Cloud9 Beginner Demo

Build a simple Lambda Function that sits behind an API gateway endpoint

Introduction

This demo aims to show users how easy it is to create, build, test and deploy a Lamda function behind an API Gateway EndPoint. The created Lamda function will be created from one of the existing templates in the SAM repository. This will be initially run locally in Cloud9, where we will debug the application to understand it, change it to suit our needs and then deploy in into AWS.

Prerequisites

An AWS account
A Cloud 9 environment already set up
- Lamda Roles have been setup
- Default region has been configured

Setup (if no cloud 9 environment is available)

Navigate to AWS Console -> CloudFormation
Create New Stack using https://s3-us-west-2.amazonaws.com/cloud9-debug-demo-sydsummit/CFCloud9Debug.json
Next -> Enter Stack Name of Cloud9-debug-demo Next -> Next-> Create ( wait approx 2-3 mins for Cloud9 Environment to be created)

Creating a Lambda Function

Open Cloud9 IDE, Navigate to the "AWS Resources" section on the right hand navigation bar.
- You will see a Lambda section with Local and Remote Functions
- We are going to create a Local Function
Click on the Lamda icon.
- A dialog will appear "Create Servlerless Application"
- Enter Function Name - "MyFirstFunction" - Application Name Field will mirror this -> Next
- Select Node.js 6.10 and choose the "empty-Nodejs" blueprint -> Next
- Add an API Gatway trigger with Resource path /isPalindrome
- For this excercice set the security to none ( just a demo )
- Next -> Finish

So now you will have a Function called MyFirstFunction that will be executed when the /isPalindrome endpoint is hit. The editor will open in the index.js of your function - so lets write some code.

Write or paste in the following code block (replace anything thats already there)

exports.handler = (event, context, callback) => {
    // simple function to calculate if passed in string is Palindromic
    const inputString = event.inputWord;
    const reversyString = inputString.split('').reverse().join('');
    const isPalindrome = (inputString === reversyString);
    
    const result = isPalindrome ? `${inputString} is a Palindrome` : `${inputString} is not a Palindrome`;
    
    callback(null,result);
};

Debuging Lambda

Now we have a function lets run/debug it and test that it works.

hit the Run Button at the top of the screen (The run/debug panel will appear)
hit the Run button inside the run panel ( you should see an error - "cannot read property "split" undefined" ). This is because we havent passed in a string value for the word we want to test. Lets do that by using the payload area.
- add the following Json into the Payload: Textarea {"inputWord":"racecar"}
- hit the run button again and you should get the output "racecar is a Palindrome"
- now add a negative test by changing the payload to test another word and see that working ".... is not a Palindrome"
- Extra Marks - do the same using breakpoints and changing the value of the inputWord inside the debugger.
- tip - make sure you select the debug icon before hitting run.

Now we have successfuly built and tested or Lambda Paidrome checker - lets expose it to the world using the API-Gateway endpoint we created earlier

Debuging Api Gateway Endpoint

We can also test API gateway endpoint locally - We need to first change our Lambda function so it will work behind an API-Gateway endpoint, by changing how it handles its requests and responses

change line 3 of the function to read the queryString Parameters - (we will use query strings for request data - you could also use the body of the request and deserialise the JSON out of that)

const inputString = event.queryStringParameters.string;

change the callback method to be Api Gateway friendly

callback(null, {
     statusCode: 200,
     headers: {
         'Content-Type': 'application/json',
     },
     body: JSON.stringify(result),
     isBase64Encoded: false
 });

change the run/debug panel so that the API Gateway (local) is the system under test - ( there is a drop down on the RHS of the run/debug button panel.
Under Test Payload you should now see entries for Path, Method and Query String
Add the following into the Query String field inputWord=racecar
See the response is correct and test a few more words to be sure.

So we have our Palidrome checker exposed from an API-Gateway endpoint - the last thing we need to do is deploy it into AWS.

Deploy and Test

To Deploy the Function we simply do the following:

Navigate to the "AWS Resources" section on the right hand navigation bar
Right Click on the function and select Deploy
To test you can either use Cloud9 and the APIGateway (remote) option - or open up the AWS console and test using the API Gateway service Stage URL

Thats it - pretty easy and a great way to Build out your function quickly.

Extra Marks ## (Optional)

add some error handling to your function
try changing the request to accept JSON payload instead of a query string parameter hand deserialise it
Inpect the yml file and see it is completely SAM compliant - try and use SAM on the command line to deploy it.

Tear Down

Login to AWS console -> CLoudFormation
Select Cloud9-debug-demo -> Actions -> Delete Stack

wmlba/scratchpad