Objective

To create two AWS lambdas that communicate via a database. One lambda will insert a new customer record into a Customer table. The other lambda will respond to that customer record creation by inserting the information required for a welcome email into a PendingEmail table.

Before you start

To cut out a lot of complex pre-requisites, this workshop uses a single AWS account. In order to avoid the different groups clashing with eachother, each group will be assigned an animal name that they will use as the prefix for all the different resources they create as part of this exercise (DynamoDB tables, lambdas, microservices, etc.)

The session leaders will provide you with a password for the AWS user you will use to create all your resources.

The supporting artefacts for the exercises can be found in the public repository https://github.com/andylongshaw/serverless

Getting ready

1. Make sure you can log into the AWS console for the account using the AWS user for your group: <MyAnimal>Admin

   Navigate to https://612280087062.signin.aws.amazon.com/console

   Enter the username and password provided to log in

2. MAKE SURE you select "EU (Ireland)" as your default location (from drop-down in the top-right of the console)

3. Use the main console search bar to locate the individual consoles you will need during this exercise:

   lambda (Compute -> Lambda)

   DynamoDB (Database -> DynamoDB)

   API gateway (Application Services -> API Gateway)

   CloudFormation (Management Tools -> CloudFormation)

Create and test a simple microservice

In this part, you will create a simple microservice based on a lambda that is callable through the API gateway

1. Locate the cloudformation folder in the github repository. Make sure you have the two CloudFormation .template files on your local disk.

2. Create a new CloudFormation Stack based on the instructions below

   In the AWS Console, navigate to the CloudFormation console

   Click "Create Stack"

   Select Template -> Choose a template -> Upload a template to Amazon S3 -> Choose File

   Select the file spa2017-apigateway-lambda.template

   Specify details -> Stack name - set this to <MyAnimal>APIGatewayStack

   Parameters -> namePrefixParameter - change this from spa2017 to <MyAnimal>

   Options -> just click "Next"

   Review -> click the "I acknowledge that..." checkbox -> Create

   Wait until the Status of this new stack in the list of Stacks is CREATE_COMPLETE

3. In the AWS console, navigate to the Lambda console and locate the lambda you just created (<MyAnimal>-ApiHandlerLambda)

4. Examine the different tabs on the lambda

   Code, configuration, monitoring

5. Test invoking your lambda in the lambda console

   Click the "Test" button

   Check that the selected test message is the one listed as "Hello World" in the dropdown list

   Click "Save and test"

   Review the execution results at the bottom of the page

6. Test it from the API gateway

   Navigate to the API Gateway console

   Select the <MyAnimal>-LambdaApi

   Under /customer click "GET"

   Click the "Test" link near the lightning flash

   Click the "Test" button

   Examine the response/logs

Make your simple microservice store data in DynamoDB

In this part you will update your lambda function to write a record to the CustomerTable.

1. Create a second Cloud Formation Stack based on the instructions below

   In the AWS Console, navigate to the CloudFormation console

   Select "Create Stack"

   Select Template -> Choose a template -> Upload a template to Amazon S3 -> Choose File

   Select the file spa2017-dynamodb-lambda.template

   Specify details -> Stack name - set this to <MyAnimal>DynamoDBStack

   Parameters -> namePrefixParameter - change this from spa2017 to <MyAnimal>

   Options -> just click "Next"

   Review -> click the "I acknowledge that..." checkbox -> Create

   Wait until the Status of the stack in the list of Stacks is CREATE_COMPLETE

2. In the AWS console, navigate to the DynamoDB console and examine the <MyAnimal>-CustomerTable

   Click "Tables" in the left-hand navigation bar

   Click on the table name

   Examine the information in the Overview tab

   Look at the Items tab - this is where you can see your records appear

3. In the AWS console, navigate to the Lambda console and replace the code in your API gateway lambda with the example code from the repository

   Open the file solutions/python/aws_lambda_customer_rest.py

   Click on the Code tab in your <MyAnimal>-ApiHandlerLambda lambda function

   Paste the code from the file into the code box, completely replacing what was there before

   Click the "Save" button

   Review the code to make sure you can roughly follow what it does

4. Change the table_name variable to be <MyAnimal>-CustomerTable and click "Save"

5. Run the test to insert a record

   Click the "Actions" button and select "Configure test event"

   Locate the messages folder in the github repository and paste the contents of the customer_rest_message_insert.json file into the input test event box, completely replacing what was there before

   Click "Save and Test"

   Examine the execution result and the log output

6. Navigate to the DynamoDB console and check that it contains a Customer matching the data in the message file above

7. Run the test to retrieve the record

   Navigate to the <MyAnimal>-ApiHandlerLambda lambda

   Click the "Actions" button and select "Configure test event"

   Locate the messages folder in the github repository and paste the contents of the customer_rest_message_retrieve.json file into the input test event box, completely replacing what was there before

   Click "Save and Test"

   Examine the execution result and the log output

8. Try adjusting the json files to add/retrieve different records

Create and test the Customer Email trigger

In this part you will create a lambda that will react to the writing of the CustomerTable record by updating a second table (PendingEmailTable).

1. In the DynamoDB console, navigate to the <MyAnimal>-PendingEmailTable

   Click "Tables" in the left-hand navigation bar

   Click on the table name

   Examine the information in the Overview tab

   Look at the Items tab - this is where you can see your records appear

2. Now navigate to the <MyAnimal>-CustomerTable and look at the Triggers tab

   You should see an entry for a lambda called <MyAnimal>-DynamoChangeHandlerLambda

3. Click through the lambda link in the Trigger and populate the trigger lambda with the example code from the repository

   Open the file solutions/python/aws_lambda_email_trigger.py

   Click on the Code tab in your <MyAnimal>-DynamoChangeHandlerLambda lambda function

   Paste the code from the file into the code box, completely replacing what was there before

   Click the "Save" button

   Review the code to make sure you can roughly follow what it does

4. Change the table_name variable to be <MyAnimal>-PendingEmailTable and click "Save"

5. Run the test to insert a record

   Click the "Actions" button and select "Configure test event"

   Locate the messages folder in the github repository and paste the contents of the new_customer_trigger_message.json file into the input test event box, completely replacing what was there before

   Click "Save and Test"

   Examine the execution result and the log output

6. Navigate to the DynamoDB console and check that the <MyAnimal>-PendingEmailTable contains a pending email matching the data in the message file above

Run the lambda combination end-to-end

1. In the AWS console, navigate back to the <MyAnimal>-ApiHandlerLambda lambda

2. Edit the test to change it to the POST/create version of the message and change the customerId and name to something distinctive

3. Run the test

4. Check you get the right customer record in <MyAnimal>-CustomerTable

5. Check you get an appropriate email record in <MyAnimal>-PendingEmailTable

What now?

Well, you could either try the extension exercises or you could go and get a well-earned coffee.