Posts Tagged ‘docker’

Scaling Scala: How to Dockerize Using Kubernetes

Kubernetes is the new kid on the block, promising to help deploy applications into the cloud and scale them more quickly. Today, when developing for a microservices architecture, it’s pretty standard to choose Scala for creating API servers.

Microservices are replacing classic monolithic back-end servers with multiple independent services that communicate among themselves and have their own processes and resources.

If there is a Scala application in your plans and you want to scale it into a cloud, then you are at the right place. In this article I am going to show step-by-step how to take a generic Scala application and implement Kubernetes with Docker to launch multiple instances of the application. The final result will be a single application deployed as multiple instances, and load balanced by Kubernetes.

All of this will be implemented by simply importing the Kubernetes source kit in your Scala application. Please note, the kit hides a lot of complicated details related to installation and configuration, but it is small enough to be readable and easy to understand if you want to analyze what it does. For simplicity, we will deploy everything on your local machine. However, the same configuration is suitable for a real-world cloud deployment of Kubernetes.

Scale Your Scala Application with Kubernetes

Be smart and sleep tight, scale your Docker with Kubernetes.

What is Kubernetes?

Before going into the gory details of the implementation, let’s discuss what Kubernetes is and why it’s important.

You may have already heard of Docker. In a sense, it is a lightweight virtual machine.

Docker gives the advantage of deploying each server in an isolated environment, very similar to a stand-alone virtual machine, without the complexity of managing a full-fledged virtual machine.

For these reasons, it is already one of the more widely used tools for deploying applications in clouds. A Docker image is pretty easy and fast to build and duplicable, much easier than a traditional virtual machine like VMWare, VirtualBox, or XEN.

Kubernetes complements Docker, offering a complete environment for managing dockerized applications. By using Kubernetes, you can easily deploy, configure, orchestrate, manage, and monitor hundreds or even thousands of Docker applications.

Kubernetes is an open source tool developed by Google and has been adopted by many other vendors. Kubernetes is available natively on the Google cloud platform, but other vendors have adopted it for their OpenShift cloud services too. It can be found on Amazon AWS, Microsoft Azure, RedHat OpenShift, and even more cloud technologies. We can say it is well positioned to become a standard for deploying cloud applications.

Prerequisites

Now that we covered the basics, let’s check if you have all the prerequisite software installed. First of all, you need Docker. If you are using either Windows or Mac, you need the Docker Toolbox. If you are using Linux, you need to install the particular package provided by your distribution or simply follow the official directions.

We are going to code in Scala, which is a JVM language. You need, of course, the Java Development Kit and the scala SBT tool installed and available in the global path. If you are already a Scala programmer, chances are you have those tools already installed.

If you are using Windows or Mac, Docker will by default create a virtual machine named default with only 1GB of memory, which can be too small for running Kubernetes. In my experience, I had issues with the default settings. I recommend that you open the VirtualBox GUI, select your virtual machine default, and change the memory to at least to 2048MB.

VirtualBox memory settings

The Application to Clusterize

The instructions in this tutorial can apply to any Scala application or project. For this article to have some “meat” to work on, I chose an example used very often to demonstrate a simple REST microservice in Scala, called Akka HTTP. I recommend you try to apply source kit to the suggested example before attempting to use it on your application. I have tested the kit against the demo application, but I cannot guarantee that there will be no conflicts with your code.

So first, we start by cloning the demo application:

git clone https://github.com/theiterators/akka-http-microservice

Next, test if everything works correctly:

cd akka-http-microservice
sbt run

Then, access to http://localhost:9000/ip/8.8.8.8, and you should see something like in the following image:

Akka HTTP microservice is running

Adding the Source Kit

Now, we can add the source kit with some Git magic:

git remote add ScalaGoodies https://github.com/sciabarra/ScalaGoodies
git fetch --all
git merge ScalaGoodies/kubernetes

With that, you have the demo including the source kit, and you are ready to try. Or you can even copy and paste the code from there into your application.

Once you have merged or copied the files in your projects, you are ready to start.

Starting Kubernetes

Once you have downloaded the kit, we need to download the necessary kubectl binary, by running:

bin/install.sh

This installer is smart enough (hopefully) to download the correct kubectl binary for OSX, Linux, or Windows, depending on your system. Note, the installer worked on the systems I own. Please do report any issues, so that I can fix the kit.

Once you have installed the kubectl binary, you can start the whole Kubernetes in your local Docker. Just run:

bin/start-local-kube.sh

The first time it is run, this command will download the images of the whole Kubernetes stack, and a local registry needed to store your images. It can take some time, so please be patient. Also note, it needs direct accesses to the internet. If you are behind a proxy, it will be a problem as the kit does not support proxies. To solve it, you have to configure the tools like Docker, curl, and so on to use the proxy. It is complicated enough that I recommend getting a temporary unrestricted access.

Assuming you were able to download everything successfully, to check if Kubernetes is running fine, you can type the following command:

bin/kubectl get nodes

The expected answer is:

NAME        STATUS    AGE
127.0.0.1   Ready     2m

Note that age may vary, of course. Also, since starting Kubernetes can take some time, you may have to invoke the command a couple of times before you see the answer. If you do not get errors here, congratulations, you have Kubernetes up and running on your local machine.

