Monthly Archives: September 2016

Minikube – Rapid Dev & Testing for Kubernetes

One of the attendees from Kubernetes for Java Developers training suggested to try minikube for simplified Kubernetes dev and testing. This blog will show how to get started with minikube using a simple Java application.

minikube-logo

Minikube starts a single node Kubernetes cluster on your local machine for rapid development and testing. Requirements lists the exact set of requirements for different operating systems.

This blog will show:

  • Start one node Kubernetes cluster
  • Run Couchbase service
  • Run Java application
  • View Kubernetes Dashboard

All Kubernetes resource description files used in this blog are at github.com/arun-gupta/kubernetes-java-sample/tree/master/maven.

Start Kubernetes Cluster using Minikube

Create a new directory with the name minikube.

In that directory, download kubectl CLI:

Download minikube CLI:

Start the cluster:

The list of nodes can be seen:

More details about the cluster can be obtained using the kubectl cluster-info command:

Behind the scenes, a Virtual Box VM is started.

Complete set of commands supported can be seen by using --help:

Run Couchbase Service

Create a Couchbase service:

This will start a Couchbase service. The service is using the pods created by the replication controller. The replication controller creates a single node Couchbase server.

The configuration file is at github.com/arun-gupta/kubernetes-java-sample/blob/master/maven/couchbase-service.yml and looks like:

Run Java Application

Run the application:

The configuration file is at github.com/arun-gupta/kubernetes-java-sample/blob/master/maven/bootiful-couchbase.yml and looks like:

This is run-once job which runs a Java (Spring Boot) application and upserts (insert or update) a JSON document in Couchbase.

In this job, COUCHBASE_URI environment variable value is set to couchbase-service. This is the service name created earlier. Docker image used for this service is arungupta/bootiful-couchbase and is created using fabric8-maven-plugin as shown at github.com/arun-gupta/kubernetes-java-sample/blob/master/maven/webapp/pom.xml#L57-L68. Specifically, the command for the Docker image is:

This ensures that COUCHBASE_URI environment variable is overriding spring.couchbase.bootstrap-hosts property as defined in application.properties of the Spring Boot application.

Kubernetes Dashboard

Kubernetes 1.4 included an updated dashboard. For minikube, this can be opened using the following command:

 

The default view is shown below:
minikube-dashboard-1-4

But in our case, a few resources have already been created and so this will look like as shown:

minikube-dashboard-couchbase

Notice, our Jobs, Replication Controllers and Pods are shown here.

Shutdown Kubernetes Cluster

The cluster can be easily shutdown:

couchbase.com/containers provide more details about running Couchbase using different orchestration frameworks. Further references:

Source: blog.couchbase.com/2016/september/minikube-rapid-dev–testing-kubernetes

Getting Started with Kubernetes 1.4 using Spring Boot and Couchbase

Kubernetes 1.4 was released earlier this week. Read the blog announcement and CHANGELOG. There are quite a few new features in this release but the key ones that I’m excited about are:

This blog will show:

  • Create a Kubernetes cluster using Amazon Web Services
  • Create a Couchbase service
  • Run a Spring Boot application that stores a JSON document in Couchbase

All the resource description files in this blog are at github.com/arun-gupta/kubernetes-java-sample/tree/master/maven.

Start Kubernetes Cluster

Download binary github.com/kubernetes/kubernetes/releases/download/v1.4.0/kubernetes.tar.gz and extract

Include kubernetes/cluster in PATH

Start a 2-node Kubernetes cluster:

The log will be shown as:

This shows that the Kubernetes cluster has started successfully.

Deploy Couchbase Service

Create Couchbase service and replication controller:

The configuration file is at github.com/arun-gupta/kubernetes-java-sample/blob/master/maven/couchbase-service.yml.

This creates a Couchbase service and the backing replication controller. Name of the service is couchbase-service. This will be used later by the Spring Boot application to communicate with the database.

Check the status of pods:

Note, how the pod status changes from ContainerCreating to Running. The image is downloaded and started in the meanwhile.

Run Spring Boot Application

Run the application:

The configuration file is at github.com/arun-gupta/kubernetes-java-sample/blob/master/maven/bootiful-couchbase.yml. In this service, COUCHBASE_URI environment variable value is set to couchbase-service. This is the service name created earlier.

Docker image used for this service is arungupta/bootiful-couchbase and is created using fabric8-maven-plugin as shown at github.com/arun-gupta/kubernetes-java-sample/blob/master/maven/webapp/pom.xml#L57-L68. Specifically, the command for the Docker image is:

This ensures that COUCHBASE_URI environment variable is overriding spring.couchbase.bootstrap-hosts property as defined in application.properties of the Spring Boot application.

