Mitigating Your Data Bloat with Partitioning & Data Virtualization ~ Andy Yun

Part 4 of a 4-part series about:
Hidden Pathways to Achieving Peak SQL Server Performance

Mitigating Your Data Bloat with Partitioning & Data Virtualization

Are you managing a VLDB where cold data must be retained, but it’s dragging down maintenance, backups, and query performance? What if you could shrink your database footprint while still letting users query that data exactly as they always have? In this session, we’ll start by reviewing the classic approach using partitioning and filegroup strategies. Then we’ll dive deep into a modern, more flexible solution: Data Virtualization. You’ll learn how external tables/CETAS can enable seamless access to archived data stored in object storage, with zero changes to your users’ T-SQL code. This isn’t just about saving space — it’s about redefining what “archiving” can mean in today’s SQL Server environments. You’ll leave with practical techniques — both classic and modern — for reducing database bloat, improving manageability, lowering storage needs, and preserving fast, transparent access to historical data.

Session Goals: Will explore two solutions to managing build-up of old, stale data in very large databases (VLDBs). First 1/3 will explore the topic of partitioning and filegroup trickery, in conjunction with storage tiering. The remaining 2/3rd's will explore Data Virtualization introduced in SQL Server 2022. Session will focus on "old vs new" and showcase the practical benefits of Data Virtualization.
Session Prerequisites: Familiarity with partitioning will be helpful but not required.