Dockerizing Your Scala App

Now that you have Kubernetes up and running, you can deploy your application in it. In the old days, before Docker, you had to deploy an entire server for running your application. With Kubernetes, all you need to do to deploy your application is:

  • Create a Docker image.
  • Push it in a registry from where it can be launched.
  • Launch the instance with Kubernetes, that will take the image from the registry.

Luckily, it is way less complicated that it looks, especially if you are using the SBT build tool like many do.

In the kit, I included two files containing all the necessary definitions to create an image able to run Scala applications, or at least what is needed to run the Akka HTTP demo. I cannot guarantee that it will work with any other Scala applications, but it is a good starting point, and should work for many different configurations. The files to look for building the Docker image are:

docker.sbt
project/docker.sbt

Let’s have a look at what’s in them. The file project/docker.sbt contains the command to import the sbt-docker plugin:

addSbtPlugin("se.marcuslonnberg" % "sbt-docker" % "1.4.0")

This plugin manages the building of the Docker image with SBT for you. The Docker definition is in the docker.sbt file and looks like this:

imageNames in docker := Seq(ImageName("localhost:5000/akkahttp:latest"))

dockerfile in docker := {
  val jarFile: File = sbt.Keys.`package`.in(Compile, packageBin).value
  val classpath = (managedClasspath in Compile).value
  val mainclass = mainClass.in(Compile, packageBin).value.getOrElse(sys.error("Expected exactly one main class"))
  val jarTarget = s"/app/${jarFile.getName}"
  val classpathString = classpath.files.map("/app/" + _.getName)
    .mkString(":") + ":" + jarTarget
  new Dockerfile {
    from("anapsix/alpine-java:8")
    add(classpath.files, "/app/")
    add(jarFile, jarTarget)
    entryPoint("java", "-cp", classpathString, mainclass)
  }
}

To fully understand the meaning of this file, you need to know Docker well enough to understand this definition file. However, we are not going into the details of the Docker definition file, because you do not need to understand it thoroughly to build the image.

The beauty of using SBT for building the Docker image is that
the SBT will take care of collecting all the files for you.

Note the classpath is automatically generated by the following command:

val classpath = (managedClasspath in Compile).value

In general, it is pretty complicated to gather all the JAR files to run an application. Using SBT, the Docker file will be generated with add(classpath.files, "/app/"). This way, SBT collects all the JAR files for you and constructs a Dockerfile to run your application.

The other commands gather the missing pieces to create a Docker image. The image will be built using an existing image APT to run Java programs (anapsix/alpine-java:8, available on the internet in the Docker Hub). Other instructions are adding the other files to run your application. Finally, by specifying an entry point, we can run it. Note also that the name starts with localhost:5000 on purpose, because localhost:5000 is where I installed the registry in the start-kube-local.sh script.

Building the Docker Image with SBT

To build the Docker image, you can ignore all the details of the Dockerfile. You just need to type:

sbt dockerBuildAndPush

The sbt-docker plugin will then build a Docker image for you, downloading from the internet all the necessary pieces, and then it will push to a Docker registry that was started before, together with the Kubernetes application in localhost. So, all you need is to wait a little bit more to have your image cooked and ready.

Note, if you experience problems, the best thing to do is to reset everything to a known state by running the following commands:

bin/stop-kube-local.sh
bin/start-kube-local.sh

Those commands should stop all the containers and restart them correctly to get your registry ready to receive the image built and pushed by sbt.

Starting the Service in Kubernetes

Now that the application is packaged in a container and pushed in a registry, we are ready to use it. Kubernetes uses the command line and configuration files to manage the cluster. Since command lines can become very long, and also be able to replicate the steps, I am using the configurations files here. All the samples in the source kit are in the folder kube.

Our next step is to launch a single instance of the image. A running image is called, in the Kubernetes language, a pod. So let’s create a pod by invoking the following command:

bin/kubectl create -f kube/akkahttp-pod.yml

You can now inspect the situation with the command:

bin/kubectl get pods

You should see:

NAME                   READY     STATUS    RESTARTS   AGE
akkahttp               1/1       Running   0          33s
k8s-etcd-127.0.0.1     1/1       Running   0          7d
k8s-master-127.0.0.1   4/4       Running   0          7d
k8s-proxy-127.0.0.1    1/1       Running   0          7d

Status actually can be different, for example, “ContainerCreating”, it can take a few seconds before it becomes “Running”. Also, you can get another status like “Error” if, for example, you forget to create the image before.

You can also check if your pod is running with the command:

bin/kubectl logs akkahttp

You should see an output ending with something like this:

[DEBUG] [05/30/2016 12:19:53.133] [default-akka.actor.default-dispatcher-5] [akka://default/system/IO-TCP/selectors/$a/0] Successfully bound to /0:0:0:0:0:0:0:0:9000

Now you have the service up and running inside the container. However, the service is not yet reachable. This behavior is part of the design of Kubernetes. Your pod is running, but you have to expose it explicitly. Otherwise, the service is meant to be internal.

Creating a Service

Creating a service and checking the result is a matter of executing:

bin/kubectl create -f kube/akkahttp-service.yaml
bin/kubectl get svc

You should see something like this:

NAME               CLUSTER-IP   EXTERNAL-IP   PORT(S)    AGE
akkahttp-service   10.0.0.54                  9000/TCP   44s
kubernetes         10.0.0.1     <none>        443/TCP    3m

Note that the port can be different. Kubernetes allocated a port for the service and started it. If you are using Linux, you can directly open the browser and type http://10.0.0.54:9000/ip/8.8.8.8 to see the result. If you are using Windows or Mac with Docker Toolbox, the IP is local to the virtual machine that is running Docker, and unfortunately it is still unreachable.