Get the logs:

The main output statement to look in this is

This indicates that the JSON document is upserted (either inserted or updated) in the Couchbase database.

Kubernetes Dashboard

Kubernetes Dashboard is look more comprehensive and claimed to have 90% parity with the CLI. Use kubectl.sh config view command to view the configuration information about the cluster. It looks like:

The clusters.cluster.server property value shows the location of Kubernetes master. The users property show two users that can be used to access the dashboard. Second one uses basic authentication and so copy the username and password property value. In our case, Dashboard UI is accessible at https://52.40.9.27/ui.

kubernetes-dashboard-1-4

All the Kubernetes resources can be easily seen in this fancy dashboard.

Shutdown Kubernetes Cluster

Finally, shutdown the Kubernetes cluster:

couchbase.com/containers provide more details about running Couchbase using different orchestration frameworks.

Further references:

Source: blog.couchbase.com/2016/september/kubernetes-1.4-spring-boot-couchbase

Deployment Pipeline using Docker, Jenkins, Java and Couchbase

This blog explains how to create a Deployment Pipeline using Jenkins and Docker for a Java application talking to a database.

Jenkins support the creation of pipelines. They are built with simple text scripts that use a Pipeline DSL (domain-specific language) based on the Groovy programming language.

The script, typically called Jenkinsfile, defines multiple steps to execute both simple and complex tasks according to the parameters that you establish. Once created, pipelines can build code and orchestrate the work required to drive applications from commit to delivery.

A pipeline consists of steps, node and stage. A pipeline is executed on a node – a computer that is part of Jenkins installation. A pipeline often consists of multiple stages. A stage consists of multiple steps. Read Getting Started with Pipeline for more details.

For our application, here is the basic flow:

docker-pipeline-jenkins

Complete source code for the application used is at github.com/arun-gupta/docker-jenkins-pipeline.

The application is defined in the webapp directory. It opens a connection to the Couchbase database and stores a simple JSON document using Couchbase Java SDK. The application also has a test that verifies that the database indeed contains the document that was persisted.

Many thanks to @alexsotob for helping me with Jenkins configuration.

Let’s get started!

Download and Install Jenkins

  • Download Jenkins from jenkins.io. This was tested with Jenkins 2.21.
  • Start Jenkins:
    This command starts Jenkins by specifying the home directory where all the configuration information is stored. It also defines the port on which Jenkins is listening, 9090 in this case.
  • First start of Jenkins shows the following message in the console:
    Copy the password shown here. This will be used to unlock Jenkins.
  • Access the Jenkins console at localhost:9090 and paste the password:docker-pipeline-jenkins-unlockClick on Next.
  • Create the first admin user as shown:
    docker-pipeline-jenkins-create-admin-user
    Click on Save and Finish.
  • Click on Install suggested plugins:docker-pipeline-jenkins-install-suggested-plugins
    A bunch of default plugins are installed:docker-pipeline-jenkins-installing-suggested-plugins
    Found it surprising that Ant and Subversion are the default plugins.
  • Login screen is prompted.
    docker-pipeline-jenkins-login
    Enter the username and password specified earlier.
  • Finally, Jenkins is ready to use:
    docker-pipeline-jenkins-start-using

That’s quite a bit of steps to get started with basic Jenkins. Do I really have to jump through all these hoops to get started with Jenkins? Is there an easier, simpler, dumber, lazier way to start Jenkins? Follow Convention-over-Configuration and give me one-click pre-configured installation.

Install Jenkins Plugins

Install the required plugins in Jenkins.

  1. If your Java project is built using Maven, then you need to configure Maven in Jenkins. Click on Manage Jenkins, Global Tool Configuration, Maven installations, and specify the location of Maven.docker-pipeline-jenkins-configure-maven
    Name the tool as Maven3 as that is the name used in the configuration later.Again a bit lame, why can’t Jenkins pick up the default location of Maven instead of expecting the user to specify a location.
  2. Click on Manage Jenkins, Manage Plugins, Available tab, search for docker pipe. Select CloudBees Docker Pipeline, click on Install without restart.
    docker-pipeline-jenkins-pipeline-plugin
    Click on Install without restart.Docker Pipeline Plugin plugin understands the Jenkinsfile and executes the commands listed there.
  3. Next screen shows the list of plugins that are installed:docker-pipeline-jenkins-pipeline-plugin-restart-jenkins
    The last line shows that CloudBees Docker Pipeline plugin is installed successfully. Select Restart Jenkins checkbox. This will install restart Jenkins as well.

