DEPLOYMENT AND STATE MANAGEMENT TEMPLATE FOR SECOND-LEVEL DISTRIBUTED CACHES (L2) IN KUBERNETES CLUSTERS, IN ORDER TO ENSURE ZERO-DOWNTIME OPERATIONS

Romanovsky Petr Yurievich, Smetanin Viktor Maksimovich, Kapliyenko Alexandra Mikhailovna, Samborenko Evgeny Alekseevich, Rusakov Aleksei Mikhailovich

Within the framework of this scientific work, the current problem of ensuring high availability and fault tolerance of distributed second-level (L2) caching systems in the Kubernetes container orchestration environment is investigated. The author conducts an in-depth analysis of the limitations of standard controllers (Deployment), revealing their unsuitability for servicing stateful applications with strict requirements for maintaining the state and stability of network identification. The article proposes and practically implements a specialized deployment template based on the Redis Cluster and the StatefulSet controller, which integrates Headless service mechanisms to provide deterministic addressing and Persistent Volume Claims to save data on local node storage. The practical significance of the work is confirmed by a series of experiments in the K3s cluster, including simulation of emergency failures and scheduled updates. The test results (verified by data write and read operations after node restart) prove the possibility of providing zero-downtime operations and significantly reducing cache recovery time, which prevents degradation of the performance of underlying databases in highly loaded systems.

Keywords:Kubernetes, orchestration, Redis Cluster, L2 cache, StatefulSet, fault tolerance, zero-downtime, cloud computing, Persistent Volume, Headless Service, K3s, High Availability, High Availability, Stateful storage, data sharding