I want to stress here that this is not a problem of Kubernetes, rather it is a limitation of the Docker Toolbox, which in turn depends on the constraints imposed by virtual machines like VirtualBox, which act like a computer within another computer. To overcome this limitation, we need to create a tunnel. To make things easier, I included another script which opens a tunnel on an arbitrary port to reach any service we deployed. You can type the following command:

bin/forward-kube-local.sh akkahttp-service 9000

Note that the tunnel will not run in the background, you have to keep the terminal window open as long as you need it and close when you do not need the tunnel anymore. While the tunnel is running, you can open: http://localhost:9000/ip/8.8.8.8 and finally see the application running in Kubernetes.

Final Touch: Scale

So far we have “simply” put our application in Kubernetes. While it is an exciting achievement, it does not add too much value to our deployment. We’re saved from the effort of uploading and installing on a server and configuring a proxy server for it.

Where Kubernetes shines is in scaling. You can deploy two, ten, or one hundred instances of our application by only changing the number of replicas in the configuration file. So let’s do it.

We are going to stop the single pod and start a deployment instead. So let’s execute the following commands:

bin/kubectl delete -f kube/akkahttp-pod.yml
bin/kubectl create -f kube/akkahttp-deploy.yaml

Next, check the status. Again, you may try a couple of times because the deployment can take some time to be performed:

NAME                                   READY     STATUS    RESTARTS   AGE
akkahttp-deployment-4229989632-mjp6u   1/1       Running   0          16s
akkahttp-deployment-4229989632-s822x   1/1       Running   0          16s
k8s-etcd-127.0.0.1                     1/1       Running   0          6d
k8s-master-127.0.0.1                   4/4       Running   0          6d
k8s-proxy-127.0.0.1                    1/1       Running   0          6d

Now we have two pods, not one. This is because in the configuration file I provided, there is the value replica: 2, with two different names generated by the system. I am not going into the details of the configuration files, because the scope of the article is simply an introduction for Scala programmers to jump-start into Kubernetes.

Anyhow, there are now two pods active. What is interesting is that the service is the same as before. We configured the service to load balance between all the pods labeled akkahttp. This means we do not have to redeploy the service, but we can replace the single instance with a replicated one.

We can verify this by launching the proxy again (if you are on Windows and you have closed it):

bin/forward-kube-local.sh akkahttp-service 9000

Then, we can try to open two terminal windows and see the logs for each pod. For example, in the first type:

bin/kubectl logs -f akkahttp-deployment-4229989632-mjp6u

And in the second type:

bin/kubectl logs -f akkahttp-deployment-4229989632-s822x

Of course, edit the command line accordingly with the values you have in your system.

Now, try to access the service with two different browsers. You should expect to see the requests to be split between the multiple available servers, like in the following image:

Kubernets in action

Conclusion

Today we barely scratched the surface. Kubernetes offers a lot more possibilities, including automated scaling and restart, incremental deployments, and volumes. Furthermore, the application we used as an example is very simple, stateless with the various instances not needing to know each other. In the real world, distributed applications do need to know each other, and need to change configurations according to the availability of other servers. Indeed, Kubernetes offers a distributed keystore (etcd) to allow different applications to communicate with each other when new instances are deployed. However, this example is purposefully small enough and simplified to help you get going, focusing on the core functionalities. If you follow the tutorial, you should be able to get a working environment for your Scala application on your machine without being confused by a large number of details and getting lost in the complexity.

This article was written by Michele Sciabarra, a Toptal Scala developer.

Getting Started with Docker: Simplifying Devops

If you like whales, or are simply interested in quick and painless continuous delivery of your software to production, then I invite you to read this introductory Docker Tutorial. Everything seems to indicate that software containers are the future of IT, so let’s go for a quick dip with the container whales Moby Dock andMolly.

Docker, represented by a logo with a friendly looking whale, is an open source project that facilitates deployment of applications inside of software containers. Its basic functionality is enabled by resource isolation features of the Linux kernel, but it provides a user-friendly API on top of it. The first version was released in 2013, and it has since become extremely popular and is being widely used by many big players such as eBay, Spotify, Baidu, and more. In the last funding round, Docker has landed a huge $95 million.

Transporting Goods Analogy

The philosophy behind Docker could be illustrated with a following simple analogy. In the international transportation industry, goods have to be transported by different means like forklifts, trucks, trains, cranes, and ships. These goods come in different shapes and sizes and have different storing requirements: sacks of sugar, milk cans, plants etc. Historically, it was a painful process depending on manual intervention at every transit point for loading and unloading.

It has all changed with the uptake of intermodal containers. As they come in standard sizes and are manufactured with transportation in mind, all the relevant machineries can be designed to handle these with minimal human intervention. The additional benefit of sealed containers is that they can preserve the internal environment like temperature and humidity for sensitive goods. As a result, the transportation industry can stop worrying about the goods themselves and focus on getting them from A to B.

And here is where Docker comes in and brings similar benefits to the software industry.

