Example for Simulating AWS API errors

日本語

This is an example application that uses a custom AWS endpoint to simulate an AWS error occurrence on your local machine.

Abstract

This Python application uses the AWS SDK to connect to a DynamoDB table. And when building the DynamoDB client, a custom endpoint is specified as the connection destination to simulate an error.

Specifically, specify the endpoint with the following code.

dynamodb = boto3.resource('dynamodb', endpoint_url = endpoint_url)

If endpoint_url is None, SDK connect AWS endpoint.

The error message returned from this sample endpoint is different from the actual one.

If you want to simulate on AWS...

If you want to simulate this on AWS, you can use API Gateway's Mock response for your custom endpoint.

Pre-Requirements

• docker (using Nginx for custom endpoint server)
• DynamoDB Table on AWS :
  For use case of successful access. If need not, you may not to create it and skip "3-a" too
  • Table name is "test" or specified environment value TABLE_NAME
  • Partition key column name is pk, type is "String"
  • Create an item with the string "1" as the partition key (pk).
• AWS Credential that is granted get item for the table above.

How to run example

1. Install boto3

if needed, intall boto3

pip install

2. Run custom endpoint server

Run docker container.

docker run \
  -it \
  --rm \
  --name ddb-endpoint \
  -p 8080:80 \
  -v ${PWD}/endpoint/400.json:/usr/share/nginx/html/400.json \
  -v ${PWD}/endpoint/500.json:/usr/share/nginx/html/500.json \
  -v ${PWD}/endpoint/nginx.conf:/etc/nginx/nginx.conf \
  nginx:latest

or run ./run-endpoint.sh

Nginx container is ready and...
If access http://localhost:8080/4xx/, you can get 400 error.
If access http://localhost:8080/5xc/, you can get 500 error.

3-a. Run app.py with successful access

In this case, using an endpoint on AWS.

TABLE_NAME=test \
python app.py

You can see like followint output:

2022-01-28T00:53:10+0900 __main__:44 [INFO] <module> : Endpoint URL : None
2022-01-28T00:53:10+0900 __main__.App:18 [DEBUG] __init__ : Constructor invoked.
2022-01-28T00:53:11+0900 __main__.App:31 [INFO] test_get_item : {'pk': '1', 'value': '...'}

3-b. Run app.py with 400 error.

In this case, using a custom endpoint.

TABLE_NAME=test \
ENDPOINT_URL=http://localhost:8080/4xx/ \
python app.py

You can see like following output:

2022-01-28T00:58:37+0900 __main__:44 [INFO] <module> : Endpoint URL : http://localhost:8080/4xx/
2022-01-28T00:58:37+0900 __main__.App:18 [DEBUG] __init__ : Constructor invoked.
2022-01-28T00:58:38+0900 __main__.App:27 [ERROR] test_get_item : 400 Error
2022-01-28T00:58:38+0900 __main__.App:28 [ERROR] test_get_item : Error Code: 400
2022-01-28T00:58:38+0900 __main__.App:29 [ERROR] test_get_item : An error occurred (400) when calling the GetItem operation: {
    "Error": {
        "Code": "SomeServiceException",
        "Message": "Details/context around the exception or error"
    },
    "ResponseMetadata": {
        "RequestId": "1234567890ABCDEF",
        "HostId": "host ID data will appear here as a hash",
        "HTTPStatusCode": 400,
        "HTTPHeaders": {"header metadata key/values will appear here"},
        "RetryAttempts": 1
    }
}

3-c. Run app.py with 500 error.

In this case, using a custom endpoint.

TABLE_NAME=test \
ENDPOINT_URL=http://localhost:8080/5xx/ \
python app.py

You can see like following output: It will take some time to complete as the SDK will retry due to a 500 error. You can see the retries in the access log of the nginx container.

2022-01-28T00:59:57+0900 __main__:44 [INFO] <module> : Endpoint URL : http://localhost:8080/5xx/
2022-01-28T00:59:57+0900 __main__.App:18 [DEBUG] __init__ : Constructor invoked.
2022-01-28T01:00:22+0900 __main__.App:27 [ERROR] test_get_item : 500 Error
2022-01-28T01:00:22+0900 __main__.App:28 [ERROR] test_get_item : Error Code: 500
2022-01-28T01:00:22+0900 __main__.App:29 [ERROR] test_get_item : An error occurred (500) when calling the GetItem operation (reached max retries: 9):

License

MIT