Бенюх, ЛадаГлибовець, Андрій2021-12-142021-12-142021-12-10Бенюх Л.І. Розроблення архітектури системи проведення високонавантажувального тестування / Бенюх Л. І., Глибовець А. М. // Наукові записки НаУКМА. Комп'ютерні науки. - 2021. - Т. 4. - С. 88-92. - https://doi.org/10.18523/2617-3808.2021.4.88-922617-38082617-7323https://doi.org/10.18523/2617-3808.2021.4.88-92https://ekmair.ukma.edu.ua/handle/123456789/21895У роботі проаналізовано основний інструментарій здійснення навантажувального тестування і тестування продуктивності, наведено приклади масштабування таких тестів і централізованої звітності метрик. Описано розроблену методологію та основні принципи побудови сучасної архітектури для ефективної реалізації підсистеми навантажувального тестування у безперервному постачанні коду.Testing system performance and its importance at the same time is difficult to overestimate or underestimate. It would be much more correct to talk about the timeliness of this activity. Virtually any digital system built on modern approaches and technologies can work without any critical problems with its own performance. At the same time, for any system, especially when it becomes popular, it is very likely that there will be a time when it will not be able to cope with the ever-increasing load and become unstable. However, most companies that develop and maintain their own digital solutions – from websites to any other digital systems – often focus primarily on the functionality of the system and its compliance, rather than on the performance of the system as a whole. Such intentions are quite natural, because the system must properly perform the functions expected of it. When companies start to face performance problems, they try not to optimize the software as soon as possible, but to add more capacity – vertical and horizontal scaling. This strategy works, but it has limitations. After all, the addition of additional resources cannot be endless and sooner or later rests either on the architecture of the system, or in the capabilities of the company itself, and so on. Therefore it is recommended to carry out stress testing in advance, plan time and resources to have enough time to correct errors, and generally understand the boundaries of the system. At the same time, in order to organize full-fledged stress testing, trained specialists, tools and infrastructure are needed, especially when we are talking about heavy workload. As part of this work, an analysis of various tools for the implementation of stress testing and performance testing, scaling of such tests and centralized reporting of metrics. As a result, approaches and principles for the construction of a modern architecture for the implementation of the load testing subsystem in the continuous supply of code were proposed.ukнавантажувальне тестуваннятестування продуктивностіKubernetesмасштабування тестіввеликонавантажувальні системицентралізована звітність результатів тестуванняCI/CDбезперервне постачання кодустаттяload testingperformance testingKubernetesscaling testshigh-load systemcentralized reporting of test resultsCI / CDcontinuous code deliveryarticleArticle