
МагиÑтерÑÐºÐ°Ñ Ð²Ñ‹Ð¿ÑƒÑÐºÐ½Ð°Ñ Ñ€Ð°Ð±Ð¾Ñ‚Ð° поÑвÑщена повышению производительноÑти центра обработки данных (ЦОД) путем мониторинга метрик оценки ÑффективноÑти ÑкÑплуатации реÑурÑов. Ð’ работе были выбраны инÑтрументы Ð´Ð»Ñ Ñбора, обработки и раÑчета метрик на оÑнове проведенного анализа научных публикаций ÑкÑпертов предметной облаÑти, определены оÑновные метрики Ð´Ð»Ñ Ð¾Ñ†ÐµÐ½ÐºÐ¸ ÑффективноÑти иÑÐ¿Ð¾Ð»ÑŒÐ·Ð¾Ð²Ð°Ð½Ð¸Ñ Ñ€ÐµÑурÑов ЦОД, проведена оценка полученных результатов. Ð’ работе предложены методы Ð¿Ð¾Ð²Ñ‹ÑˆÐµÐ½Ð¸Ñ Ð¿Ñ€Ð¾Ð¸Ð·Ð²Ð¾Ð´Ð¸Ñ‚ÐµÐ»ÑŒÐ½Ð¾Ñти ЦОД, Ð²ÐºÐ»ÑŽÑ‡Ð°Ñ Ð¸Ñпользование метрик оценки ÑффективноÑти, автоматизацию процеÑÑов Ñбора и анализа показателей и повышение производительноÑти раÑчета метрик Ñ Ð¿Ð¾Ð¼Ð¾Ñ‰ÑŒÑŽ парадигмы Map Reduce и кÑширование промежуточных результатов. Также предложено иÑпользование индекÑа B-Tree в релÑционной базе данных Ð´Ð»Ñ Ð¾Ð¿Ñ‚Ð¸Ð¼Ð¸Ð·Ð°Ñ†Ð¸Ð¸ доÑтупа к раÑÑчитанным метрикам. Предложенные методы были реализованы в программном ÑредÑтве, которое позволÑет отÑлеживать ключевые метрики и принимать Ñ€ÐµÑˆÐµÐ½Ð¸Ñ Ð¿Ð¾ улучшению инфраÑтруктуры и оптимизацию затрат на реÑурÑÑ‹ ЦОД. Результаты реализации прототипа и теÑÑ‚Ð¸Ñ€Ð¾Ð²Ð°Ð½Ð¸Ñ Ð¿Ñ€Ð¾Ð´ÐµÐ¼Ð¾Ð½Ñтрировали ÑффективноÑть предложенных методов.
The masters thesis is dedicated to improving the performance of a data center by monitoring resource utilization metrics. The study analyzed scientific publications based on data from open sources of the data center. Metrics were used to determine the consumption of key resources such as CPU, RAM, and power consumption, and results were compared. Based on this, tools were selected for collecting, processing, and calculating metrics. The study proposes methods for improving data center performance, including the use of resource utilization metrics, process automation, and improved computing performance through the Map Reduce paradigm, caching intermediate results with Redis. The study also proposes using B-Tree index in the PostgreSQL relational database to optimize access to calculated metrics. The proposed methods were implemented in software that allows tracking key metrics and making decisions to improve infrastructure and optimize data center resource costs. The results of the prototype implementation and testing demonstrated the effectiveness of the proposed methods, which was confirmed by comparing metric- based indicators with original resource indicators.
monitoring, мониÑоÑинг, metric, меÑÑики, опÑимизаÑÐ¸Ñ ÑабоÑÑ ÑенÑÑа обÑабоÑки даннÑÑ, optimization of datacenter operations, monitoring system, ÑиÑÑема мониÑоÑинга
monitoring, мониÑоÑинг, metric, меÑÑики, опÑимизаÑÐ¸Ñ ÑабоÑÑ ÑенÑÑа обÑабоÑки даннÑÑ, optimization of datacenter operations, monitoring system, ÑиÑÑема мониÑоÑинга
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 0 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Average | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Average | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Average |
