Deploying NGINX and NGINX Plus with Docker

Note: This post was updated in November 2017 to bring our Docker commands up to current standards and to provide a better and safer-to-use NGINX Plus Dockerfile.

Docker is an open platform for building, shipping, running, and orchestrating distributed applications. As software applications, the open source NGINX product and the enhanced and commercially supported version, NGINX Plus, are great use cases for Docker, and we make an NGINX image available on Docker Hub, the repository of Docker images. This article describes how you can deploy the open source NGINX product using this image from Docker Hub, or create and deploy your own Docker image of NGINX Plus.

Introduction

The Docker open platform includes the Docker Engine – the open source runtime that builds, runs, and orchestrates containers – and Docker Hub, a hosted service where Dockerized applications are distributed, shared, and collaborated on by the entire development community or within the confines of a specific organization.

Docker containers enable developers to focus their efforts on application “content” by separating applications from the constraints of infrastructure. Dockerized applications are instantly portable to any infrastructure – laptop, bare‑metal server, VM, or cloud – making them modular components that can be readily assembled and reassembled into fully featured distributed applications and continuously innovated on in real time.

For more information about Docker, see What is Docker? or the full Docker documentation.

Using the NGINX Docker Image

You can create an NGINX instance in a Docker container using the NGINX image from Docker Hub.

Let’s start with a very simple example. To launch an instance of NGINX running in a container and using the default NGINX configuration, run this command:

# docker run --name mynginx1 -p 80:80 -d nginx \
fcd1fb01b14557c7c9d991238f2558ae2704d129cf9fb97bb4fadf673a58580d

This command creates a container named mynginx1 based on the NGINX image. The command returns the long form of the container ID, which is used in the name of log files; see Managing Logging.

The -d option specifies that it run in detached mode, which means that the container continues to run until stopped but does not respond to commands run on the command line. We discuss later how to interact with the container.

The -p option tells Docker to map the ports exposed in the container by the NGINX image – port 80 – to the specified port on the Docker host. The first parameter specifies the port in the Docker host, while the second parameter is mapped to the port exposed in the container.

To verify that the container was created and is running, and to see the port mappings, we run docker ps. (We’ve split the output across multiple lines here to make it easier to read.)

# docker ps
CONTAINER ID  IMAGE        COMMAND               CREATED        STATUS        ...
fcd1fb01b145 nginx:latest "nginx -g 'daemon of 16 seconds ago Up 15 seconds ...

... PORTS NAMES
... 0.0.0.0:80->80/tcp mynginx1

The PORTS field in the output reports that port 80 on the Docker host is mapped to port 80 in the container. Another way to verify that NGINX is running is to make an HTTP request to that port. The code for the default NGINX welcome page appears:

# curl http://localhost:49167
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
body {
width: 35em;
margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif;
}
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p>

<p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="https://www.nginx.com/">nginx.com</a>.</p>

<p><em>Thank you for using nginx.</em></p>
</body>
</html>

Working with the NGINX Docker Container

So now we have a working NGINX Docker container, but how do we manage the content and the NGINX configuration? And what about logging?

A Note About SSH

It is common to enable SSH access to NGINX instances, but the NGINX image does not have OpenSSH installed, because Docker containers are generally intended to be for a single purpose (in this case running NGINX), and for normal operations there is no need to have shell access directly to the NGINX container. Instead we’ll use other methods supported by Docker. For a detailed discussion of alternatives to SSH access, see Why You Don’t Need to Run SSHd in Your Docker Containers.

Managing Content and Configuration Files

There are several ways you can manage both the content served by NGINX and the NGINX configuration files. Here we cover a few of the options.

Option 1 – Maintain the content and configuration on the Docker host

When the container is created we can tell Docker to mount a local directory on the Docker host to a directory in the container. The NGINX image uses the default NGINX configuration, which uses /usr/share/nginx/html as the container’s root directory and puts configuration files in /etc/nginx. For a Docker host with content in the local directory /var/www and configuration files in /var/nginx/conf, run this command:

# docker run --name mynginx2 --mount source=/var/www,target==/usr/share/nginx/html,readonly --mount source=/var/nginx/conf,target=/etc/nginx/conf,readonly -p 80:80 -d nginx

Now any change made to the files in the local directories /var/www and /var/nginx/conf on the Docker host are reflected in the directories /usr/share/nginx/html and /etc/nginx in the container. The readonly option means these directors are read only inside the container.

Option 2 – Copy files from the Docker host

Another option is to have Docker copy the content and configuration files from a local directory on the Docker host during container creation. Once a container is created, the files are maintained by creating a new container when files change or by modifying the files in the container. A simple way to copy the files is to create a Dockerfile with commands that are run during generation of a new Docker image based on the NGINX image from Docker Hub. For the file‑copy (COPY) commands in the Dockerfile, the local directory path is relative to the build context where the Dockerfile is located.

