Docker: Lightweight Linux Containers for Consistent Development and Deployment
Containers vs. Other Types of Virtualization
So what exactly is a container and how is it different from hypervisor-based virtualization? To put it simply, containers virtualize at the operating system level, whereas hypervisor-based solutions virtualize at the hardware level. While the effect is similar, the differences are important and significant, which is why I'll spend a little time exploring the differences and the resulting differences and trade-offs.
Both containers and VMs are virtualization tools. On the VM side, a hypervisor makes siloed slices of hardware available. There are generally two types of hypervisors: "Type 1" runs directly on the bare metal of the hardware, while "Type 2" runs as an additional layer of software within a guest OS. While the open-source Xen and VMware's ESX are examples of Type 1 hypervisors, examples of Type 2 include Oracle's open-source VirtualBox and VMware Server. Although Type 1 is a better candidate for comparison to Docker containers, I don't make a distinction between the two types for the rest of this article.
Containers, in contrast, make available protected portions of the operating system—they effectively virtualize the operating system. Two containers running on the same operating system don't know that they are sharing resources because each has its own abstracted networking layer, processes and so on.
Operating Systems and Resources:
Since hypervisor-based virtualization provides access to hardware only, you still need to install an operating system. As a result, there are multiple full-fledged operating systems running, one in each VM, which quickly gobbles up resources on the server, such as RAM, CPU and bandwidth.
Containers piggyback on an already running operating system as their host environment. They merely execute in spaces that are isolated from each other and from certain parts of the host OS. This has two significant benefits. First, resource utilization is much more efficient. If a container is not executing anything, it is not using up resources, and containers can call upon their host OS to satisfy some or all of their dependencies. Second, containers are cheap and therefore fast to create and destroy. There is no need to boot and shut down a whole OS. Instead, a container merely has to terminate the processes running in its isolated space. Consequently, starting and stopping a container is more akin to starting and quitting an application, and is just as fast.
Both types of virtualization and containers are illustrated in Figure 2.
Figure 2. VMs vs. Containers
Isolation for Performance and Security:
Processes executing in a Docker container are isolated from processes running on the host OS or in other Docker containers. Nevertheless, all processes are executing in the same kernel. Docker leverages LXC to provide separate namespaces for containers, a technology that has been present in Linux kernels for 5+ years and is considered fairly mature. It also uses Control Groups, which have been in the Linux kernel even longer, to implement resource auditing and limiting.
The Docker dæmon itself also poses a potential attack vector because it currently runs with root privileges. Improvements to both LXC and Docker should allow containers to run without root privileges and to execute the Docker dæmon under a different system user.
Although the type of isolation provided is overall quite strong, it is arguably not as strong as what can be enforced by virtual machines at the hypervisor level. If the kernel goes down, so do all the containers. The other area where VMs have the advantage is their maturity and widespread adoption in production environments. VMs have been hardened and proven themselves in many different high-availability environments. In comparison, Docker and its supporting technologies have not seen nearly as much action. Docker in particular is undergoing massive changes every day, and we all know that change is the enemy of security.
Docker and VMs—Frenemies:
Now that I've spent all this time comparing Docker and VMs, it's time to acknowledge that these two technologies can actually complement each other. Docker runs just fine on already-virtualized environments. You obviously don't want to incur the cost of encapsulating each application or component in a separate VM, but given a Linux VM, you can easily deploy Docker containers on it. That is why it should not come as a surprise that the officially supported way of using Docker on non-Linux systems, such as OS X and Windows, is to install a Precise64 base Ubuntu virtual machine with the help of Vagrant. Simple detailed instructions are provided on the http://www.docker.io site.
The bottom line is that virtualization and containers exhibit some similarities. Initially, it helps to think of containers as very lightweight virtualization. However, as you spend more time with containers, you come to understand the subtle but important differences. Docker does a nice job of harnessing the benefits of containerization for a focused purpose, namely the lightweight packaging and deployment of applications.
