Microsoft SQL Server runs a large share of enterprise applications. Moving it to the cloud can reduce infrastructure management, improve resilience and, with the right approach, lower licensing costs. This guide compares the main ways to move SQL Server to Google Cloud and outlines a practical migration plan.

Option 1: Cloud SQL for SQL Server

Cloud SQL is Google Cloud’s fully managed database service. It handles patching, backups, high availability and replication, so your team manages databases instead of servers.

• Best for: most standard SQL Server workloads that do not need operating-system-level access.
• Consider: some advanced features and configurations differ from self-managed SQL Server, so check compatibility early.

Option 2: SQL Server on Compute Engine

Running SQL Server on virtual machines gives full control over configuration, features and versions, similar to on-premises.

• Best for: workloads that need specific features, third-party agents or custom configurations.
• Consider: your team remains responsible for patching, backups and high availability.

Option 3: Modernize to PostgreSQL

Migrating from SQL Server to Cloud SQL for PostgreSQL or AlloyDB removes SQL Server licensing costs and moves you to an open-source engine. It requires converting schemas, T-SQL code and application queries.

• Best for: applications you are actively developing, where long-term licensing savings justify conversion work.
• Consider: plan for code conversion, testing and performance tuning.

A practical migration plan

1. Assess

Inventory databases, sizes, versions, features in use, dependencies, and performance and availability requirements. Review licensing, including whether existing licenses can be used in the cloud.

2. Choose a target per database

Not every database needs the same approach. Many organizations move most databases to Cloud SQL and modernize selected applications to PostgreSQL over time.

3. Prepare the foundation

Set up networking, identity, security and monitoring in a well-designed GCP landing zone, including hybrid connectivity to on-premises systems during the transition.

4. Migrate data

Use backup and restore, replication or Database Migration Service, depending on the target and the downtime you can accept. Continuous replication keeps cutover windows short.

5. Test thoroughly

Validate data, run application and performance tests, and rehearse the cutover and rollback steps.

6. Cut over and optimize

After go-live, rightsize instances, tune queries and indexes, and review costs as part of your Google Cloud FinOps routine.

Related reading

Moving Oracle as well? See our Oracle to Google Cloud migration guide. For the bigger picture, explore our Google Cloud migration services.

Plan your SQL Server migration

Auriqo helps enterprises in India and the USA assess, plan and execute SQL Server migrations to Google Cloud. Book a migration assessment.