase web server using a computer or smartphone. The results showed that the built system could monitor current consumption, detect lamplight, and detect room temperature in real-time. This system can be used to detect faulty electrical equipment and determine its position so that repairs can be carried out immediately. However, the type of damage has not been identified.

Perancangan Server Kantor Desa Tomuan Holbung Berbasis Client Server

Abstract− In the era of regional autonomy, the village is a very important part in determining the success of development and in direct contact with the community, especially in public services. Server is a medium that has an important role to provide internet access services for its users. So that the network can be used optimally, it is necessary to have a server. This study aims to design a server in the Tomuan Holbung Village Office, Asahan Regency by building an efficient, flexible server infrastructure and optimizing resource use, so that it can be adjusted to the workload so that hardware resources are optimal. This research was conducted to be a solution for the use of hardware resources provided by the Server machine that has not been used optimally, so efficient steps are needed to overcome this problem. The results of the research obtained are in the form of server design and optimizing the use of resources and managing servers efficiently in terms of time and cost.&#x0D;

Power-optimized Deployment of Key-value Stores Using Storage Class Memory

High-performance flash-based key-value stores in data-centers utilize large amounts of DRAM to cache hot data. However, motivated by the high cost and power consumption of DRAM, server designs with lower DRAM-per-compute ratio are becoming popular. These low-cost servers enable scale-out services by reducing server workload densities. This results in improvements to overall service reliability, leading to a decrease in the total cost of ownership (TCO) for scalable workloads. Nevertheless, for key-value stores with large memory footprints, these reduced DRAM servers degrade performance due to an increase in both IO utilization and data access latency. In this scenario, a standard practice to improve performance for sharded databases is to reduce the number of shards per machine, which degrades the TCO benefits of reduced DRAM low-cost servers. In this work, we explore a practical solution to improve performance and reduce the costs and power consumption of key-value stores running on DRAM-constrained servers by using Storage Class Memories (SCM). SCMs in a DIMM form factor, although slower than DRAM, are sufficiently faster than flash when serving as a large extension to DRAM. With new technologies like Compute Express Link, we can expand the memory capacity of servers with high bandwidth and low latency connectivity with SCM. In this article, we use Intel Optane PMem 100 Series SCMs (DCPMM) in AppDirect mode to extend the available memory of our existing single-socket platform deployment of RocksDB (one of the largest key-value stores at Meta). We first designed a hybrid cache in RocksDB to harness both DRAM and SCM hierarchically. We then characterized the performance of the hybrid cache for three of the largest RocksDB use cases at Meta (ChatApp, BLOB Metadata, and Hive Cache). Our results demonstrate that we can achieve up to 80% improvement in throughput and 20% improvement in P95 latency over the existing small DRAM single-socket platform, while maintaining a 43–48% cost improvement over our large DRAM dual-socket platform. To the best of our knowledge, this is the first study of the DCPMM platform in a commercial data center.

The significance of product design in the circular economy: A sustainable approach to the design of data centre equipment as demonstrated via the CEDaCI design case study

Society depends on fast, uninterrupted 24/7 data transfer in almost every aspect of our lives. Data centres (DC) across the globe process billions of gigabytes of data every day. There are currently about 7 million data centres globally and the sector is expected to grow fivefold by 2030, inevitably increasing already high demand for millions of tonnes of resources, including critical raw materials (CRM). The infrastructure for recycling sectoral waste is lacking, wasting millions of tonnes of materials. Many virgin materials are located in conflict zones, which jeopardises the supply chain that the sector heavily relies on, consequently posing a danger of severe material shortages in the future. Present design thinking does not balance optimisation of the performance and operational energy requirements and designing for circularity. To date, server design has mainly been focused on energy efficiency. However, a Circular Economy (CE) entails consideration of a complete life cycle of the product, including embodied impacts. Because sectoral data processing and energy requirements continue to grow, sustainable design and innovation in line with the 2020 EU Circular Economy Action Plan are crucial in promoting best practices and managing the sector’s energy demand in the future. This paper describes a proposed sustainable server design developed as part of the CEDaCI (a Circular Economy for the Data Centre Industry) project and based on the design requirements outlined by the EU Circular Economy Action Plan 2020 together with other directives. It presents recommendations for good practice methods, futureproofing data centre equipment design for a Circular Economy and building a critical foundation for sustainable hardware design that can be developed in the future.

DESIGN AND IMPLEMENTATION OF A WIRELESS SENSOR NETWORK WEB SERVER

Remote control networks have become a key area of research within wireless communication technology. Zigbee technology fills the gap in the market for low-cost, low-power wireless solutions. This paper presents the design and application of a wireless sensor network web server based on the S3C2410 microcontroller and the Zigbee protocol. The server enables real-time monitoring and control of remote objects. Using a B/S architecture and modular design approach, both hardware and software components are detailed, with an emphasis on key implementation techniques. The system has been demonstrated to operate stably with excellent scalability, making it suitable for applications in smart homes, monitoring systems, industrial control, and other domains.

Design of reliable video image transmission system for underwater vehicle

In order to improve the transmission performance of underwater vehicle motion video image, a reliable transmission system of underwater vehicle motion video image is designed. The hardware part of the system includes the design of the transmission server and the design of the transmission client; the software part of the system mainly includes two parts: compression processing and lossless transmission of the motion video image of the underwater vehicle, which completes the software design of the system and realises the design of the whole system. The test results show that the proposed transmission system has good transmission performance for the motion video image of the vehicle in shallow water.