Create Jenkins Job

Let’s create a job in Jenkins that will run the pipeline.

  1. After Jenkins restarts, it shows the login screen. Enter the username and password created earlier. This brings you back to Installing Plugins/Upgrades page. Click on the Jenkins icon in the top left corner to see the main dashboard:docker-pipeline-jenkins-dashboard
  2. Click on create new jobs, give the name as docker-jenkins-pipeline and choose the type as Pipeline:docker-pipeline-jenkins-create-projectClick on OK.
  3. Configure Pipeline as shown:
    docker-pipeline-jenkins-configure-pipelineLocal git repo is used in this case. You can certainly choose a repo hosted on github. Further, this repo can be configured with a git hook or poll at a constant interval to trigger the pipeline.Click on Save to save the configuration.

Run Jenkins Build

Before you start the job, Couchbase database need to be explicitly started as:

This will be resolved after #9 is fixed.  Make sure you can access Couchbase at http://localhost:8091, use Administrator as the login and password as the password. Click on Data Buckets tab and see the books bucket created.

docker-pipeline-couchbase-books

Click on Build Now and you should see an output similar to:

docker-pipeline-jenkins-build-run

All green is good!

Let’s try to understand what happened behind the scene.

Jenkinsfile describes how the pipeline is built. At the top level, it has four stages – Package, Create Docker Image, Run Application and Run Tests. Each stage is shown as a box in Jenkins dashboard. Total time taken for each stage is shown in the box.

Let’s understand what happens in each stage.

  • Package – Application source code lives in the webapp directory. Maven command mvn clean package -DskipTests is used to create a JAR file of the application. Note that the maven project also includes the tests and are explicitly skipped using -DskipTests. Typically, tests would be in a separate downstream project.Maven project creates a far JAR file of the application and includes all the dependencies.
  • Create Docker Image – Docker image of the application is built using the Dockerfile in the webapp directory. The image simply includes the fat JAR and runs it using java -jar.Each image is tagged with the build number using ${env.BUILD_NUMBER}.
  • Run Application – Running the application involves running the application Docker container.IP address of the database container is identified using the docker inspect command.The database container and the application container are both running in the default bridge network. This allows the two containers to communicate with each other. Another enhancement would be to run the pipeline in a swarm mode cluster. This would require to create and use an overlay network.
  • Run Tests – Tests are run against the container using the mvn test command. If the tests pass the image is pushed to Docker Hub. The test results are captured either way.This stage also shows the usage of try/catch/finally block in Jenkinsfile.If the tests pass then the image is pushed to Docker Hub. In this case, it is available at hub.docker.com/r/arungupta/docker-jenkins-pipeline/tags/.

Some TODOs …

  • Move the tests to a downstream project (#7)
  • Use Git hook or poll to trigger pipeline (#8)
  • Automate database startup/shutdown (#9)
  • Run pipeline in a cluster of Docker Engines with Swarm mode (#10)
  • Show alternate configuration to push image to bintray (#11)

Another pain point is that global variables syntax does not seem to be documented anywhere. It is only available at <JENKINS-HOST>:<JENKINS-PORT>/job/docker-jenkins-pipeline/pipeline-syntax/globals. This is again slightly lame!

not impossible, just not implemented yet” #sadpanda

Some further references to read:

More information about Couchbase:

Feel free to file bugs at github.com/arun-gupta/docker-jenkins-pipeline/issues or send PR.

Source: blog.couchbase.com/2016/september/deployment-pipeline-docker-jenkins-java-couchbase

Docker Service and Swarm Mode to Create Couchbase Cluster

Docker 1.12 introduced Services. A replicated, distributed and load balanced service can be easily created using docker service create command. A “desired state” of the application, such as run 3 containers of Couchbase, is provided and the self-healing Docker engine ensures that that many containers are running in the cluster. If a container goes down, another container is started. If a node goes down, containers on that node are started on a different node.

This blog will show how to setup a Couchbase cluster using Docker Services.

Many thanks to @marcosnils, another fellow Docker Captain, to help me debug the networking!

Couchbase Cluster

A cluster of Couchbase Servers is typically deployed on commodity servers. Couchbase Server has a peer-to-peer topology where all the nodes are equal and communicate to each other on demand. There is no concept of master nodes, slave nodes, config nodes, name nodes, head nodes, etc, and all the software loaded on each node is identical. It allows the nodes to be added or removed without considering their “type”. This model works particularly well with cloud infrastructure in general.

