Polyglot Persistence in Distributed Databases for Point in Time Recovery and Failure Handling in MySQL Replicated Environment

Polyglot Persistence in Distributed Databases for Point in Time Recovery and Failure Handling in MySQL Replicated Environment

연구 분야: Databases

학회: 2023 7th International Conference on Intelligent Computing and Control Systems (ICICCS)

Polyglot Persistence refers to using different databases in the same architecture to achieve availability, scalability, consistency and fault tolerance. This research study proposes a novel approach to deal with failures in distributed databases. Most businesses use distributed databases and replication techniques to achieve high availability and scalability as their reliance on digital data grows. Even with backup solutions such as snapshots and binary logs, there are challenges such as data loss due to disc corruption or network issues. This research study proposes a polyglot architecture that includes a NoSQL rapid access key-value storage and cache to improve system resilience and reduce reliance on primary server databases. The proposed method saves timestamps and query logs to extra storage, allowing for query reconstruction and data restoration in the event of disc corruption or server failure.