How Is It Different from Virtual Machines?

At a quick glance, virtual machines and Docker containers may seem alike. However, their main differences will become apparent when you take a look at the following diagram:

Applications running in virtual machines, apart from the hypervisor, require a full instance of the operating system and any supporting libraries. Containers, on the other hand, share the operating system with the host. Hypervisor is comparable to the container engine (represented as Docker on the image) in a sense that it manages the lifecycle of the containers. The important difference is that the processes running inside the containers are just like the native processes on the host, and do not introduce any overheads associated with hypervisor execution. Additionally, applications can reuse the libraries and share the data between containers.

As both technologies have different strengths, it is common to find systems combining virtual machines and containers. A perfect example is a tool named Boot2Docker described in the Docker installation section.

Docker Architecture

Docker Architecture

At the top of the architecture diagram there are registries. By default, the main registry is the Docker Hub which hosts public and official images. Organizations can also host their private registries if they desire.

On the right-hand side we have images and containers. Images can be downloaded from registries explicitly (docker pull imageName) or implicitly when starting a container. Once the image is downloaded it is cached locally.

Containers are the instances of images – they are the living thing. There could be multiple containers running based on the same image.

At the centre, there is the Docker daemon responsible for creating, running, and monitoring containers. It also takes care of building and storing images. Finally, on the left-hand side there is a Docker client. It talks to the daemon via HTTP. Unix sockets are used when on the same machine, but remote management is possible via HTTP based API.

Installing Docker

For the latest instructions you should always refer to the official documentation.

Docker runs natively on Linux, so depending on the target distribution it could be as easy as sudo apt-get install docker.io. Refer to the documentation for details. Normally in Linux, you prepend the Docker commands with sudo, but we will skip it in this article for clarity.

As the Docker daemon uses Linux-specific kernel features, it isn’t possible to run Docker natively in Mac OS or Windows. Instead, you should install an application called Boot2Docker. The application consists of a VirtualBox Virtual Machine, Docker itself, and the Boot2Docker management utilities. You can follow the official installation instructions for MacOS and Windows to install Docker on these platforms.

Using Docker

Let us begin this section with a quick example:

docker run phusion/baseimage echo "Hello Moby Dock. Hello Molly."

We should see this output:

Hello Moby Dock. Hello Molly.

However, a lot more has happened behind the scenes than you may think:

  • The image ‘phusion/baseimage’ was download from Docker Hub (if it wasn’t already in local cache)
  • A container based on this image was started
  • The command echo was executed within the container
  • The container was stopped when the command exitted

On first run, you may notice a delay before the text is printed on screen. If the image had been cached locally, everything would have taken a fraction of a second. Details about the last container can be retrieved by by running docker ps -l:

CONTAINER ID		IMAGE					COMMAND				CREATED			STATUS				PORTS	NAMES
af14bec37930		phusion/baseimage:latest		"echo 'Hello Moby Do		2 minutes ago		Exited (0) 3 seconds ago		stoic_bardeen 

Taking the Next Dive

As you can tell, running a simple command within Docker is as easy as running it directly on a standard terminal. To illustrate a more practical use case, throughout the remainder of this article, we will see how we can utilize Docker to deploy a simple web server application. To keep things simple, we will write a Java program that handles HTTP GET requests to ‘/ping’ and responds with the string ‘pong\n’.

import java.io.IOException;
import java.io.OutputStream;
import java.net.InetSocketAddress;
import com.sun.net.httpserver.HttpExchange;
import com.sun.net.httpserver.HttpHandler;
import com.sun.net.httpserver.HttpServer;

public class PingPong {

    public static void main(String[] args) throws Exception {
        HttpServer server = HttpServer.create(new InetSocketAddress(8080), 0);
        server.createContext("/ping", new MyHandler());
        server.setExecutor(null);
        server.start();
    }

    static class MyHandler implements HttpHandler {
        @Override
        public void handle(HttpExchange t) throws IOException {
            String response = "pong\n";
            t.sendResponseHeaders(200, response.length());
            OutputStream os = t.getResponseBody();
            os.write(response.getBytes());
            os.close();
        }
    }
}

Dockerfile

Before jumping in and building your own Docker image, it’s a good practice to first check if there is an existing one in the Docker Hub or any private registries you have access to. For example, instead of installing Java ourselves, we will use an official image: java:8.

To build an image, first we need to decide on a base image we are going to use. It is denoted by FROMinstruction. Here, it is an official image for Java 8 from the Docker Hub. We are going to copy it into our Java file by issuing a COPY instruction. Next, we are going to compile it with RUN. EXPOSE instruction denotes that the image will be providing a service on a particular port. ENTRYPOINT is an instruction that we want to execute when a container based on this image is started and CMD indicates the default parameters we are going to pass to it.

FROM java:8
COPY PingPong.java /
RUN javac PingPong.java
EXPOSE 8080
ENTRYPOINT ["java"]
CMD ["PingPong"]

After saving these instructions in a file called “Dockerfile”, we can build the corresponding Docker image by executing:

docker build -t toptal/pingpong .

The official documentation for Docker has a section dedicated to best practices regarding writing Dockerfile.

Running Containers

When the image has been built, we can bring it to life as a container. There are several ways we could run containers, but let’s start with a simple one:

docker run -d -p 8080:8080 toptal/pingpong

where -p [port-on-the-host]:[port-in-the-container] denotes the ports mapping on the host and the container respectively. Furthermore, we are telling Docker to run the container as a daemon process in the background by specifying -d. You can test if the web server application is running by attempting to access ‘http://localhost:8080/ping’. Note that on platforms where Boot2docker is being used, you will need to replace ‘localhost’ with the IP address of the virtual machine where Docker is running.

On Linux:

curl http://localhost:8080/ping

On platforms requiring Boot2Docker:

curl $(boot2docker ip):8080/ping

If all goes well, you should see the response:

pong

Hurray, our first custom Docker container is alive and swimming! We could also start the container in an interactive mode -i -t. In our case, we will override the entrypoint command so we are presented with a bash terminal. Now we can execute whatever commands we want, but exiting the container will stop it:

docker run -i -t --entrypoint="bash" toptal/pingpong

There are many more options available to use for starting up the containers. Let us cover a few more. For example, if we want to persist data outside of the container, we could share the host filesystem with the container by using -v. By default, the access mode is read-write, but could be changed to read-only mode by appending :ro to the intra-container volume path. Volumes are particularly important when we need to use any security information like credentials and private keys inside of the containers, which shouldn’t be stored on the image. Additionally, it could also prevent the duplication of data, for example by mapping your local Maven repository to the container to save you from downloading the Internet twice.

Docker also has the capability of linking containers together. Linked containers can talk to each other even if none of the ports are exposed. It can be achieved with –link other-container-name. Below is an example combining mentioned above parameters:

docker run -p 9999:8080 
    --link otherContainerA --link otherContainerB 
    -v /Users/$USER/.m2/repository:/home/user/.m2/repository 
    toptal/pingpong
 Other Container and Image Operations

Unsurprisingly, the list of operations that one could apply to the containers and images is rather long. For brevity, let us look at just a few of them:

  • stop – Stops a running container.
  • start – Starts a stopped container.
  • commit – Creates a new image from a container’s changes.
  • rm – Removes one or more containers.
  • rmi – Removes one or more images.
  • ps – Lists containers.
  • images – Lists images.
  • exec – Runs a command in a running container.

Last command could be particularly useful for debugging purposes, as it lets you to connect to a terminal of a running container:

docker exec -i -t <container-id> bash

Docker Compose for the Microservice World

If you have more than just a couple of interconnected containers, it makes sense to use a tool like docker-compose. In a configuration file, you describe how to start the containers and how they should be linked with each other. Irrespective of the amount of containers involved and their dependencies, you could have all of them up and running with one command: docker-compose up.

Docker in the Wild

Let’s look at three stages of project lifecycle and see how our friendly whale could be of help.

Development

Docker helps you keep your local development environment clean. Instead of having multiple versions of different services installed such as Java, Kafka, Spark, Cassandra, etc., you can just start and stop a required container when necessary. You can take things a step further and run multiple software stacks side by side avoiding the mix-up of dependency versions.

With Docker, you can save time, effort, and money. If your project is very complex to set up, “dockerise” it. Go through the pain of creating a Docker image once, and from this point everyone can just start a container in a snap.

You can also have an “integration environment” running locally (or on CI) and replace stubs with real services running in Docker containers.

Testing / Continuous Integration

With Dockerfile, it is easy to achieve reproducible builds. Jenkins or other CI solutions can be configured to create a Docker image for every build. You could store some or all images in a private Docker registry for future reference.

With Docker, you only test what needs to be tested and take environment out of the equation. Performing tests on a running container can help keep things much more predictable.

Another interesting feature of having software containers is that it is easy to spin out slave machines with the identical development setup. It can be particularly useful for load testing of clustered deployments.

Production

Docker can be a common interface between developers and operations personnel eliminating a source of friction. It also encourages the same image/binaries to be used at every step of the pipeline. Moreover, being able to deploy fully tested container without environment differences help to ensure that no errors are introduced in the build process.

You can seamlessly migrate applications into production. Something that was once a tedious and flaky process can now be as simple as:

docker stop container-id; docker run new-image

And if something goes wrong when deploying a new version, you can always quickly roll-back or change to other container:

docker stop container-id; docker start other-container-id

… guaranteed not to leave any mess behind or leave things in an inconsistent state.

Summary

A good summary of what Docker does is included in its very own motto: Build, Ship, Run.

  • Build – Docker allows you to compose your application from microservices, without worrying about inconsistencies between development and production environments, and without locking into any platform or language.
  • Ship – Docker lets you design the entire cycle of application development, testing, and distribution, and manage it with a consistent user interface.
  • Run – Docker offers you the ability to deploy scalable services securely and reliably on a wide variety of platforms.

Have fun swimming with the whales!

Part of this work is inspired by an excellent book Using Docker by Adrian Mouat.

This article was written by RADEK OSTROWSKI, a Toptal Java developer.

Developing for the Cloud in the Cloud: BigData Development with Docker in AWS

Why you may need it?

I am a developer, and I work daily in Integrated Development Environments (IDE), such as Intellij IDEA or Eclipse. These IDEs are desktop applications. Since the advent of Google Documents, I have seen more and more people moving their work from desktop versions of Word or Excel to the cloud using an online equivalent of a word processor or a spreadsheet application.

