eCommerce Application

Real-world project from Java Web Developer Nanodegree at Udacity.

The application allows users to perform basic operations for online shopping, including creating new accounts, adding items to the cart, submitting orders and viewing their order history.

The project requires to add authorization with implementing proper security and hashing, analyze the metrics, and automate the configuration and deployment of these systems and the application.

Summary

• Demo
• Technologies
• Deployments
• Analytics

Demo

CI/CD

cicd.mov

Splunk

splunk.mov

Technologies

• Add authorization using Spring Security with username/password and JSON Web Tokens (JWT) combinations.
• Write tests with JUnit and Mockito, and meet at least 80% code coverage level.
• Identify the correct metrics for logging with SLF4J and Log4J, in order to monitor the system.
• Index and analyze the metrics with Splunk.
• Use Jenkins to integrate with the version control and build a CI/CD pipeline to deploy the application to the Tomcat server in a Docker container residing on the AWS EC2 instance.

Deployments

There are 4 types of deployments implemented:

• diff-ec2-freestyle: Jenkins Freestyle Project that deploys the application to the Tomcat in a new container residing on a different EC2 instance from the one where Jenkins is running.

• diff-ec2-maven: Similar to diff-ec2-freestyle, but it's built with the Maven Integration Plugin.

• same-ec2-freestyle: Jenkins Freestyle Project that deploys the application to the Tomcat in a new container residing on the same EC2 instance as the one where Jenkins is running.

• same-ec2-maven: Similar to same-ec2-freestyle, but it's built with the Maven Integration Plugin.

Analytics

The project identifys the correct metrics to log, to monitor the system and index metrics to Splunk.

Metrics

The code traces the following metrics to log:

• CreateUser request successes
• CreateUser request failures
• UsernameNotFoundException when login
• Order requests successes
• Order requests failures

Dashboards

There are two dashboards created for success rate per day, per hour and per minute of any one create user and submit order.

Watch demo for details.

Alerts

There is one alert set up to monitor logins with invalid username. Everytime the alert is triggerd, an email will be sent to the given email address.

Watch demo for details.