In our example, the content is in the content directory and the configuration files are in the conf directory, both subdirectories of the directory where the Dockerfile is located. The NGINX image has the default NGINX configuration files, including default.conf, in the /etc/nginx/conf.d directory. Since we instead want to use the configuration files from the host, we include RUN commands that delete the default files:

FROM nginx
RUN rm /etc/nginx/conf.d/default.conf
COPY content /usr/share/nginx/html
COPY conf /etc/nginx

We create our own NGINX image by running the following command from the directory where the Dockerfile is located. Note the period (“.”) at the end of the command. It defines the current directory as the build context, which contains the Dockerfile and the directories to be copied.

# docker build -t mynginx_image1 .

Now we run this command to create a container called mynginx3 based on the mynginx_image1 image:

# docker run --name mynginx3 -p 80:80 -d mynginx_image1

If we want to make changes to the files in the container, we use a helper container as described in Option 3.

Option 3 – Maintain files in the container

As mentioned in A Note About SSH, we can’t use SSH to access the NGINX container, so if we want to edit the content or configuration files directly we have to create a helper container that has shell access. For the helper container to have access to the files, we must create a new image that has the proper Docker data volumes defined for the image. Assuming we want to copy files as in Option 2 while also defining volumes, we use the following Dockerfile:

FROM nginx
COPY content /usr/share/nginx/html
COPY conf /etc/nginx
VOLUME /usr/share/nginx/html
VOLUME /etc/nginx

We then create the new NGINX image by running the following command (again note the final period):

# docker build -t mynginx_image2 .

Now we run this command to create an NGINX container based on the image:

# docker run --name mynginx4 -p 80:80 -d mynginx_image2

We then run this command to start a helper container mynginx4_files that has a shell, enabling us to access the content and configuration directories of the mynginx4 container we just created:

# docker run -i -t --volumes-from mynginx4 --name mynginx4_files debian /bin/bash
root@b1cbbad63dd1:/#

The new mynginx4_files helper container runs in the foreground with a persistent standard input (the -i option) and a tty (the -t option). All volumes defined in mynginx4 are mounted as local directories in the helper container.

The debian argument means that the helper container uses the Debian image from Docker Hub. Because the NGINX image also uses Debian (and all of our examples so far use the NGINX image), it is most efficient to use Debian for the helper container, rather than having Docker load another operating system. The /bin/bash argument means that the bash shell runs in the helper container, presenting a shell prompt that you can use to modify files as needed.

To start and stop the container, run the following commands:

# docker start mynginx4_files
# docker stop mynginx4_files

To exit the shell but leave the container running, press Ctrl+p followed by Ctrl+q. To regain shell access to a running container, run this command:

# docker attach mynginx4_files

To exit the shell and terminate the container, run the exit command.

Managing Logging

You can configure either default or customized logging.

Using default logging

The NGINX image is configured to send the main NGINX access and error logs to the Docker log collector by default. This is done by linking them to stdout and stderr; all messages from both logs are then written to the file /var/lib/docker/containers/container-ID/container-ID-json.log on the Docker host. The container-ID is the long-form ID returned when you create a container. To display it, run this command:

# docker inspect --format '{{ .Id }}' container-name

You can use both the Docker command line and the Docker Engine API to extract the log messages. On the command line, run this command:

# docker logs container-name

To use the Docker Remote API, issue a GET request using the Docker Unix sock:

curl --unix-sock /var/run/docker-sock http://localhost/containers/container-name/logs?stdout=1&stderr=1

To include only access log messages in the output, include only stdout=1; to limit the output to error log messages, include only stderr=1. To learn about other available options, see the Docker documentation.

Using customized logging

If you want to implement another method of log collection, or if you want to configure logging differently in certain configuration blocks (such as server{} and location{}), define a Docker volume for the directory or directories in which to store the log files in the container, create a helper container to access the log files, and use whatever logging tools you like. To implement this, create a new image that contains the volume or volumes for the logging files.

For example, to configure NGINX to store log files in /var/log/nginx/log, we can start with the Dockerfile from Option 3 and simply add a VOLUME definition for this directory:

FROM nginx
COPY content /usr/share/nginx/html
COPY conf /etc/nginx
VOLUME /var/log/nginx/log

We can then create an image as described above and use it to create an NGINX container and a helper container that have access to the logging directory. The helper container can have any desired logging tools installed.

Controlling NGINX

Since we do not have direct access to the command line of the NGINX container, we cannot use the nginx command to control NGINX. Fortunately we can use signals to control NGINX, and Docker provides the kill command for sending signals to a container.

To reload the NGINX configuration, run this command:

# docker kill -s HUP container-name

To restart NGINX, run this command to restart the container:

# docker restart container-name

Deploying NGINX Plus with Docker

So far we have discussed Docker for the open source NGINX product, but you can also use it with the commercial product, NGINX Plus. The difference is you first need to create an NGINX Plus image, because as a commercial offering NGINX Plus is not available at Docker Hub. Fortunately, this is quite easy to do.

Note: You should never upload your NGINX Plus images to a public repository such as Docker Hub. Uploading your NGINX Plus images to any public repository, including Docker Hub, violates your license agreement.

