Welcome to this lab exercise where we will explore feature management using Unleash.
In this lab, we will have a play with AWS Comprehend and feature toggles. Please spend some time to read about the different services and tools used in this lab
Amazon Comprehend is an AWS service that offers powerful natural language processing (NLP) capabilities. It enables businesses to analyze and extract valuable insights from unstructured text data, such as customer reviews, social media content, and documents. With features like sentiment analysis, named entity recognition, and language detection, AWS Comprehend helps businesses understand and harness the information contained within textual content.
Sentiment analysis, a key feature of AWS Comprehend, allows businesses to automatically determine the emotional tone and polarity of text, classifying it as positive, negative, neutral, or mixed. By gauging the sentiment expressed in customer feedback, reviews, or social media posts, organizations can swiftly grasp public sentiment about their products or services, facilitating data-driven decision-making and proactive engagement with their audience. This feature is a valuable tool for monitoring and understanding customer opinions, improving brand reputation, and enhancing user experiences.
AWS Lambda is a serverless computing service that allows developers to run code without provisioning or managing servers. It automatically scales applications by running code in response to triggers such as changes in data, HTTP requests, or even specific system events. Lambda functions are event-driven and can be written in various programming languages, including Python, Node.js, Java, and more. With AWS Lambda, users are only charged for the compute time they consume, making it a cost-effective option for executing small, on-demand functions without maintaining dedicated server infrastructure. It's commonly used for microservices architectures, data processing, and automating cloud operations.
AWS Serverless Application Model (AWS SAM) is an open-source framework designed to simplify the development, deployment, and management of serverless applications on AWS. It extends AWS CloudFormation to provide a simplified syntax for defining serverless resources, such as AWS Lambda functions, API Gateway endpoints, DynamoDB tables, and more. With SAM, developers can easily define and deploy complex serverless applications in a declarative template, manage configurations, and package their code for deployment. SAM also includes the AWS SAM CLI, which allows for local testing and debugging of serverless applications, making it a powerful tool for both development and operational efficiency. By streamlining the creation of serverless resources, AWS SAM enables rapid iteration and seamless integration with other AWS services.
Unleash is an open-source feature management platform that enables developers to control the release of features in their applications using feature toggles. It allows for the gradual rollout of new features, enabling teams to enable or disable them without deploying new code. This helps reduce risks during deployment, improve testing, and provide a more flexible development process. Unleash supports various toggle strategies, such as user-specific, gradual rollouts, or environment-based toggles, allowing for fine-grained control over which users or environments see new features. This makes it ideal for continuous delivery and A/B testing scenarios.
In this lab you will
- Use the Unleash UI to create a feature toggle
- Utilize GitHub Codespaces, a cloud-based development environment, to build and run an AWS Lambda function
- Observe how we can switch the toggle on or off and observe the resulting responses
- We will test a gradual rollout strategy and observe that the toggle state is randomized with a percentage of a probability of being on/off
- If the feature toggle is enabled, it will pass the request body to AWS Comprehend
- If the feature toggle is disabled, it will return a mock result of positive, with no confidence score
- We will deploy the Lambda function and test the API endpoint
Observe the effects of setting the toggle on or off by viewing the results in our browser.
The unleash instance is running at https://eu.app.unleash-hosted.com/eubb1043/
- Log into Unleash with credentials given in the class.
- You might also have been given an invite code.
- Please remind your instructor that your user will need "Editor" or "Admin" access. He frequeltly forgets to do this!
When in the Unleash UI
- Click on the "Default" Project
- Click on the "New Feature Flag" button.
- Enter a name for the feature toggle, including your name to avoid naming conflicts (e.g. glenn_toggle)
- Choose the "Release" toggle type.
- Leave the remaining values as their default.
- Click on the "Create Feature Flag" button.
- On the main feature flag page, enable the flag for the "Development" environment
-
Fork this repository to your own GitHub account by clicking the "Fork" button in the top right corner of the GitHub page
-
Once forked, navigate to your forked repository
-
Click the green "Code" button and select the "Codespaces" tab
-
Click "Create codespace on main" to launch your development environment
-
Wait for the Codespace to initialize - this may take a minute or two
-
Once ready, you'll have a fully configured VS Code environment in your browser
-
Familiarize yourself with Codespaces by exploring the interface - it works just like VS Code!
- You will need to create a Terminal in your codespace, when asked, chose the "Continue working in codespaces"
The terminal is available at the bottom of the screen in your Codespace (if you don't see it, go to Terminal > New Terminal from the menu).
When you enable the terminal, GitHub Codespaces might have to do more initialization. Grab a coffee!
On the left side in the Codespace, there is a file browser (Explorer) where you can see all the files in the repository.
