DynamoDB access in Lambda@Edge

Question

Need your help in understanding some concepts. I have a web application that uses Lambda@Edge on the CloudFront. This lambda function accesses the DynamoDB - making around 10 independent queries. This generates occasional errors, though it works perfectly when I test the lambda function stand alone. I am not able to make much sense out of the cloudfront logs, and Lambda@Edge does not show up in the CloudWatch.

I have a feeling that the DynamoDB queries are the culprit. (because that is all I am doing in the Lambda function) To make sure, I replicated the data over all regions. But that has not solved the problem. I increased the timeout and memory allocated to the lambda function. But that has not helped in any way. But, reducing the number of DB queries seems to help.

Can you please help me understand this? Is it wrong to make DB queries in the Lambda@Edge? Is there a way to get detailed logs of the Lambda@Edge?

Frederik Frederik · Accepted Answer · 2021-06-22T20:08:53

Over a year too late, but you never know someone benefits of it. Lambda@Edge does not run in a specific region, hence, if you connect to a DynamoDB table, you need to define the region in which this table can be found.

In NodeJS this would result in the below:

// Load the AWS SDK for Node.js
var AWS = require('aws-sdk');
// Set the region 
AWS.config.update({region: 'REGION'});

// Create DynamoDB document client
var docClient = new AWS.DynamoDB.DocumentClient({apiVersion: '2012-08-10'});

As F_SO_K mentioned, you can find your CloudWatch logs in the region closest to you. How to find out which region that would be (in case you're the only one using that specific Lambda@Edge, you can have a look in this documentation)

DynamoDB access in Lambda@Edge

2 Answers