Skip to content

A technique for running SonarQube in Kubernetes. Also, a gathering of code analysis plugins for Gradle.

Notifications You must be signed in to change notification settings

javajon/code-analysis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

30 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Static Code Analysis

Within this project is a growing collection of techniques for ensuring your software development team is continuously analyzing their code, taking heed of the feedback from the tools and shifting the quality improvements to earlier in the development process. It's advisable you Shift left your static code analysis to reduce inferior, embarrassing and insecure solutions delivered to your customers.

In this repository there is an example project called microservice. This simple microservice is based on Java ,Spring Boot and MyBatis. Within Gradle builds a microservice and deploys it to a Docker registry. The Gradle build also contains an check and sonarqube task and these are described below.

Currently, this repository addresses two techniques oriented near the Java ecosystem.

  1. Leveraging Gradle plugins for analysis reporting,
  2. Leveraging SonarQube on Kubernetes.

Analysis reports with Gradle plugins

A few code analysis plugins have been added to the Gradle build. To create the analysis reports run the Gradle task check:

gradlew check

Once this task completes explore the code analysis reports in the build/reports directory. The current list of plugins can be inspected in the gradle/analysis directory.

SonarQube with Kubernetes

Setting up your SonarQube services(s) as fragile snowflakes is both common and not a recommended technique. Any developer should be able to quickly start a service or rely on a team service that matches the same behaviors. The latest SonarQube version, it's plugins and it's configurations should also be easily adjustable. Your software development lifecycle processes (SDLC) should embrace the versioned configuration and deployment of SonarQube across a variety of cattle (not pets) targets.

Follow these instructions to setup a personal SonarQube engine and dashboard. With this you have a strong static code analysis tool backing your code changes all before you submit your work for pull requests. Within SonarQube there are plugins such as for Checkstyle, PMD and Spotbugs. The Spotbugs plugin includes rules for vulnerabilities such as the OWASP top 10.

The code analysis tools within your IDE offers similar rules and can find many of the same technical debts, violations, vulnerabilities and anti-patterns. SonarQube provides a language and IDE agnostic way of supplying cross team analysis rules.

With a running SonarQube the SonarLint plugin for your IDE can be connected the service's endpoint for the rules and engine. SonarLint, within the IDE, is an effect shift left technique to ensure debt is reduced before committing and costly pull requests with peers. Your peers will appreciate your work when you preempt and reduce chatter about low level topics in your pull requests.

Setting up SonarQube on Kubernetes

Kubernetes setup

With Katacoda

Learn how to run Sonarqube on Kubernetes with this interactive Katacoda tutorial.

With Minikube

  1. Install Minikube
  2. Install kubectl command line tool for Kubernetes
  3. Install Helm, a package manager for Kubernetes based applications
  4. Set the profile to use: minikube profile minikube-sonarqube
  5. Start Minikube: minikube start
  6. Verify minikube status and kubectl version run correctly
  7. Initialize Helm with: helm init

SonarQube setup

Install SonarQube using the public, stable Helm chart:

SONAR_CONTEXT=https://raw.githubusercontent.com/javajon/code-analysis/master
helm install stable/sonarqube --name my-sonar --namespace sonarqube -f $SONAR_CONTEXT/sonarqube-values.yaml

Inside the sonarqube-values.yaml file is a variety of settings that will superceed the defaults of the chart. The image tag specifies the SonarQube version where the tag can be found here. As of this writing, there may be newer container tag and plugin versions. The first time, it will take a few minutes for SonarQube and its backing Postgres datastore to start and respond.

Sonarqube access

At the end of the Helm install there is a note section. The SonarQube service by default is exposed with a configuration for a cloud LoadBalancer. If you are running this Helm chart on a cloud based cluster then follow the note to export the SERVICE_IP. If you are running MiniKube then the exposed service is exposed in a different way (more on this here).

Access to the service:

export NODE_PORT=$(kubectl get --namespace sonarqube -o jsonpath="{.spec.ports[0].nodePort}" services my-sonar-sonarqube)
export NODE_IP=$(kubectl get nodes --namespace sonarqube -o jsonpath="{.items[0].status.addresses[0].address}")
export SONAR_SERVICE=$(echo http://$NODE_IP:$NODE_PORT)

An equivalent, shorter form using Minikube is:

export SONAR_SERVICE=$(minikube service my-sonar-sonarqube -n sonarqube --url)

Given the above, the SONAR_SERVICE should be assigned to something like this: http://192.168.99.10x:3xxxx. Point your browser to this endpoint to access SonarQube at echo $SONAR_SERVICE.

Now SonarQube is up and running on your cluster and you should be able to get to the dashboard and login as administrator with admin/admin.

Analyze with SonarQube

The Gradle build within the microservice folder also contains the SonarQube plugin that adds a Gradle task sonarqube. This will analyze the source and publish the results to a SonarQube service.

  • Clone this repository with Git
  • Using Gradle in the microservices folder run:
gradlew -Dsonar.host.url=$SONAR_SERVICE sonarqube

Once the analysis completes, navigate back to the SonarQube dashboard and the project analysis will appear.

A New Tool For Your Toolbox

The Minikube instance that is running is called minikube-sonarqube and it can be readily stopped and started with the Minikube stop and start commands. This virtual machine can be easily reconstituted ready to receive future analysis. To restart this instance with a fresh Bash terminal follow these condensed 2 steps:

  1. Start Minikube with a preconfigured SonarQube: minikube start -p minikube-sonarqube && minikube profile minikube-sonarqube
  2. Run an analysis: gradlew -Dsonar.host.url=$(minikube service my-sonar-sonarqube -n sonarqube --url) sonarqube

Technology stack

Additional information

About

A technique for running SonarQube in Kubernetes. Also, a gathering of code analysis plugins for Gradle.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages