
- 15 Nov, 2026
- Database Migration
Migrating from Greenplum to Apache Cloudberry requires more than transferring table data. A successful migration also considers database objects, distribution policies, permissions, external integrations and application compatibility.
Assess the source environment
Before starting, inventory:
- Databases, schemas and table sizes
- Partitioned tables and partition boundaries
- Distribution keys and storage options
- Views, functions, procedures and triggers
- Roles, grants and ownership
- External tables and data integrations
- Application connections and dependencies
This assessment helps identify migration risks before the target environment is populated.
Plan the migration approach
Choose migration methods based on data volume, downtime requirements, object compatibility and available tooling.
Validate the chosen approach with representative tables before migrating the full environment. Large tables, partitioned tables and objects with dependencies may need additional attention.
Validate data and database objects
After data transfer, compare source and target results. Useful checks include:
- Table and partition row counts
- Data type and schema comparisons
- Distribution and storage configuration
- View and function validity
- Object ownership and permissions
- Application read and write operations
Row counts alone do not prove that a migration is complete. Include application-level validation and checks for important business data.
Review application access
Create appropriate application roles and grant only the required privileges. Review access to existing objects as well as default privileges for future objects.
External tables and data integrations should be tested separately because they can depend on operating system services, network access and external configuration.
Plan post-migration maintenance
Review statistics, query plans, resource settings, monitoring and backup procedures after migration. Performance characteristics may change when workloads run on a different platform or infrastructure configuration.
Conclusion
A structured migration combines inventory, controlled data movement, object validation, access testing and post-migration tuning. Treating these as separate workstreams makes the migration easier to verify and operate.




