Навчально-науковий інститут атомної та теплової енергетики (НН ІАТЕ)

Постійне посилання на фонд

https://ela.kpi.ua/handle/123456789/62

Переглянути

Перегляд Навчально-науковий інститут атомної та теплової енергетики (НН ІАТЕ) за Ключові слова "004.23:004.272.44"

Зараз показуємо 1 - 1 з 1
  • Ескіз
    ДокументВідкритий доступ
    Event-oriented architecture in the system for processing large data streams
    (Державний університет телекомунікацій, 2023) KuzmInykh, V. O.; Otrokh, S. I.; Xu, B.; Taranenko, R. A.; Kublii, L. I.
    The article discusses approaches to the implementation of the microservice architecture of the system for processing large data flows when collecting information on the main event-oriented approach in the organization of managing the sequence of use of individual microservices, which provide access to information sources and the extraction of relevant data corresponding to the user's request. this is important when processing large data flows from heterogeneous infor-mation sources, especially when the task is to minimize the total time of collection and processing of large data flows. this, in turn, poses the task of minimizing the number of requests to information sources to obtain a sufficient number of units relevant to the data search request. the creation of effective systems for processing big data requires constant development of approaches to the architecture of building software applications. an approach to the construction of the software system architecture is proposed, which makes it possible to manage the selection of microservices adaptively, in accordance with the events that occur during information collection, and thus form the choice of information sources based on the evaluation of the effectiveness of obtaining relevant information from them. the event-oriented microservice architecture makes it possible to adapt the operation of the software system to the loads on individual microservices and to increase the efficiency of their work by correcting calls to information sources based on the analysis of relevant events that occur during the collection and initial processing of the received data. Based on the analysis of the current state of information collection, new instances of containers can be created to extract information from the most effective sources and, thus, improve the efficiency of the system as a whole. the use of event-oriented microservice architecture can be especially effective in the development of various information and analytical systems that have the need, according to user requests, to refer to various sources of information based on their relevance, and to process large data streams when collecting information.