Электронный каталог

Статья

Tsamtsurov, E.
Development of a Software Suite for Testing Server Hardware : Abstract / E.Tsamtsurov, N.Balashov, K.Lukyanov // Физика элементарных частиц и атомного ядра. Письма. – 2025. – Т. 22, № 5. – P. 980. – URL: https://www1.jinr.ru/Pepan_letters/panl_2025_5/15_Tsamtsurov_ann.pdf.

Testing of server equipment prior to its operation is crucial for ensuring reliable and smooth operation of systems at the Multifunctional Information and Computing Complex of the Joint Institute for Nuclear Research. The main purpose of testing is to identify hidden defects that may arise under critical loads on the equipment. There are various empirical methods described in production standards used to detect equipment failures. The paper presents an automated system for testing server equipment, including automation of system installation, launching tests, and collecting test logs. In the current implementation of the system, testing is carried out using the method of Highly Accelerated Stress Screening (HASS). A key part of the system is the monitoring subsystem required for collecting and analyzing temperature data from the tested components. Temperature metrics analysis during the testing phase allows for determining the duration of testing with a given accuracy. In addition to the monitoring tools such as Node Exporter, Prometheus, Prometheus Gateway and Grafana, the system uses Stress-ng to load the equipment with synthetic tests. All of these subsystems are freely distributed, and the proposed system can be easily implemented for similar testing in comparable infrastructures.



Спец.(статьи,препринты) = Ц 84 а2 - Многомашинные комплексы вычислительных средств. Вычислительные системы и сети. Параллельные вычисления. Квантовые компьютеры
ОИЯИ = ОИЯИ (JINR)2025