High availability model of a web service from a system administrator’s perspective

Based on cloud providers’ reports on service outages, it has become clear that how a web service is deployed is of great importance. Clearly, using one service supplier is insufficient because it introduces single points of failure. In this paper, a novel high-availability multi-cloud model intended for a web service is proposed, which is free from such shortcomings yet preserves convenient assets of computing clouds. The methodology used to improve web service availability should involve several cloud suppliers and devise management techniques that control access to them. This is achieved by means of the server availability tracking algorithm, which controls client apps’ access to the service. Moreover, typical benefits and problems involved in choosing IT infrastructure for a web service are elaborated. State-of-the-art cloud computing models, such as IaaS, PaaS, SaaS, BPaaS, and INaaS, are outlined. Operating systems statistics used for web services are included. Open-source monitoring software solutions are gathered, which help administrators to monitor and govern web servers.

---

Opracowanie rekordu ze środków MNiSW, umowa nr SONP/SP/546092/2022 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2024).