Oblikovanje platforme za razporejanje opravil in spremljanje storitev

Šuligoj, Nino (author), Rožanc, Igor (mentor)

Diplomska naloga zajema oblikovanje porazdeljene platforme za razporejanje opravil in spremljanje storitev. V nalogi se soočimo z iskanjem ustrezne rešitve, ki bi trenuten sistem spremenila v porazdeljen sistem z visoko razpoložljivostjo in omogočenim spremljanjem storitev. S predstavljeno arhitekturo, delovanjem in zahtevami lahko določimo primeren sistem, ki omogoča vse, kar potrebujemo za nadgradnjo. Raziščemo obstoječe rešitve, jih primerjamo med seboj in izberemo najbolj ustrezno orodje, to je Kubernetes. To orodje namestimo in vanj umestimo naš sistem. Dodamo mu še funkcijo za spremljanje storitev. Po zaključenem praktičnem delu predstavimo še analizo uspešnosti in (uspešno) ovrednotimo rezultat izvedbe, v zaključku izpostavimo dosežke in napovemo možne dopolnitve sistema.

• Language: Slovenian
• Keywords: razporejevalnik opravil, Kubernetes, organizacija vsebnikov, porazdeljen sistem
• Work type: Bachelor thesis/paper
• Typology: 2.11 - Undergraduate Thesis
• Organization: FRI - Faculty of Computer and Information Science
• Year: 2023
• PID: 20.500.12556/RUL-150756
• COBISS.SI-ID: 169249283
• Publication date in RUL: 22.09.2023

Secondary language

• Language: English
• Title: Designing a task scheduling and service monitoring platform

Abstract

The thesis covers the design of a distributed platform for task scheduling and service monitoring. In the thesis, we search for a suitable solution to turn the existing system into a distributed system with high availability and service monitoring. With the presented architecture, operation and requirements, we define a suitable system that provides everything we need for its upgrading. We research existing solutions, compare them and choose the most appropriate tool, which is Kubernetes. We install this tool and put our system in it. Then, we upgrade it with functionality of monitoring services. After finishing the practically part, we present the performance analysis and successfuly evaluate the result of the implementation. In the conclusion, we emphasize the achievements and announce possible additional features.

• Keywords: taskscheduler, Kubernetes, docker orchestration tool, distributed system