There are obvious reasons for using a cloud to keep your work. Today, compared to the traditional desktop business applications, some web applications do not have a significant disadvantage in functionalities. The content is available wherever there is a web browser, and these days, that’s almost everywhere. Collaboration and sharing are easier, and losing files is less likely.

Unfortunately, these cloud advantages are not as common in the world of software development as is for business applications. There are some attempts to provide an online IDE, but they are nowhere close to traditional IDEs.

That is a paradox; while we are still bound to our desktop for daily coding, the software is now spawned on multiple servers. Developers needs to work with stuff they cannot keep any more on their computer. Indeed, laptops are no longer increasing their processing power; having more than 16GB of RAM on a laptop is rare and expensive, and newer devices, tablets, for example, have even less.

However, even if it is not yet possible to replace classic desktop applications for software development, it is possible to move your entire development desktop to the cloud. The day I realized it it was no longer necessary to have all my software on my laptop, and noticing the availability of web version of terminals and VNC, I moved everything to the cloud. Eventually, I developed a build kit for creating that environment in an automated way.

Developer in the cloud

What is the cloud about for a developer? Developing in it, of course!

In this article I present a set of scripts to build a cloud-based development environment for Scala and big data applications, running with Docker in Amazon AWS, and comprising of a web-accessible desktop with IntelliJ IDE, Spark, Hadoop and Zeppelin as services, and also command line tools like a web based SSH, SBT and Ammonite. The kit is freely available on GitHub, and I describe here the procedure for using it to build your instance. You can build your environment and customize it to your particular needs. It should not take you more than 10 minutes to have it up and running.

What is in the “BigDataDevKit”?

My primary goal in developing the kit was that my development environment should be something I can simply fire up, with all the services and servers I work with, and then destroy them when they are no longer needed. This is especially important when you work on different projects, some of them involving a large number of servers and services, as when you work on big data projects.

My ideal cloud-based environment should:

  • Include all the usual development tools, most importantly a graphical IDE.
  • Have the servers and services I need at my fingertips.
  • Be easy and fast to create from scratch, and expandable to add more services.
  • Be entirely accessible using only a web browser.
  • Optionally, allow access with specialized clients (VNC client and SSH client).

Leveraging modern cloud infrastructure and software, the power of modern browsers, a widespread availability of broadband, and the invaluable Docker, I created a development environment for Scala and big data development that, for the better, replaced my development laptop.

Currently, I can work at any time, either from a MacBook Pro, a Surface Tablet, or even an iPad (with a keyboard), although admittedly the last option is not ideal. All these devices are merely clients; the desktop and all the servers are in the cloud.

Docker and Amazon AWS!

My current environment is built using following online services:

  • Amazon Web Services for the servers.
  • GitHub for storing the code.
  • Dropbox to save files.

I also use a couple of free services, like DuckDns for dynamic IP addresses and Let’s encrypt to get a free SSL certificate.

In this environment, I currently have:

  • A graphical desktop with Intellij idea, accessible via a web browser.
  • Web accessible command line tools like SBT and Ammonite.
  • Hadoop for storing files and running MapReduce jobs.
  • Spark Job Server for scheduled jobs.
  • Zeppelin for a web-based notebook.

Most importantly, the web access is fully encrypted with HTTPS, for both web-based VNC and SSH, and there are multiple safeguards to avoid losing data, a concern that is, of course, important when you do not “own” the content on your physical hard disk. Note that getting a copy of all your work on your computer is automatic and very fast. If you lose everything because someone stole your password, you have a copy on your computer anyway, as long as you configured everything correctly.

Using a Web Based Development Environment with AWS and Docker

Now, let’s start describing how the environment works. When I start work in the morning, the first thing is to log into the Amazon Web Services console where I see all my instances. Usually, I have many development instances configured for different projects, and I keep the unused ones turned off to save billing. After all, I can only work on one project at a time. (Well, sometimes I work on two.)

Screen 1

So, I select the instance I want, start it, I wait for a little or go grab a cup of coffee. It’s not so different to turning on your computer. It usually takes a bunch of seconds to have the instance up and running. Once I see the green icon, I open a browser, and I go to a well known URL: https://msciab.duckdns.org/vnc.html. Note, this is my URL; when you create a kit, you will create your unique URL.

Since AWS assigns a new IP to each machine when you start, I configured a dynamic DNS service, so you can always use the same URL to access your server, even if you stop and restart it. You can even bookmark it in your browser. Furthermore, I use HTTPS, with valid keys to get total protection of my work from sniffers, in case I need to manage passwords and other sensitive data.

Screen 2

Once loaded, the system will welcome you with a Web VNC web client, NoVNC. Simply log in and a desktop appears. I use a minimal desktop, intentionally, just a menu with applications, and my only luxury is a virtual desktop (since I open a lot of windows when I develop). For mail, I still rely on other applications, nowadays mostly other browser tabs.

In the virtual machine, I have what I need to develop big data applications. First and foremost, there is an IDE. In the build, I use the IntelliJ Idea community edition. Also, there is the SBT build tool and a Scala REPL, Ammonite.

Screen 3

The key features of this environment, however, are services deployed as containers in the same virtual machine. In particular, I have:

  • Zeppelin, the web notebook for using Scala code on the fly and doing data analysis (http://zeppelin:8080)
  • The Spark Job Server, to execute and deploy spark jobs with a Rest interface (http://sparkjobserver:8080).
  • An instance of Hadoop for storing and retrieving data from the HDFS (http://hadoop:50070).

