Using Gitlab CI to build and deploy a Java application to AWS Elastic Beanstalk¶

Overview¶

You will learn: - how build advanced CI pipeline with: codestyle checks, unit tests, API tests, performance tests & security checks - learn about cloud services and how to use AWS for deployments

we increase the complexity of the application we are building and deploying
we will be building and testing a Java application and we will deploy it to Amazon Web Services (AWS)
the pipeline is more complex, mostly the same GitLab CI principles will be used.
this course is focused on building pipelines
check the Resourced for links to articles and other video tutorials

Introduction to the Java application¶

About the project - simple Java application that represents a simple car fleet management solution. - use IDE called IntelliJ - check the Resources for some tutorials that can help you get started - together we will be building the CI/CD pipeline - you can work on the .gitlab-ci.yml file without IntelliJ - to run the project from IntelliJ click on Gradle > cars-api > Tasks > application > bootRun - the app exposes an API that allows you to add, view, and remove cars from a database.

What is an API? - an API is a program that does not have a graphical interface, like a website, for example - an API offers raw data, typicall form a database - an API can use used by a front-end application to display the data in the browser - work with the API by using Postman

Running the application in Postman¶

if you have started the application in IntelliJ, you cannot view this in a browser as it does not have a user interface (UI)
use a software development tool called Postman, which is free to download and install
you will find the Postman collection attached to this lecture
can click on “Import” and select both the collection and the environment

Continuous Integration (CI) pipeline overview¶

the first step is the CI pipeline.
the CI pipeline typically has a few stages: build, code quality, test, packaging the application for later use.
The purpose of the CI pipeline is to ensure that the artifact is releasable

most programs go through a build stage.
build process will take the source code and transform it into something that can be executed on a computer
we call this process compilation
for this project, build process will translate source code into Java bytecode that can be executed on the Java Virtual Machine (JVM)
the output is a jar file (which is an archive) that contains this bytecode
to run the build process locally, I will use a tool called Gradle, which is just a build tool
Gradle helps make the build process much simpler

Let’s do the same thing now in GitLab:

define new pipeline file: .gitlab-ci.yml and add it to Git
add build stage & job
publish artifact

Smoke test¶

what is a smoke test?
a straightforward test that checks if something works
if you were responsible for testing a computer, you will only push the power button and see if something appears on the screen
we will start the Java application and see if it works by calling the /health endpoint using cURL

CI/CD pipeline recap¶

Now we have an artifact (or a package of software) and we are ready for deployment.

So what to do next?

There are two oppisite directions in which we can go:

deploy on your own intrastructure (aka server that we control and manage)
deploy using a cloud provider, like Amazon AWS, Google Cloud, Microsoft Azure and many

The advantage using a cloud provider is that you only rent the infrastructure for the time you are using it. Let’s take this example:

If you build an ecommerce shop and you want to host it, you will need to buy a server, place it somewhere save, make backups, handle software updates, fix any hardware failures and so on. During peak times (for example before Christmas), you may need to buy an additional server that will stay idle for the rest of the year.

Using a cloud provider, you can focus on actually building and maintaining the application and forget about the hardware and scalability issues.

Which option makes more sense, it is up to you. For some of the reasons mentioned above, cloud services have risen grately in popularity in the last years.

The following lectures will show to use use Amazon AWS to deploy the Java application.

Brief introduction to Amazon Web Services (AWS)¶

Amazon Web Services (AWS) is a cloud platform offering over 170 services available in data centers located all over the world
AWS offers services that include virtual servers, managed databases, file storage, content delivery, and many others
while creating an account is free, AWS is a paid service
a basic usage of many services is free in the first year
adding a debit/credit card is mandatory

Serverless computing with AWS Elastic Beanstalk¶

with AWS you rent a virtual machine that has a dedicated CPU, memory and disk
with a virtual server you still need to handle software updates, back-ups, monitoring, and any other aspects that ensure your application is running as expected
AWS Elastic Beanstalk is a way to deploy an application but let AWS handle the hardware and software needed to run it
AWS Elastic Beanstalk has a serverless architecture
serverless architecture: you don’t care about the hardware and software needed to run your application
using a serverless architecture is the easiest way to deploy an application in the cloud

Manually deploying a Java application to AWS Elastic Beanstalk¶

use the AWS console interface to upload the jar file
provide the jar file as a resource

How to deploy to AWS from GitLab CI¶

the goal is to avoid any manual interaction
AWS offers a CLI tool that can be used from GitLab CI

Getting started with AWS S3¶

AWS S3 is the main storage service for the entire AWS platform
we will use S3 to upload the artifact

GitLab Group settings¶

the GitLab group functionality is a way to organize similar projects into group
having a group allows you to configure environment variables that are available in multiple projects

How to upload a file to AWS S3 from GitLab CI¶

create user, set environment variables - AWS_ACCESS_KEY_ID - AWS_SECRET_ACCESS_KEY - S3_BUCKET
upload the artifact to S3 using the copy command (aws s3 cp)

How to deploy a Java application to AWS Elastic Beanstalk using the AWS CLI¶

deploying to AWS EB involves two steps - creating a new application version referencing the artifact from S3 - updating the production environment with the latest application version

Create an application version¶

the Java application allows us to insert the application version in one of the info endpoints
this allows us to check which version is currently deployed

Verify the application version after deployment¶

we should ensure that the right application version was deployed
we can do this by checking the /info endpoint which will tell us the current application version

Ensuring coding standards with tools for codestyle checking¶

Most projects want to have a consistent code style and to follow some conventions
automated tools are used to assist with this task
some do static code analysis, as the inspection performed without actually running the code
other tools do a dynamic code analysis, which will actually run the code in order to perform the inspection
one simple one that can be used for Java projects is PMD
PMD is a static code analysis tool
PMD can help find unused variables, problematic code blocks and overall to enforce generally accepted best practices
PMD already has a large set of predefined rules but will also allow you to configure or add new rules, as needed

Introduction to unit testing in CI pipelines¶

what is a unit test
unit tests are responsible for testing only single units of code, typically one class
the execution time is very fast and give instant feedback
use JUnit or similar
for Java projects, JUnit is the most popular framework for writing unit tests

Unit test stage: Run JUnit tests with GitLab CI¶

covers how to add execute JUnit tests
the JUnit XML report is publish as an artifact and recognized by GitLab CI

How to structure a CI/CD pipeline in GitLab CI?¶

when structuring a pipeline, you need to take into account multiple factors
what is the logical flow of the pipeline
which jobs can be grouped together and which not
run jobs that have a similar execution time in parallel
put jobs that are easy to run in the beginning (fail fast principle)

API test stage: Run Postman API tests in GitLab CI¶

for this application, API tests are the highest level of tests we can execute
you can use Postman to write tests
Newman is the CLI tool from Postman
Newman allows you to run Postman tests from the CLI

GitLab Pages¶

allows you to publish HTML websites directly from a repository
great option for creating dashboards
you can publish HTML reports