Data protection becomes critical as your deployment moves beyond development into production. Keep a backup and recovery strategy in mind even before you move into production.

Backup strategies

Weaviate supports automated backups to local or cloud storage.

If using Serverless Cloud → Automated backups included

• Daily backups with point-in-time recovery
• No configuration required, managed by Weaviate

If using Enterprise Cloud → Configurable managed backups

• Custom backup frequency and retention policies
• Managed by Weaviate with your storage preferences

If self-hosting → Manual backup configuration required

• S3, GCS, or Azure blob storage integration
• You manage backup schedules, monitoring, and retention
• Example: On-premises deployment with existing backup infrastructure

Recovery planning

Weaviate's HNSW snapshotting feature significantly reduces the recovery time required, as the HNSW index does not need to be rebuilt from scratch.

However, keep recovery time and resources required in mind as you develop with Weaviate.

Key considerations

1. Recovery Time Objective (RTO) - How quickly do you need to be back online?
2. Recovery Point Objective (RPO) - How much data loss is acceptable?
3. Testing - Regularly test your backup and restore procedures

Operational considerations

• Backup storage costs - Vector databases can generate large backups
• Network bandwidth - Large datasets require significant transfer capacity
• Cross-region redundancy - Consider backup storage in different regions
• Retention policies - Balance compliance needs with storage costs

Learn more

• Configuration: Backups
• Blog: Hardening your Weaviate OSS installation

What's next?

Let's explore real-world industry patterns that combine all these architectural decisions into proven solutions.