Microsoft SQL Server powers mission-critical enterprise workloads across countless organizations, but this power comes with expensive enterprise licensing. Companies are increasingly exploring ways to manage these costs, from right-sizing existing workloads to considering open-source alternatives and complete SQL Server migration.
In cloud computing, where costs directly correlate with utilization, the first step should always be ensuring your workloads are properly sized and appropriately licensed. While this isn’t a new concept, many organizations have already taken this step using tools like Optimize 24×7 to understand and optimize their workloads.
Why Modernize?
Even with properly sized SQL Server workloads, costs can remain prohibitive. This reality has sparked new conversations around modernization, with open-source databases emerging as a compelling cost-saving approach.
The modernization landscape has evolved beyond relational databases. Non-relational databases now serve as better fits for specific components of monolithic workloads. For this discussion, we’ll focus on the relational aspect, specifically PostgreSQL.
PostgreSQL has become a leading candidate in replacement discussions due to its open-source licensing and rapidly expanding extension libraries. From its origins at UC Berkeley in the 1980s, evolving from Ingres to today’s version 17, PostgreSQL now ranks as the 4th most popular relational database. According to db-engines.com, it’s experiencing significant market share growth while competitors face declines.
Self-Managed vs. Managed Databases
When running PostgreSQL databases within AWS, you have several deployment options:
Self-Managed Databases
You start with a base Linux image and handle all installation and configuration. This approach places full ownership on you for backups, patching, upgrades, high availability, disaster recovery, performance tuning, and troubleshooting.
Managed Databases
AWS handles the heavy lifting of database administration, including OS and database installation, patching, upgrades, configuration, data replication, and point-in-time recovery. This allows your DBAs to focus on application-impactful work like performance tuning and implementing new functionality.
Managed offerings include Amazon RDS, which closely mimics self-managed architecture while providing the administrative benefits described above. Amazon Aurora goes further by enhancing PostgreSQL through separated storage and compute tiers, adding native cloud capabilities that improve scalability, performance, and multi-region functionality. This elevates Aurora PostgreSQL to enterprise-class status while simplifying backup complexity, reader node management, and storage scaling.
Where to Begin?
No two database workloads are identical when evaluating open-source alternatives. Each varies in data volume, schema complexity, and SQL Server functionality usage. Understanding these differences enables you to calculate overall conversion costs (both development hours and runtime expenses) and prioritize the most cost-effective conversions.
Starting with low-risk, low-complexity workloads provides an easier migration path than tackling complex systems immediately. This approach helps you build skills and refine processes before addressing more challenging databases.
While AWS and third-party vendors provide database conversion tools, today’s workload complexities demand deeper understanding of both SQL Server and PostgreSQL capabilities. Success requires accurately assessing workloads, designing cost-effective architectures, ensuring proper functionality mapping to PostgreSQL, leveraging modern PostgreSQL capabilities, and avoiding the migration of technical debt.
Next Steps
Evolve Cloud Services maintains a team of database specialists available through the no-cost AWS Modernization Viability Assessment (MVA). We can evaluate your workloads, calculate conversion costs, and guide your modernization journey.
For questions about the modernization process or transitioning to open-source database solutions, please Contact Us.
Authored by Phil Ekins, Director of Database Services at Evolve Cloud Services