Redshift Snapshots & DR | Overview.

Redshift Snapshots & DR  - Overview.
Scope:

• Intro,
• Snapshots in Amazon Redshift Types,
• How Snapshots Work,
• Snapshot Copying (To another region or account for DR),
• Restoring from Snapshots,
• Disaster Recovery (DR) Strategies for different Levels for Redshift,
• Security & Compliance,
• Best Practices.

Intro:

• Amazon Redshift snapshots play a crucial role in data backup and disaster recovery (DR) by providing point-in-time copies (replication) of a data warehouse. 
• This functionality is available for both provisioned and serverless configurations.

1. Snapshots in Amazon Redshift Types

• twtech Redshift Snapshots are backups of its Redshift cluster data, stored in Amazon S3 or data warehouse. 
• They capture the state of twtech cluster at a point in time.

Types

1. Automated Snapshots
• Managed by Redshift.
• Created on a schedule (default: every 8 hours, or after 5 GB of changes).
• Retention period configurable (1–35 days).
• Stored in same AWS region as cluster.
2. Manual Snapshots
• User-initiated.
• Retained until explicitly deleted.
• Useful for before upgrades, schema changes, or DR purposes.

 2. How Snapshots Work

• Redshift uses incremental snapshots:
• First snapshot → full copy.
• Subsequent snapshots → only changed blocks since last snapshot.
• Snapshots are stored in S3
• Snapshots are stored in S3, but fully managed (twtech doesn’t see the S3 bucket).
• No cluster downtime → snapshots run in background.

3. Snapshot Copying (To another region or account for DR):

• Cross-Region Snapshot Copy:
• Automatically copy automated/manual snapshots to a different region.
• Useful for geo-redundancy and DR scenarios.
• Cross-Account Snapshot Sharing:
• Share manual snapshots with other AWS accounts.
• Enables multi-tenant setups or DR in a secondary account.

 4. Restoring from Snapshots

• A new cluster can be restored from a snapshot:
• Creates a fresh cluster with the same schema & data.
• Can scale up/down to a different node type during restore.
• twtech can rename clusters or switch endpoints for application continuity.

 5. Disaster Recovery (DR) Strategies for different Levels for Redshift:

1. Basic DR (Single Region)
• Automated & manual snapshots in same region.
• Restore in case of cluster failure.
2. Cross-Region DR
• Use Cross-Region Snapshot Copy.
• Restore in another AWS region if primary region is down.
3. Cross-Account DR
• Share snapshots with a secondary AWS account.
• Protects against accidental deletion or account compromise.
4.  Amazon Redshift RA3 nodes for data warehousing (RA3) + Managed Storage Advantage
• Since RA3 uses Amazon S3 as managed storage, recovery can be faster and more cost-efficient.

 6. Security & Compliance

• Snapshots are encrypted:
• By default with AWS-managed keys.
• Can use KMS CMKs for custom encryption.
• Snapshot encryption status:
• Restored clusters must match the encryption setting.
• Audit & Compliance:
• Snapshots are durable (S3’s 11 9s durability).
• Cross-region copies support compliance with DR regulations.

 7. Best Practices

• Enable Automated Snapshots + Cross-Region Snapshot Copy.
• Take Manual Snapshots before schema or workload changes.
• Set retention policies to balance cost vs recovery needs.
• Test DR drills: restore in secondary region & validate queries.
• Consider cross-account snapshots for added security.