National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”

Imagine constructing your infrastructure like a multi-layered cake. This metaphor isn’t just for culinary enthusiasts; it’s a practical framework for organizing your repository’s folder structure, selecting the appropriate tools, and enhancing security right from the start.

This chapter is where things get more complex, but don’t worry—we’ll start with the basics and build from there.

It’s impossible to predict exactly what kind of application you’re going to create, but rest assured—we’re prepared for anything. You might need to add another layer or folder where you’ll set up a database or other types of storage. However, I want to share how I tackled my own challenges with great success. The solution I developed not only solved my problem but also earned high praise from everyone who used it. They were impressed, and I received a lot of positive feedback. Let’s dive into the issue I faced and how I overcame it.

GitLab is more than just a Git repository manager—it’s a comprehensive DevOps platform that integrates source control, CI/CD pipelines, and even project management tools all in one place. Developers appreciate GitLab for its flexibility and powerful features that go beyond what GitHub or Bitbucket typically offer. One of GitLab’s standout features is its built-in CI/CD capabilities, which allow developers to automate testing, deployment, and more directly from within the platform.

In a typical Kubernetes setup, you start with a standard configuration: three master nodes and multiple worker nodes. This setup provides a balance of high availability and resource distribution, ensuring your cluster is resilient and scalable. However, managing these nodes, especially in a cloud environment like AWS, can become complex and resource-intensive.

In Kubernetes, a storage class is a way to define the “classes” of storage available in your cluster. It abstracts the underlying storage provider (e.g., AWS EBS, Azure Disk, Google Persistent Disk) and provides a way to dynamically provision storage volumes. A storage class defines the type of storage (like SSDs or spinning disks), the performance characteristics, and other parameters that determine how the storage behaves.

Okay, imagine you are the CTO of a big IT company providing a platform as a service. So you have hundreds of thousands of applications running at the same time. Applications of very different kinds. Where would you start? Well, if you are like me, you would start with the database. Kubernetes team chose etcd for that role. Let me explain why.

The paper notes that the analytic hierarchy process is one of the most widely used multi-criteria methods, and its calculation procedures have repeatedly become the object of various improvements and modifications. The presence of many variants of the analytic hierarchy process may result in the dependence of the determination of the best alternatives on the used variant of this method. The paper describes three variants of the analytic hierarchy process: original, simplified, and fuzzy. Based on solving a known example, these three variants confirmed that they can lead to different rankings. The paper concludes that it is advisable to carry out calculations according to several variants and to decide the superiority of one alternative over another only in case of convergence of the results of their comparison according to the majority of used variants to increase the validity of the recommendations for the person making the decision.

A frontrunning attack is one of the most common attacks on smart contracts. Its essence is in manipulating the order of transaction inclusion in a block to gain an advantage by altering the transaction processing sequence. This attack poses a particular threat to the conduct of p2p auctions for “green” energy trading. In this article, different types of such attacks are considered, analyzed, and formalized in step-by-step execution algorithms. Next, a model is proposed to estimate the probability of success of such an attack. An explicit formula is derived for the probability of success of a displacement attack and an insertion attack, which are particular cases of a frontrunning attack. The probability of success is shown to depend on network parameters and the ratio between the transaction fees created by an honest and a malicious users. Numerical examples of practical applications of the derived formula are provided, further confirming the correctness of the analytical results.

The authors analyze the influence of variable technological loads on the main line of the rolling mill and present a mathematical model of dynamic processes in group electric drives of crimping technological machine units (rolling mills). The dependence of the moment of technological resistance on the speeds of the working rolls is shown to be determined by the formula that takes into account the peculiarities of the technological load. The study of the influence of variable technological loads on the main line of the rolling mill showed that an increase in the coefficient of friction on the upper roll is the most dangerous. It was found out that in these cases, there can be a significant mismatch of the speeds of the working rolls, and the drop in the moment of technological resistance on the upper roll is not compensated by the increase in the moment of technological resistance on the lower roll; therefore, self-oscillations, as well as unstable oscillation modes, are excited in the elastic links of the main line of the rolling mill.

A multichannel queueing system with a Poisson input flow is considered. The input flow rate depends on the current state of the semi-Markov process. This state also determines the type of the customer. Each channel contains several service lines. To be serviced, a customer requires several lines depending on the respective distribution. If the channel does not have sufficient service lines, it is possible to redirect this customer to another channel. The service time has a general distribution function depending on the customer’s type and the number of lines required for servicing. A fast simulation method is proposed to evaluate the blocking probability of customers of a certain type demanding a given number of service lines. A comparison with the Monte Carlo method is made using a numerical example, and the change in the relative error of the estimate as the blocking probability decreases is analyzed.

Forecasting the destruction of shallow subway stations by explosions is critical for ensuring the safety of passengers and subway personnel. The article discusses the methods of modeling and analysis of destruction in the Ansys program that may occur as a result of explosions at shallow stations. Using modern computer models and software for the simulation of explosive processes, the behavior of structures under the influence of blast waves is investigated. Particular attention is paid to identifying the most vulnerable areas of the plants and developing recommendations for increasing their stability. Various explosion scenarios are considered, including different explosion powers and locations. In addition, materials and design solutions that can minimize the consequences of explosions are studied. The research results can be used to develop effective safety measures and improve existing engineering solutions. Thus, the work aims to improve the level of subway safety by introducing scientifically based methods for predicting and preventing station destruction from explosions.

Analysis of the provision of protection of critical energy infrastructure facilities that are often destroyed under the action of a shock wave, which affects the stability of Ukraine’s energy systems. Modeling of a rocket explosion and the distribution of destructive pressures from shock waves in the conditions of dense urban development was carried out using the ANSYS AUTODYN software product. The calculated parameters of the boundary of the shock wave distribution 15 ms after detonation were determined, and the maximum pressure values of the blast wave at the manometric points of the transformer substation structures were determined in different conditions of their location: surface location and underground. The purpose of the work is to develop a methodology for assessing the stability of transformer substations as elements of the critical infrastructure of an urbanized space in war conditions during the bombardment of residential areas with rockets. This will make it possible to develop measures to preserve the structural integrity of the transformer substation building itself. Methods. Finite element analysis of the stress-strain state of a concrete structure based on the elastic-plastic model and the generalized Hook-Brown failure criterion. This made it possible to simulate the zone of intense destruction of the concrete structure of the underground and surface location.

The causes of welding impacts on metal are briefly considered: the formation of non-uniform microstructure, the formation of residual stresses and deformations, and the appearance of defects. The characteristics of the mutually reinforcing influence of welding impacts and corrosion are analyzed at three levels: micro-level processes, welding-joint level processes, and structural-level processes. The complex processes are presented simply and clearly in the form of original diagrams. Examples of “metal-corrosive environment” systems, which are the most common in the operation of welded structures, are given.

The classification of types of corrosion is given. The nature and schemes of processes of electrochemical, pitting, crevice, intergranular, knife, transcrystalline corrosion are considered. Mechanochemical corrosion mechanisms (including hydrogen embrittlement, stress corrosion, corrosion cracking, and corrosion fatigue) are examined in detail, as they are the main factors in the destruction of welded pipelines, pressure vessels, tanks, and ship structures. Corrosion mechanisms are illustrated with visual diagrams.

Standard methods for evaluating the corrosion resistance of welded joints are provided. Best practices of using the acoustic emission method to determine the type of corrosion, including the stage of corrosion cracking, when monitoring the technical condition of welded structures are shown. Possible directions for the use of artificial neural networks for the detection of the type of corrosion damage are shown.