One of Docker's killer features is the ability to find, download and start container images that were created by other developers quickly. The place where images are stored is called a registry, and Docker Inc. offers a public registry also called the Central Index. You can think of the registry along with the Docker client as the equivalent of Node's NPM, Perl's CPAN or Ruby's RubyGems.
In addition to various base images, which you can use to create your own Docker containers, the public Docker Registry features images of ready-to-run software, including databases, content management systems, development environments, Web servers and so on. While the Docker command-line client searches the public Registry by default, it is also possible to maintain private registries. This is a great option for distributing images with proprietary code or components internally to your company. Pushing images to the registry is just as easy as downloading. It requires you to create an account, but that is free as well. Lastly, Docker Inc.'s registry has a Web-based interface for searching for, reading about, commenting on and recommending (aka "starring") images. It is ridiculously easy to use, and I encourage you to click the link in the Resources section of this article and start exploring.
Hands-On with Docker
The client libraries are great for accessing the dæmon programmatically, but the more common use case is to issue instructions from the command line, which is the second way the Docker binary can be used, namely as a command-line client to the REST-based dæmon.
Third, the Docker binary functions as a client to remote repositories of images. Tagged images that make up the filesystem for a container are called repositories. Users can pull images provided by others and share their own images by pushing them to the registry. Registries are used to collect, list and organize repositories.
Let's see all three ways of running the docker executable in action. In this example, you'll search the Docker repository for a MySQL image. Once you find an image you like, you'll download it, and tell the Docker dæmon to run the command (MySQL). You'll do all of this from the command line.
Figure 3. Pulling a Docker Image and Launching a Container
Start by issuing the
docker search mysql command, which then displays
a list of images in the public Docker registry that match the keyword
"mysql". For no particular reason other than I know it works, let's
download the "brice/mysql" image, which you do with the
brice/mysql command. You can see that Docker downloaded not only the
specified image, but also the images it was built on. With the
images command, you list the images currently available locally, which
includes the "brice/mysql" image. Launching the container with the
option to detach from the currently running container, you now have MySQL
running in a container. You can verify that with the
which lists containers, rather than images. In the output, you also see
the port on which MySQL is listening, which is the default of 3306.
But, how do you connect to MySQL, knowing that it is running inside
a container? Remember that Docker containers get their own network
interface. You need to find the IP address and port at which the mysqld
server process is listening. The
<imageId> provides a
lot of info, but since all you need is the IP address, you can just grep for
that when inspecting the container by providing its hash
5a9005441bb5 | grep IPAddress. Now you can connect with the standard MySQL
CLI client by specifying the host and port options. When you're done with
the MySQL server, you can shut it down with
It took seven commands to find, download and launch a Docker container
to get a MySQL server running and shut it down after you're done. In
the process, you didn't have to worry about conflicts with installed
software, perhaps a different version of MySQL, or dependencies. You
used seven different Docker commands: search, pull, images, run, ps,
inspect and stop, but the Docker client actually offers 33 different
commands. You can see the full list by running
help from the
command line or by consulting the on-line manual.
Before exercising Docker in the above example, I mentioned that the client communicates with the dæmon and the Docker Registry via REST-based Web services. That implies that you can use a local Docker client to interact with a remote dæmon, effectively administering your containers on a remote machine. The APIs for the Docker dæmon, Registry and Index are nicely documented, illustrated with examples and available on the Docker site (see Resources).
- Readers' Choice Awards--Nominate Your Apps & Gadgets Now!
- Memory Ordering in Modern Microprocessors, Part I
- Source Code Scanners for Better Code
- diff -u: What's New in Kernel Development
- RSS Feeds
- Non-Linux FOSS: AutoHotkey
- Using Django and MongoDB to Build a Blog
- Tech Tip: Really Simple HTTP Server with Python
- Security Hardening with Ansible