Migrating SQL Server to PostgreSQL at Enterprise Scale

Avinash Vallarapu (HexaCluster) and Michal Nosek (Percona) walk through a complex, large-scale SQL Server to PostgreSQL migration project. They’ll show how the right expertise, a genuine understanding of the key differences between the two engines, combined with the right tooling turns a hard, custom-scoped project into a repeatable one to migrate the entire SQL Server estate. Percona's deep expertise in open source databases is complemented by HexaCluster's HexaRocket, which delivers end-to-end migration: nearly 100% Schema and Database Objects Migration with Validation, Data Migration with Validation, CDC for live replication, and reverse replication for rollback.

The session shows how even complex migrations can be completed with limited effort, and how the same strategy scales to run many migrations at once. Application code changes are surfaced up front by DMAT, HexaCluster's Application and Database Migration Assessment Tool, so developers and administrators move faster and reach true success at scale.

We cover the mapping decisions that quietly change results if handled naively, how HexaRocket turns them into a standardized, config-driven pipeline that runs the same way across database number 1 and database number 200, and how to choose among Community PostgreSQL, Percona Distribution for PostgreSQL (well-tested extensions and TDE), Amazon RDS, and Aurora PostgreSQL.

If you own a SQL Server estate and PostgreSQL is on your roadmap, you will leave with a repeatable method for moving the whole fleet, not just one database – to maximize the benefits in short time.

Who should attend

• Database architects and DBAs running SQL Server (on-premise, Azure SQL, Azure SQL Managed Instance, AWS RDS and other platforms)
• Platform and infrastructure engineers planning to run PostgreSQL (self-managed or DBaaS)
• Application and backend engineers who own the T-SQL that has to keep working after cutover
• Engineering leaders scoping a SQL Server modernization or cost-reduction program, especially across many databases

What you will take away

• A data type and T-SQL mapping reference for what quietly causes data loss or behavior change if mapped naively.
• A repeatable, config-driven method for migrating tens or hundreds of databases the same way, in parallel.
• A working model for near-zero-downtime cutover with CDC, and a rehearsed rollback with reverse replication.
• A clear view of which PostgreSQL target (Community, Percona Distribution, RDS, Aurora) fits your security, cost, and operational needs.