Edge Computing

Creating Serverless Functions (FaaS) with AWS Lambda and Golang

Serverless architecture is a new paradigm that is being used in the development of modern applications. The term serverless refers to the abstraction of servers from application development.  One advantage of serverless computing over server computing is that the cloud provider takes care of the servers while you concentrate on application development.

Function-as-a-Service (FaaS) is a serverless method of running modular code on the cloud. You can use FaaS to write code that is triggered by events.

For example, when a user clicks on a button in your web application, a specific cloud function is executed.

In this tutorial, we'll build a user profile generator function.

The function will accept some user data and return a profile based on the data.

After uploading the generator function to AWS Lambda, we'll create a client program that communicates with the user-profile-generating function on AWS Lambda.

If you are new to serverless computing, it is highly recommended to go through these articles:

Requirements

AWS account to access AWS lambda dashboard.
Use git clone https://github.com/Bamimore-Tomi/faas-golang.git to download the source code for this demo.
Install Go 1.x runtime on your machine.

Setting up an AWS Lambda function

After creating your AWS account, on the main dashboard, search for Lambda.

search-aws-lambda

When you click on Lambda, the AWS Lambda dashboard will be visible, as shown below:

aws-lambda-dashboard

Click on the Create function button to create a new AWS Lambda function.

aws-create-lambda-setting

Here, we specify the name of the function as user-profile.

On the function's dashboard, we will upload the function we want to run in the cloud.

aws-lambda-function-dashboard

In the next section, we'll look at how to make a function that runs on AWS Lambda.

Creating a user-profile function in Golang

In this section, we will see how to create a function that converts it into a zip file before we upload it on AWS Lambda.

We need to install the aws-lambda-go package using:

go mod init
go get github.com/aws/aws-lambda-go/lambda

package main

import (
	"fmt"
	"github.com/aws/aws-lambda-go/lambda"
)

// Struct for the input the program expects from the client
type InfoEvent struct {
	Firstname string `json:"firstname"`
	Lastname string `json:"lastname"`
	Age int `json:"age"`
}

// Struct for the output the server will send back to the client
type Response struct {
	Profile string `json:"profile"`
}

// Event handler, this function handles requests from clients
func HandleInfoEvent(event InfoEvent) (Response, error) {
	return Response{Profile: fmt.Sprintf("Their name is %s %s, they are %d ", event.Firstname, event.Lastname, event.Age)}, nil
}

func main() {
	lambda.Start(HandleInfoEvent)
}

Navigate to the server folder to see this source code.

In the program above, we created the HandleInfoEvent function. This method runs in the cloud.

Use the following steps to compile and zip the program:

Installation - MacOS and Linux

Compile the executable:

GOOS=linux go build main.go

Create the zip file with the following command:

zip user-profile-function.zip main

Installation - Windows

Download the build-lambda-zip tool from GitHub:

go.exe get -u github.com/aws/aws-lambda-go/cmd/build-lambda-zip

You can now use the tool downloaded earlier to create the zip file:

Set GOOS=linux

Go build -o main main.go

%USERPROFILE%\Go\bin\build-lambda-zip.exe -output main.zip main

You can now upload the generated zip file to AWS Lambda.

You can test the function to check if it works well.

First, we need to change the handler's name from the default hello to main.

The Lambda function handler is the method in your function code that processes events.

aws-lambda-edit-hanler

Click on Edit:

aws-lambda-runtime-settings

Change the handler field from hello to main and click Save.

Testing the function on AWS Lambda

On the main dashboard, click on Test:

aws-lambda-dashboard-test

Input this sample data in the JSON editor on the page and click on Test.

{
	"firstname": "Oluwatomisin",
	"lastname": "Bamimore",
	"age": 19
}

aws-lambda-test-prompt

The result will be:

aws-lambda-test-result

Calling the Lambda function from Golang

We can also execute the same function locally.

For that, we have to install the AWS Go SDK as shown:

go get github.com/aws/aws-sdk-go

We can now use this code segment to execute the function in the cloud and see the results locally.

package main

import (
	"encoding/json"
	"fmt"
	"os"
	"github.com/aws/aws-sdk-go/aws"
	"github.com/aws/aws-sdk-go/service/lambda"
	"github.com/joho/godotenv"
	"github.com/aws/aws-sdk-go/aws/credentials"
	"github.com/aws/aws-sdk-go/aws/session"
)

type Info struct {
	Firstname string `json:"firstname"`
	Lastname string `json:"lastname"`
	Age int `json:"age"`
}

type Response struct {
	Profile string `json:"profile"`
}

func main() {
	// Load the environment variable which has the IAM credentials needed to connect to AWS
	godotenv.Load()
	// Initialize new aws session
	sess := session.Must(session.NewSessionWithOptions(session.Options{SharedConfigState: session.SharedConfigEnable}))
	// Initialize new aws client using IAM credentials
	client := lambda.New(sess, &aws.Config{Region: aws.String("us-east-1"), Credentials: credentials.NewStaticCredentials(os.Getenv("aws_access_key_id"), os.Getenv("aws_secret_access_key"), "")})
	// Prepare request parameters
	request := Info{Firstname: "Oluwatomisin", Lastname: "Bamimore", Age: 16}

	payload, err := json.Marshal(request)

	if err != nil {
		fmt.Println("Error marshalling request")
		os.Exit(0)
	}

	result, err := client.Invoke(&lambda.InvokeInput{FunctionName: aws.String("user-profile"), Payload: payload})

	if err != nil {
		fmt.Println("Error calling user-profile function", err)
		os.Exit(0)
	}

	var resp Response

	err = json.Unmarshal(result.Payload, &resp)

	if err != nil {
		fmt.Println("Error unmarshalling user-profile response")
		os.Exit(0)
	}

	// Print response from lambda function
	fmt.Println(resp.Profile)
}

You need to create an IAM user with the correct permissions.

Store the aws_access_key_id and the aws_secret_access_key in a .env file in the directory that you are currently working in.

sample-env-file