Note, these URLs are fixed but are accessible within the virtual environment. You can see their web interfaces in the following screenshot.

Screen 4

Each service runs in a separate Docker container. Without becoming too technical, you can think of this as three separate servers inside your virtual machine. The beauty of using Docker is you can add services, and even add two or three virtual machines. Using Amazon containers, you can scale your environment easily.

Last, but not least, you have a web terminal available. Simply access your URL with HTTPS and you will be welcomed with a terminal in a web page.

Screen 5

In the screenshot above you can see I list the containers, which are the three servers plus the desktop. This command line shell gives you access to the virtual machine holding the containers, allowing you to manage them. It’s as if your servers are “in the Matrix” (virtualized within containers), but this shell gives you an escape outside the “Matrix” to manage servers, and desktop. From here, you can restart the containers, access their filesystems and perform other manipulations allowed by Docker. I will not discuss in detail Docker here, but there is a vast amount of documentation on Docker website.

How to setup your instance

Do you like this so far, and you want your instance? It is easy and cheap. You can get it for merely the cost of the virtual machine on Amazon Web Services, plus the storage. The kit in the current configuration requires 4GB of RAM to get all the services running. If you are careful to use the virtual machine only when you need it, and you work, say, 160 hours a month, a virtual machine at current rates will cost 160 x $0.052, or $8 per month. You have to add the cost of storage. I use around 30GB, but everything altogether can be kept under $10.

However, this does bot include the cost of an (eventual) Dropbox (Pro) account, should you want to backup more than 2GB of code. This costs another $15 per month, but it provides important safety for your data. Also, you will need a private repository, either a paid GitHub or another service, such as Bitbucket, which offers free private repositories.

I want to stress that if you use it only when you need it, it is cheaper than a dedicated server. Yes, everything mentioned here can be setup on a physical server, but since I work with big data I need a lot of other AWS services, so I think it is logical to have everything in the same place.

Let’s see how to do the whole setup.

Prerequisites

Before starting to build a virtual machine, you need to register with the following four services:

The only one you need your credit card for is Amazon Web Services. DuckDns is entirely free, while DropBox gives you 2GB of free storage, which can be enough for many tasks. Let’s Encrypt is also free, and it is used internally when you build the image to sign your certificate. Besides these, I recommend a repository hosting service too, like GitHub or Bitbucket, if you want to store your code, however, it is not required for the setup.

To start, navigate to the GitHub BigDataDevKit repository.

Screen 6

Scroll the page and copy the script shown in the image in your text editor of choice:

Screen 7

This script is needed to bootstrap the image. You have to change it and provide some values to the parameters. Carefully, change the text within the quotes. Note you cannot use characters like the quote itself, the backslash or the dollar sign in the password, unless you quote them. This problem is relevant only for the password. If you want to play safe, avoid a quote, dollar sign, or backslashes.

The PASSWORD parameter is a password you choose to access the virtual machine via a web interface. The EMAIL parameter is your email, and will be used when you register an SSL certificate. You will be required to provide your email, and it is the only requirement for getting a free SSL Certificate from Let’s Encrypt.

To get the values for TOKEN and HOST, go to the DuckDNS site and log in. You will need to choose an unused hostname.

Screen 8

Look at the image to see where you have to copy the token and where you have to add your hostname. You must click on the “add domain” button to reserve the hostname.

Configuring your instance

Assuming you have all the parameters and have edited the script, you are ready to launch your instance. Log in to the Amazon Web Services management interface, go to the EC2 Instances panel and click on “Launch Instance”.

Screen 9

In the first screen, you will choose an image. The script is built around the Amazon Linux, and there are no other options available. Select Amazon Linux, the first option in the QuickStart list.

Screen 10

On the second screen, choose the instance type. Given the size of the software running, there are multiple services and you need at least 4GB of memory, so I recommend you select the t2.medium instance. You could trim it down, using the t2.small if you shut down some services, or even the micro if you only want the desktop.

Screen 11

On the third screen, click “Advanced Details” and paste the script you configured in the previous step. I also recommend you enable protection against termination, so that with an accidental termination you won’t lose all your work.

Screen 12

The next step is to configure the storage. The default for an instance is 8GB, which is not enough to contain all the images we will build. I recommend increasing it to 20GB. Also, while it is not needed, I suggest another block device of at least 10GB. The script will mount the second block device as a data folder.You can make a snapshot of its contents, terminate the instance, then recreate it using the snapshot and recovering all the work. Furthermore, a custom block device is not lost when you terminate the instance so have double protection against accidental loss of your data. To increase your safety even further, you can backup your data automatically with Dropbox.

Screen 13

The fifth step is naming the instance. Pick your own. The sixth step offers a way to configure the firewall. By default only SSH is available but we also need HTTPS, so do not forget to add also a rule opening HTTPS. You could open HTTPS to the world, but it’s better if it’s only to your IP to prevent others from accessing your desktop and shell, even though that is still protected with a password.

Once done with this last configuration, you can launch the instance. You will notice that the initialization can take quite a few minutes the first time since the initialization script is running and it will also do some lengthy tasks like generating an HTTPS certificate with Let’s Encrypt.

