How to display data from DynamoDB on an AWS S3 hosted static webpage

Question

I have a system which uses AWS IoT to send data to a DynamoDB table. This is simple weather data and stores time, temperature, humidity and pressure. So far that side is working well. I want to display data from this table on a webpage hosted on S3. In its most simple form this would just display the latest row of data. I had though that this would be a case of a simple client-side javascript to query the database, but looking on Amazon it gets quite complicated with Lambda functions called through API gateway using IAM to certify. Is there a simpler way to go about this? Data should be publicly readable, and non-writeable, so I thought should be easier than what I have read so far.

Answer 1

Please have a look at the Simple Web Service CDK Pattern . It helps you create a simple end-to-end service using API Gateway, a Lambda function, and access to a DynamoDB table with just a few lines of code. It is available in multiple programming languages.

As a general note: Whenever you want to provide dynamic content, you need some kind of application that takes care of it. An API Gateway backed by an AWS Lambda function is no more complicated than running a web server with all the undifferentiated heavy lifting like.network configuration, firewall setup, OS patching, and maintenance. And proper handling of identity and access control needs to be done in any case.

Answer 2

If you really want to just display the latest row, and you prefer to keep you webpage as static as possible, I would consider just writing out the latest row of dynamodb to a simple json file using whatever backend process you want, then that file can be consumed by your front end application without having to worry about IAM Credentials or even the AWS JS SDK - keep it as simple and lightweight as possible.

No need to repeatedly hit your dynamodb to pull back the same data for each page load either - which should also save you some money in the long run.

Answer 3

There is a in-browser JavaScript SDK . It allows the JavaScript on your web page to make calls directly to DynamoDB, without having to make API calls through API Gateway that trigger a Lambda function that itself makes calls to DynamoDB on your behalf.

The main consideration here is how to authenticate your web browser client. It cannot make DynamoDB API calls without AWS credentials. I'm assuming that you have no server-side component that can return AWS credentials to the client.

You could embed read-only, minimally-permissioned AWS credentials in the page itself - this is definitely not a best practice, but it's simple and in some cases might be acceptable. Other options include requiring the web site user to authenticate via Web Identity Federation, which would dynamically yield a set of usable, time-limited AWS credentials. Another, more sophisticated option, would be AWS Amplify which supports both authenticated and unauthenticated clients. This is generally preferable to hard-coding (read-only) AWS credentials in a web page but is a little more complex to set up.

There is also a blog post: Dynamic Websites Using the AWS SDK for JavaScript in the Browser .

In all these scenarios, the page itself would make API calls directly to DynamoDB, and you should ensure that the IAM role associated with the credentials is heavily restricted (just the relevant DynamoDB table(s) and just the necessary query/get permissions).