Clean Resources

Target

To avoid incurring unwanted costs after finishing the practice lab, we need to delete the created resources.

⚠️ WARNING: Deleting the Knowledge Base DOES NOT automatically delete the Vector Store (OpenSearch Serverless). You must manually delete the OpenSearch Serverless Collection as this is the costliest service in this Lab.

Implementation Steps

Step 1: Delete Knowledge Base

1. Access the Amazon Bedrock Console -> Knowledge bases.

2. Select the radio button next to your Knowledge Base name.

3. Click the Delete button.

4. A dialog box appears, enter the Knowledge Base name to confirm (or type delete).

5. Click Delete. This process takes 10-15 minutes to complete successfully, so please be patient.

Step 2: Delete Vector Store

1. Access the Amazon OpenSearch Service.
2. In the left menu, under Serverless, select Collections.
3. You will see a Collection named like bedrock-knowledge-base-....
4. Select the radio button next to that Collection name.
5. Click the Delete button.
6. Type confirm or the collection name to confirm deletion.
7. Click Delete.

Step 3: Delete Data on S3

1. Access the Amazon S3 service.
2. Select the bucket rag-workshop-demo.
3. Click the Empty button first.
   • Type permanently delete to confirm deleting all files inside.
4. After the bucket is empty, return to the Buckets list.
5. Select that bucket again and click the Delete button.
6. Enter the bucket name to confirm.

Completion

Congratulations on fully completing the Workshop “Building a RAG Application with Amazon Bedrock”. Your system has been cleaned up and is safe!