Creating a Docker Image of NGINX Plus

To generate an NGINX Plus image, first create a Dockerfile. The example we provide here uses Debian Stretch as the base Docker image. Before you can create the NGINX Plus Docker image, you have to download your version of the nginx-repo-crt and nginx-repo-key files from https://cs.nginx.com, the portal for NGINX Plus customers who have purchased or are evaluating NGINX Plus. Copy the files to the directory where the Dockerfile is located (the Docker build context).

As with open source NGINX, by default the NGINX Plus access and error logs are linked to the Docker log collector. No volumes are specified, but you can add them if desired, or each Dockerfile can be used to create base images from which you can create new images with volumes specified, as described previously.

By default, no files are copied from the Docker host as a container is created. You can add COPY definitions to each Dockerfile, or the image you create can be used as the basis for another image as described above.

NGINX Plus Dockerfile

FROM debian:stretch-slim

LABEL maintainer="NGINX Docker Maintainers "

# Download certificate and key from the customer portal (https://cs.nginx.com)
# and copy to the build context
COPY nginx-repo.crt /etc/ssl/nginx/
COPY nginx-repo.key /etc/ssl/nginx/

# Install NGINX Plus
RUN set -x \
&& apt-get update && apt-get upgrade -y \
&& apt-get install --no-install-recommends --no-install-suggests -y apt-transport-https ca-certificates gnupg1 \
&& \
NGINX_GPGKEY=573BFD6B3D8FBC641079A6ABABF5BD827BD9BF62; \
found=''; \
for server in \
ha.pool.sks-keyservers.net \
hkp://keyserver.ubuntu.com:80 \
hkp://p80.pool.sks-keyservers.net:80 \
pgp.mit.edu \
; do \
echo "Fetching GPG key $NGINX_GPGKEY from $server"; \
apt-key adv --keyserver "$server" --keyserver-options timeout=10 --recv-keys "$NGINX_GPGKEY" && found=yes && break; \
done; \
test -z "$found" && echo >&2 "error: failed to fetch GPG key $NGINX_GPGKEY" && exit 1; \
echo "Acquire::https::plus-pkgs.nginx.com::Verify-Peer \"true\";" >> /etc/apt/apt.conf.d/90nginx \
&& echo "Acquire::https::plus-pkgs.nginx.com::Verify-Host \"true\";" >> /etc/apt/apt.conf.d/90nginx \
&& echo "Acquire::https::plus-pkgs.nginx.com::SslCert \"/etc/ssl/nginx/nginx-repo.crt\";" >> /etc/apt/apt.conf.d/90nginx \
&& echo "Acquire::https::plus-pkgs.nginx.com::SslKey \"/etc/ssl/nginx/nginx-repo.key\";" >> /etc/apt/apt.conf.d/90nginx \
&& printf "deb https://plus-pkgs.nginx.com/debian stretch nginx-plus\n" > /etc/apt/sources.list.d/nginx-plus.list \
&& apt-get update && apt-get install -y nginx-plus \
&& apt-get remove --purge --auto-remove -y gnupg1 \
&& rm -rf /var/lib/apt/lists/*

# Forward request logs to Docker log collector
RUN ln -sf /dev/stdout /var/log/nginx/access.log \
&& ln -sf /dev/stderr /var/log/nginx/error.log

EXPOSE 80

STOPSIGNAL SIGTERM

CMD ["nginx", "-g", "daemon off;"]

Creating the NGINX Plus Image

With the Dockerfile, nginx-repo-crt, and nginx-repo.key files in the same directory, run the following command there to create a Docker image called nginxplus (as before, note the final period):

# docker build --no-cache -t nginxplus .

Note the --no-cache option, which tells Docker to build the image from scratch and ensures the installation of the latest version of NGINX Plus. If the Dockerfile was previously used to build an image and you do not include the --no-cache option, the new image uses the version of NGINX Plus from the Docker cache. (We purposely do not specify a version in the Dockerfile so that the file does not need to change at every new release of NGINX Plus.) Omit the --no-cache option if it’s acceptable to use the NGINX Plus version from the previously built image.

Output like the following from the docker images nginxplus command indicates that the image was created successfully:

# docker images nginxplus
REPOSITORY TAG IMAGE ID CREATED VIRTUAL SIZE
nginxplus latest ef2bf65931cf 6 seconds ago 91.2 MB

To create a container named mynginxplus based on this image, run this command:

# docker run --name mynginxplus -p 80:80 -d nginxplus

You can control and manage NGINX Plus containers in the same way as NGINX containers.

Summary

NGINX, NGINX Plus, and Docker work extremely well together. Whether you use the open source NGINX image from the Docker Hub repository or create your own NGINX Plus image, you can easily spin up new instances of NGINX and NGINX Plus in Docker containers. You can also easily create new Docker images from the base images, making your containers even easier to control and manage. Make sure that all NGINX Plus instances running in your Docker containers are covered by your subscription. For details, please contact the NGINX sales team.

There is much more to Docker than we have been able to cover in this article. More information is available at www.docker.com.

Cover image
Free O'Reilly Ebook
Your guide to everything NGINX