Blue-Green Deployment for Amazon RDS databases

 

Step-by-Step: Blue-Green Deployment for Amazon RDS

🔹 Use Case

• Safely upgrade schema/engine version.

• Avoid downtime and rollback risks.

• Test changes on an exact replica before promoting.

---

1. Create a Green (Clone) Environment

You have several options depending on your current setup:

Option A: Snapshot and Restore

1. Take a snapshot of your current RDS (Blue).

2. Restore a new DB instance from that snapshot (this becomes Green).

   aws rds create-db-instance \
     --db-instance-identifier green-db \
     --db-snapshot-identifier snapshot-of-blue
   

Option B: Read Replica (for supported engines like MySQL/PostgreSQL)

1. Create a read replica of Blue.

   aws rds create-db-instance-read-replica \
     --db-instance-identifier green-db \
     --source-db-instance-identifier blue-db
   

2. Promote it to a standalone DB instance (optional if testing writes).

   aws rds promote-read-replica \
     --db-instance-identifier green-db
   

---

2. Apply Changes to Green

• Apply schema changes, engine upgrades, or parameter group updates to the Green database.

• Use a test application or staging environment to connect to Green and validate functionality.

---

3. Validate Green

• Run integration and regression tests.

• Monitor performance, query plans, indexes, etc.

• Verify replication (if using a read replica).

• Ensure no application-breaking changes exist.

---

4. Redirect Application to Green

Once Green is validated:

Option A: Update the DB Endpoint

• Update the application/database connection string to point to Green’s endpoint.

• If you’re using Route 53, change the CNAME for the DB hostname.

Option B: Switchover using Route 53

• Use a CNAME (e.g., db.myapp.com) pointing to the current RDS instance.

• Change the CNAME to point to Green DB's endpoint.

---

5. Monitor Green

• Use CloudWatch for RDS metrics (CPU, IOPS, connections).

• Monitor application performance post-switch.

• Optionally keep Blue as backup for rollback.

---

6. Decommission Blue (Optional)

• After confirming stability, delete or snapshot the Blue instance.

  aws rds delete-db-instance \
    --db-instance-identifier blue-db \
    --skip-final-snapshot # or provide --final-db-snapshot-identifier
  

---

🧠 Best Practices

• Enable automated backups on both instances.

• Use parameter groups for environment isolation.

• Use CloudFormation or Terraform for reproducible infrastructure.