Screen 14

When you eventually see the management console “running” with a confirmation, and it is no longer “initializing”, you are ready to go.

Assuming all the parameters are correct, you can navigate to https://YOURHOST.duckdns.org.

Replace YOURHOST with the hostname you chose, but do not forget it is an HTTPS site, not HTTP, so your connection to the server is encrypted so you must write https// in the URL. The site will also present a valid certificate for Let’s Encrypt. If there are problems getting the certificate, the initialization script will generate a self-signed certificate. You will still be able to connect with an encrypted connection, but the browser will warn you it is an unknown site, and the connections are insecure. It should not happen, but you never know.

Screen 15

Assuming everything is working, you then access the web terminal, Butterfly. You can log in using the user app and the password you put in the setup script.

Once logged in, you have a bootstrapped virtual machine, which also includes Docker and other goodies, such as a Nginx Frontend, Git, and the Butterfly Web Terminal. Now, you can complete the setup by building the Docker images for your development environment.

Next, type the following commands:

git clone https://github.com/sciabarra/BigDataDevKit
cd BigDataDevKit
sh build.sh

The last command will also ask you to type a password for the Desktop access. Once done, it will start to build the images. Note the build will take a about 10 minutes, but you can see what is happening because everything is shown on the screen.

Once the build is complete, you can also install Dropbox with the following command:

/app/.dropbox-dist/dropboxd

The system will show a link you must click to enable Dropbox. You need to log into Dropbox and then you are done. Whatever you put in the Dropbox folder is automatically synced between all your Dropbox instances.

Once done, you can restart the virtual machine, and access your environment at the https://YOURHOST.dyndns.org/vnc.html URL.

You can stop your machine and restart it when you resume work. The access URL stay the same. This way, you will pay only for the time you are using it, plus monthly extra for the used storage.

Preserving your data

The following discussion requires some knowledge of how Docker and Amazon works. If you do not want to understand the details, just keep in mind following simple rule: In the virtual machine, there is an /app/Dropbox folder available, whatever you place in /app/Dropbox is preserved, and everything else is disposable and can go away. To improve security further, also store your precious code in a version control system.

Now, if you do want to understand this, read on. If you followed my directions in the virtual machine creation, the virtual machine is protected from termination, so you cannot destroy it accidentally. If you expressly decide to terminate it, the primary volume will be destroyed. All the Docker images will be lost, including all the changes you made.

However, since the folder /app/Dropbox is mounted as a Docker Volume for containers, it is not part of Docker images. In the virtual machine, the folder /app is mounted in the Amazon Volume you created, which is also not destroyed even when you expressly terminate the virtual machine. To remove the volume, you have to remove it expressly.

Do not confuse Docker volumes, which are a Docker logical entity, with Amazon Volumes, which is a somewhat physical entity. What happens is that the /app/Dropbox Docker volume is placed inside the /appAmazon volume.

The Amazon Volume is not automatically destroyed when you terminate the virtual machine, so whatever is placed in it will be preserved, until you also expressly destroy the volume. Furthermore, whatever you put in the Docker volume is stored outside of the container, so it is not destroyed when the container is destroyed. If you enabled Dropbox, as recommended, all your content is copied to the Dropbox servers, and to your hard disk if you sync Dropbox with your computer(s). Also, it is recommended that the source code be stored in a version control system.

So, if you place your stuff in version control system under the Dropbox folder, to lose your data all of this must happen:

  • You expressly terminate your virtual machine.
  • You expressly remove the data volume from the virtual machine.
  • You expressly remove the data from Dropbox, including the history.
  • You expressly remove the data from the version control system.

I hope your data is safe enough.

I keep a virtual machine for each project, and when I finish, I keep the unused virtual machines turned off. Of course, I have all my code on GitHub and backed up in Dropbox. Furthermore, when I stop working on a project, I take a snapshot of the Amazon Web Services block before removing the virtual machine entirely. This way, whenever a project resumes, for example for maintenance, all I need to do is start a new virtual machine using the snapshot. All my data goes back in place, and I can resume working.

Optimizing access

First, if you have direct internet access, not mediated by a proxy, you can use native SSH and VNC clients. Direct SSH access is important if you need to copy files in and out of the virtual machine. However, for file sharing, you should consider Dropbox as a simpler alternative.

The VNC web access is invaluable, but sometimes, it can be slower than a native client. You have access to the VNC server on the virtual machine using port 5900. You must expressly open it because it is closed by default. I recommend that you only open it to your IP address, because the internet is full of “robots” that scan the internet looking for services to hook into, and VNC is a frequent target of those robots.

Conclusion

This article explains how you can leverage modern cloud technology to implement an effective development environment. While a machine in the cloud cannot be a complete replacement for your working computer or a laptop, it is good enough for doing development work when it is important to have access to the IDE. In my experience, with current internet connections, it is fast enough to work with.

Being in the cloud, server access and manipulation is faster than having them locally. You can quickly increase (or decrease) memory, fire up another environment, create an image, and so on. You have a datacenter at your fingertips, and when you work with big data projects, well, you need robust services and lots of space. That is what the cloud provides.

The original article was written by MICHELE SCIABARRA – FREELANCE SOFTWARE ENGINEER @ TOPTAL and can be read here.

If you’d like to learn more about Toptal designers or hire one, check this out.

Search This Blog
NetworkedBlogs