Google launches 2nd generation of its Cloud SQL service, with up to 10TB per instance
Context & Ripple Effects
This closes out a year in which Google rebuilt its enterprise database story from scratch: the Cloud Bigtable launch in May brought a fast NoSQL service aimed at corporate workloads, and today's second-generation Cloud SQL answers the obvious follow-up — what about plain relational data at scale.
The 10TB-per-instance ceiling is the point of comparison: until now, customers with large transactional databases had little reason to pick Google over the incumbent managed-relational offerings, so gen 2 is less a feature drop than a removal of the excuse not to evaluate GCP.
First-order effects
- Enterprises running large MySQL-class workloads can now keep them fully managed on Google Cloud instead of sharding across smaller instances or self-managing capacity.
Second-order effects
- AWS and Microsoft must match the scale tier or cede large-database migrations, and Google's pairing of Cloud SQL with Bigtable lets it pitch a full relational-plus-NoSQL stack rather than point solutions.
Third-order effects
- If the cadence holds, managed databases become Google's enterprise beachhead — a pattern the corpus confirms with the trio reaching general availability in 2016 (Cloud SQL, Bigtable, and Datastore exiting beta) and later broadening to Microsoft SQL Server support, showing the platform absorbing rival ecosystems.
The trend: Cloud providers are turning managed databases from commodity add-ons into the primary wedge for winning enterprise workloads, with capacity ceilings and engine breadth as the competitive currency.