| Studiju veids |
maģistra akadēmiskās studijas |
| Studiju programmas nosaukums |
Biznesa informātika |
| Nosaukums |
Ietvars bezserveru skaitļošanas izmaksu novērtēšanai salīdzinājumā ar mikropakalpēm. |
| Nosaukums angļu valodā |
A Framework for Evaluating Serverless Computing as a Cost-Effective Alternative to Microservices. |
| Struktūrvienība |
33000 Datorzinātnes, informācijas tehnoloģijas un enerģētikas fakultāte |
| Darba vadītājs |
Gundars Alksnis |
| Recenzents |
Pēteris Rudzājs |
| Anotācija |
This study focuses on serverless computing as an inexpensive alternative of
microservices architecture based on a mixed-method approach utilizing Amazon Web
Services. The study develops a comprehensive evaluation framework integrating
quantitative performance benchmarking with qualitative expert insights. Controlled
experiments compare AWS Lambda and EC2 implementations across key metrics
including scalability, resource utilization, latency, and operational costs. Results
indicate serverless computing offers significant advantages for variable workloads
through automatic scaling and pay-per-invocation pricing, while microservices
provide predictable performance for sustained high-traffic scenarios. However,
challenges persist including cold start delays, vendor lock-in concerns, and
monitoring complexity. Expert interviews reveal pragmatic hybrid architectures
combining both paradigms according to workload characteristics. The framework
contributes
actionable decision-making guidance for cloud infrastructure
optimization, balancing technical performance against organizational capabilities and
strategic objectives. This research advances both academic understanding and
practical enterprise strategy in cloud-native architecture selection. |
| Atslēgas vārdi |
MĀKOŅSKAITĻOŠANAS IZMAKSU OPTIMIZĀCIJA, HIBRĪDARHITEKTŪRA, MIKROPAKALPOJUMU ARHITEKTŪRA, VEIKTSPĒJAS BENČMARKINGS, BEZSERVERU SKAITĻOŠANA |
| Atslēgas vārdi angļu valodā |
CLOUD COST OPTIMIZATION, HYBRID ARCHITECTURE, MICROSERVICES ARCHITECTURE, PERFORMANCE BENCHMARKING, SERVERLESS COMPUTING |
| Valoda |
lv |
| Gads |
2026 |
| Darba augšupielādes datums un laiks |
06.01.2026 10:15:14 |
