-
Notifications
You must be signed in to change notification settings - Fork 163
Open
Description
Problem Statement
Currently, PeerDB syncs focus on data replication but don't handle index migration from source to target databases. This creates manual overhead for users who need to recreate indexes on the target system.
Proposed Solution
Add automatic index migration support during sync operations:
Key Features:
- Detect and capture index definitions from source database
- Replicate indexes to target database during initial sync
- Support common index types (B-tree, Hash, GiST, GIN, etc.)
- Handle index creation timing (post-data load for performance)
- Provide configuration options to include/exclude specific indexes
Benefits:
- Reduces manual setup time
- Ensures query performance parity between source and target
- Minimizes risk of missing critical indexes
- Improves overall sync completeness
Implementation Considerations
- Index creation should occur after initial data load to optimize performance
- Support for partial indexes and expression-based indexes
- Handle index naming conflicts
- Provide dry-run mode to preview index migration plan
- Add metrics for index creation progress
Use Cases
- Production to staging environment replication
- Database migration projects
- Read replica setup with full schema parity
- Disaster recovery scenarios
Would appreciate feedback on this proposal and any additional requirements to consider.
Metadata
Metadata
Assignees
Labels
No labels