Zero-Downtime Database Migration at Scale: A Phased State-Machine Approach with Dual-Mode Consistency Validation
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Large-scale database migrations from on-premises infrastructure to cloud environments pose critical challenges around service continuity and data consistency. This paper presents a nine-phase state-machine framework for migrating over 500 GB of relational data to a cloud SQL database with zero service downtime. Our approach employs a dual-write and shadow-read strategy, where writes are propagated to both databases concurrently and reads are validated in parallel before any primary handover is committed. Data integrity is ensured through a multi-layered consistency validation strategy combining Sequence Watermark change tracking, application-level retries with persistent failure logging, and an asynchronous reconciliation mechanism for data not covered by direct replication. We further present a staged handover strategy that separates read and write transitions, reducing the risk of inconsistency during the cutover window. Experimental results show a latency increase bounded within 30 ms with no regression in API success rate.
Fields of Study & Subjects
Distributed systems and fault tolerance
Cloud Computing and Resource Management
Advanced Database Systems and Queries
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 27 March 2026
- Language
- EN
- Document type
- Preprint