Enterprise Data-Estate Migration to Cloud-Native Converged Analytical Platforms: Architectural Patterns, Failure Modes, and Governance Implications

Abstract

When enterprises migrate applications to the cloud, they can use unit testing and other tests of deployed versions to ensure accuracy. When migrating other enterprise assets, including analytics datasets built up over at least a decade of mergers and acquisitions and other growth, the unit of migration is the row, and accuracy considerations include ensuring that rows and other data are not lost or modified during the migration. This article discusses challenges specific to migrating analytics data assets: architectural approaches and other issues related to migrating analytics data to modern, integrated cloud platforms that provide both data warehousing and advanced analytics capabilities on a common platform using open data formats (Armbrust et al., 2021). The article draws on published information about migrations of large commercial datasets in the highly regulated financial-services industry to describe and analyze three architectural approaches to migrating analytics data assets: domain-based phased migrations, dual writes and other validation techniques, and approaches to handling data schema changes during migration. It cites twenty-one sources, including publications on data migration more generally (Fahmideh Gholami et al., 2016), lakehouse platforms (Armbrust et al., 2021), (Jain et al., 2023), microservices architectures (Ayas et al., 2023), and data governance (Volz et al., 2025). Four common problems with analytics data migrations are discussed: data quality issues that arise during or after migration but not before (Azeroual et al., 2021), (Foidl et al., 2024); tradeoffs related to data replication in geographically distributed environments that are not fully understood (Ferreira et al., 2024), (Kadam et al., 2026); new dependencies on specific cloud providers and data formats developed during migration (Opara-Martins et al., 2016), (Alhosban et al., 2024); and governance challenges, including unclear ownership, that can increase during a migration (Volz et al., 2025). A table summarizes these and other challenges and responses. A case study describes another enterprise migration in detail, including its architecture and rollback plans. The article discusses three debates in the literature: the relative importance of technical and governance issues to successful migration; the value of automated tools to manage data schema changes compared with traditional data quality efforts; and the extent to which open data table formats reduce risks of provider lock-in. It recommends that technical and governance decisions be made together to support successful migration.