A typical Couchbase cluster creation process looks like:
  • Start Couchbase: Start n Couchbase servers
  • Create cluster: Pick any server, and add all other servers to it to create the cluster
  • Rebalance cluster: Rebalance the cluster so that data is distributed across the cluster
In order to automate using Docker Services, the cluster creation is split into a “master” and “worker” service.
docker-service-couchbase-cluster
The master service has only one replica. This provides a single reference point to start the cluster creation. This service also exposes port 8091. It allows the Couchbase Web Console to be accessible from outside the cluster.
The worker service uses the exact same image as master service. This keeps the cluster homogenous which allows to scale the cluster easily.
Let’s get started!

Setup Swarm Mode on Ubuntu

  1. Launch an Ubuntu instance on Amazon. This blog used mx4.large size for the AMI.
  2. Install Docker:
  3. Docker Swarm mode is an optional feature and need to be explicitly enabled. Initialize Swarm mode:

Create Couchbase “master” Service

  1. Create an overlay network:
    This is required so that multiple Couchbase Docker containers in the cluster can talk to each other.
  2. Create a “master” service:
    This image is created using the Dockerfile here. This Dockerfile uses a configuration script to configure the base Couchbase Docker image. First, it uses Couchbase REST API to setup memory quota, setup index, data and query services, security credentials, and loads a sample data bucket. Then, it invokes the appropriate Couchbase CLI commands to add the Couchbase node to the cluster or add the node and rebalance the cluster. This is based upon three environment variables:
    • TYPE: Defines whether the joining pod is worker or master
    • COUCHBASE_MASTER: Name of the master service
    • AUTO_REBALANCE: Defines whether the cluster needs to be rebalanced
    For this first configuration file, the TYPE environment variable is set to MASTER and so no additional configuration is done on the Couchbase image.

    This service also uses the previously created overlay network named couchbase. It exposes the port 8091 that makes the Couchbase Web Console accessible outside the cluster. This service contains only one replica of the container.

  3. Check status of the Docker service:

    It shows that the service is running. The “desired” and “expected” number of replicas are 1, and thus are matching.

  4. Check the tasks in the service:

    This shows that the container is running.

  5. Access Couchbase Web Console using public IP address and it should look like:docker-service-couchbase-login
    The image used in the configuration file is configured with the Administrator username and password password. Enter the credentials to see the console:
    docker-service-couchbase-web-console
  6. Click on Server Nodes to see how many Couchbase nodes are part of the cluster. As expected, it shows only one node:docker-service-couchbase-one-active-server

Create Couchbase “worker” Service

  1. Create “worker” service:

    This RC also creates a single replica of Couchbase using the same arungupta/couchbase:swarm image. The key differences here are:

    • TYPE environment variable is set to WORKER. This adds a worker Couchbase node to be added to the cluster.
    • COUCHBASE_MASTER environment variable is passed the name of the master service,  couchbase-master.couchbase in our case. This uses the service discovery mechanism built into Docker for the worker and the master to communicate.
  2. Check service:
  3. Checking the Couchbase Web Console shows the updated output:
    docker-service-couchbase-one-pending-server
    It shows that one server is pending to be rebalanced.During the worker service creation, AUTO_REBALANCE environment variable could have been set to true or false to enable rebalance. This ensures that the node is only added to the cluster but the cluster itself is not rebalanced. Rebalancing the cluster requires to re-distribute the data across multiple nodes of the cluster. The recommended way is to add multiple nodes, and then manually rebalance the cluster using the Web Console.

Add Couchbase Nodes by Scaling Docker Service

  1. Scale the service: 

  2. Check the service:
    This shows that 2 replicas of worker are running.
  3. Check the Couchbase Web Console:
    docker-service-couchbase-two-pending-serversAs expected, two servers are now added in the cluster and pending rebalance.
  4. Optionally, you can rebalance the cluster by clicking on the Rebalance button. and it will show like:docker-service-couchbase-rebalancingAfter the rebalancing is complete, the Couchbase Web Console is updated to as as shown:
    docker-service-couchbase-rebalanced
  5. See all the running containers using docker ps:

In addition to creating a cluster, Couchbase Server supports a range of high availability and disaster recovery (HA/DR) strategies. Most HA/DR strategies rely on a multi-pronged approach of maximizing availability, increasing redundancy within and across data centers, and performing regular backups.

Now that your Couchbase cluster is ready, you can run your first sample application.

Learn more about Couchbase and Containers:

Source: http://blog.couchbase.com/2016/september/docker-service-swarm-mode-couchbase-cluster