"KeyError: 'Records'" in AWS S3 - Lambda trigger

Question

I have the following lambda function code for simply printing out the Author and metadata of an uploaded event of an S3 bucket:

from __future__ import print_function
import json
import urllib
import boto3

print('Loading function')

s3 = boto3.client('s3')


def lambda_handler(event, context):

    #print("Received event: " + json.dumps(event, indent=2))
    # bucket = event['Records'][0]['s3']['bucket']['name']

    for record in event['Records']:
        bucket = record[0]['s3']['bucket']['name']
        key = record[0]['s3']['object']['key']
        response = s3.head_object(Bucket=bucket, Key=key)

        logger.info('Response: {}'.format(response))

        print("Author : " + response['Metadata']['author'])
        print("Description : " + response['Metadata']['description'])

However, I am getting the following error while testing:

{
  "stackTrace": [
    [
      "/var/task/lambda_function.py",
      17,
      "lambda_handler",
      "for record in event['Records']:"
    ]
  ],
  "errorType": "KeyError",
  "errorMessage": "'Records'"
}

Am I doing anything wrong while accessing the bucket name and key name of the S3 object? If not, then what am I doing wrong here?

Answer 1

I had this issue yesterday and the problem is when you run a test there isn't actually any event (def lambda_handler(event, context)) for lambda to process. The event document is normally generated when an actual event occurs which is a json doc, your python code parses this document and then retrieves the keys. So the solution is for you to create an actual event by dropping a file in s3 so long as you have created event notification then this would get picked up

Answer 2

bit late to the party. But here is my first post!

EXPLANATION:

When you test in lambda panel -> def lambda_handler(event, context) <- event is injected directly.

However in AWS API its neccessary to add Mapping Template or otherwise -> event <- is empty, thus resulting in quizzing:

"errorType": "KeyError", "errorMessage": "'Records'"

this is null pointer. Records doesnt exist, since -> event <- doesnt exist.

SOLUTION:

You need to configure Integration Request inside AWS API. Click on Body Mapping Templates. Then add Mapping Template Set content type to application/json Then edit generated mapping template:

{
  "body" : $input.json('$'),
  "headers": {
    #foreach($header in $input.params().header.keySet())
    "$header": "$util.escapeJavaScript($input.params().header.get($header))" #if($foreach.hasNext),#end

    #end
  },
  "method": "$context.httpMethod",
  "params": {
    #foreach($param in $input.params().path.keySet())
    "$param": "$util.escapeJavaScript($input.params().path.get($param))" #if($foreach.hasNext),#end

    #end
  },
  "query": {
    #foreach($queryParam in $input.params().querystring.keySet())
    "$queryParam": "$util.escapeJavaScript($input.params().querystring.get($queryParam))" #if($foreach.hasNext),#end

    #end
  }  
}

And edit Lambda function:

replace:

for record in event['Records']:

with:

for record in event['query']['Records']

don't know whether stack will ping you with this answer - so i call you @Dawny33 @KevinOelen @franklinsijo

As for explanation i figured it on my own. However "mapping template" comes from https://medium.com/simple-thoughts-amplified/passing-variables-from-aws-api-gateway-to-lambda-3c5d8602081b