> ## Documentation Index
> Fetch the complete documentation index at: https://help.nops.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete Idle RDS Instance or Move to Aurora Serverless

> Delete RDS instances that nobody uses, or move them to Aurora Serverless so they pause when idle.

An RDS instance is billed for every hour it runs, even with no connections. AWS Compute Optimizer flags instances that have been idle over its lookback period and recommends deleting them, or, if they are used only occasionally, moving them to Aurora Serverless with auto-pause. nOps imports these recommendations daily through AWS Cost Optimization Hub and shows the instance's monthly cost as the saving.

| | |
| - | - |
| **Cloud** | AWS |
| **Category** | Database |
| **Service** | Amazon RDS |
| **Action** | Delete |
| **Effort** | Low |
| **Restart needed** | No |
| **Can be rolled back** | Yes |
| **Analysis window** | Last 14 days |
| **Source** | AWS Compute Optimizer |

## Why it matters

Test databases, databases of retired applications and copies made for a one-off task often keep running for months after anyone last connected.

## How nOps finds it

This recommendation comes from AWS Compute Optimizer through Cost Optimization Hub, not from an nOps rule. AWS treats an instance as idle when it had no database connections over the lookback period (14 days by default). The saving is AWS's estimate of what the instance costs.

## Before you start

Confirm with the owner that the database is not needed, including for month-end jobs, disaster recovery or audits.

## How to implement

1. Check **CloudWatch → DatabaseConnections** for the instance over a longer period than the lookback.
2. Find the owner from tags and confirm the database can go.
3. If it must be kept but is rarely used, consider Aurora Serverless v2 with auto-pause instead (migrate with a snapshot restore).
4. Otherwise delete it with a final snapshot (`aws rds delete-db-instance --db-instance-identifier DB_ID --final-db-snapshot-identifier DB_ID-final`).
5. Remove it from Terraform or CloudFormation so it is not recreated.

## Risks and considerations

* Deleting is permanent unless you keep a final snapshot. The snapshot keeps costing storage until you delete it.
* A database used only once a month or once a quarter can look idle in a 14-day window.
* To roll back, restore the instance from the final snapshot.

## FAQs

> [accordion] Should I delete it or move it to Aurora Serverless?
>
> Delete it if nobody needs it. Move it to Aurora Serverless with auto-pause if it is needed now and then, so you pay mostly for storage while it is idle.

> [accordion] Why not just stop it?
>
> RDS starts a stopped instance again after 7 days, so stopping only helps for short breaks.

## Related

* [Stop Idle RDS Instance](https://help.nops.io/inform/supported-recommendations/stop-idle-rds-instance)
* [Delete Orphaned RDS Backups](https://help.nops.io/inform/supported-recommendations/delete-orphaned-rds-backups)
* [Delete Inactive DocumentDB Clusters](https://help.nops.io/inform/supported-recommendations/delete-inactive-documentdb-clusters)
* [Supported recommendations](https://help.nops.io/inform/supported-recommendations.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.