Add an unleash token to the code
The token will be given in class, and should be added to this segment of template.yml file
Properties:
CodeUri: hello_world/
Handler: app.lambda_handler
Runtime: python3.9
Environment:
Variables:
UNLEASH_API_TOKEN: "TOKEN"
Architectures:
Also, in the hello_world folder, there is a file app.py - You need to change the following line to use your own feature toggle
if client.is_enabled('glenn_toggle'):Before you can build and run the Lambda function, you need to configure your AWS credentials. These will be provided to you in the classroom.
Open the terminal in your Codespace and run:
aws configureYou will be prompted to enter the following information:
AWS Access Key ID [None]: <enter the access key provided in class>
AWS Secret Access Key [None]: <enter the secret key provided in class>
Default region name [None]: eu-west-1
Default output format [None]: json
This configuration is necessary for:
- Running the Lambda function locally with AWS Comprehend
- Deploying the Lambda function to AWS
You can verify your configuration by running:
aws sts get-caller-identityImportant! Before going forward, make sure you understand the logic of the Lambda function. It uses the toggle to decide whether to:
- Return a mock response for the sentiment analysis
- Call the AWS comprehend service to do a proper sentiment analysis
Open the terminal in your Codespace and run:
sam build --use-container
sam local invoke -e event.jsonThe event.json file looks very comprehensive! But it's actually just how an HTTP request is passed from the API Client to the Lambda function through the AWS Service called API Gateway. Find the "body" element in the file to try out different text
The response might look like something like this;
END RequestId: 5e84abaf-fc32-4d50-86f8-4219764c8d5e
REPORT RequestId: 5e84abaf-fc32-4d50-86f8-4219764c8d5e Init Duration: 0.14 ms Duration: 1151.45 ms Billed Duration: 1152 ms Memory Size: 512 MB Max Memory Used: 512 MB
{"statusCode": 200, "headers": {"Content-Type": "application/json"}, "body": "{\"sentiment\": \"positive\", \"confidence\": 0.9991305470466614}"}
What exactly am I looking at? The JSON you see here is a result of the AWS Comprehend service's sentiment analysis on the text given in the HTTP request body. Is it positive? Is it negative, or neutral?
Play around with it, change the body. How Positive can you get? Puppies and cake?
Try to toggle your feature on and off. When the toggle is off, the lambda should return positive sentiment, but with no confidence. When enabled, it should return 200 ok.
Depending on what you want to focus on - AWS deployment of your function, feature toggles and so on, please feel free to pick any of the bonus challenges.
sam deploy --guidedPlease note that you do not need to provide the "guided" flag after the first deployment has been done. During the deployment process, provide the following input, but use your own name in the stack name.
Setting default arguments for 'sam deploy'
=========================================
Stack Name [sam-app]: your name or something
AWS Region [eu-west-1]:
#Shows you resources changes to be deployed and require a 'Y' to initiate deploy
Confirm changes before deploy [y/N]: N
#SAM needs permission to be able to create roles to connect to the resources in your template
Allow SAM CLI IAM role creation [Y/n]: Y
#Preserves the state of previously provisioned resources when an operation fails
Disable rollback [y/N]: N
HelloWorldFunction may not have authorization defined, Is this okay? [y/N]: Y
Save arguments to configuration file [Y/n]: Y
SAM configuration file [samconfig.toml]:
SAM configuration environment [default]:
This will take some time, and when it is done - the Endpoint given to your lambda will be displayed - something like this (Just an example)
Key HelloWorldApi
Description API Gateway endpoint URL for Prod stage for Hello World function
Value https://6ztkdjfii8.execute-api.eu-west-1.amazonaws.com/Prod/hello/ The lambda function is deployed with the domain name / URL given by value in your output.
If you have Postman installed, or another API client - the URL should be similar to this https://26gfk7hsl6.execute-api.eu-west-1.amazonaws.com/Prod/sentiment
Warning! Please note that there should be /Prod/sentiment in the URL, sometimes SAM outputs /Prod/Hello in the output
The method is POST, and the BODY should be a RAW text
You can test the API with curl from the terminal in your Codespace:
URL=<url from output of SAM deploy>
curl -X POST $URL -H "Content-Type: application/json" -d 'Sharknado is an absolute disaster, and not in the fun, campy way you might expect. The plot is absurd, and not in a clever or entertaining way, The CGI is laughably bad, with the sharks looking more like floating clip art than any real threat.' | jq- Go to the Unleash UI and find your toggle.
- Find the development environment and click the "strategy" button
- Change to Randomized stickiness - and 50%
- Observe that the toggle will evaluate to true in about 50% of the lambda invocations
- Try different settings for the toggle (percentage, sticky randomness etc)
- Look at the documentation for the comprehend client and https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/comprehend.html
- Can you add more functionality? How about detect_dominant_language? Keyword extraction, Toxic language- or entity detection.
- Explore the Unleash IO and see if you can create a gradual roll-out strategy for your toggle!
- Explore the Python SDK for unleash https://docs.getunleash.io/reference/sdks/python