Skip to content

Latest commit

 

History

History
183 lines (126 loc) · 5.95 KB

BestPractices.md

File metadata and controls

183 lines (126 loc) · 5.95 KB
Raw

Docker and Node.js Best Practices

Table of Contents

Environment Variables

Run with NODE_ENV set to production. This is the way you would pass in secrets and other runtime configurations to your application as well.

-e "NODE_ENV=production"

Global npm dependencies

If you need to install global npm dependencies, it is recommended to place those dependencies in the non-root user directory. To achieve this, add the following line to your Dockerfile

ENV NPM_CONFIG_PREFIX=/home/node/.npm-global

ENV PATH=$PATH:/home/node/.npm-global/bin # optionally if you want to run npm global bin without specifying path

Upgrading/downgrading Yarn

If you need to upgrade/downgrade yarn, you can do so by issuing the following commands in your Dockerfile:

FROM node:6

ENV YARN_VERSION 1.5.1

RUN curl -fSLO --compressed "https://yarnpkg.com/downloads/$YARN_VERSION/yarn-v$YARN_VERSION.tar.gz" \
    && tar -xzf yarn-v$YARN_VERSION.tar.gz -C /opt/ \
    && ln -snf /opt/yarn-v$YARN_VERSION/bin/yarn /usr/local/bin/yarn \
    && ln -snf /opt/yarn-v$YARN_VERSION/bin/yarnpkg /usr/local/bin/yarnpkg \
    && rm yarn-v$YARN_VERSION.tar.gz

If you're using an Alpine-based image, curl won't be present, so you'll need to make sure it's installed while using it:

FROM node:6-alpine

ENV YARN_VERSION 1.5.1

RUN apk add --no-cache --virtual .build-deps-yarn curl \
    && curl -fSLO --compressed "https://yarnpkg.com/downloads/$YARN_VERSION/yarn-v$YARN_VERSION.tar.gz" \
    && tar -xzf yarn-v$YARN_VERSION.tar.gz -C /opt/ \
    && ln -snf /opt/yarn-v$YARN_VERSION/bin/yarn /usr/local/bin/yarn \
    && ln -snf /opt/yarn-v$YARN_VERSION/bin/yarnpkg /usr/local/bin/yarnpkg \
    && rm yarn-v$YARN_VERSION.tar.gz \
    && apk del .build-deps-yarn

Handling Kernel Signals

Node.js was not designed to run as PID 1 which leads to unexpected behaviour when running inside of Docker. For example, a Node.js process running as PID 1 will not respond to SIGTERM (CTRL-C) and similar signals. As of Docker 1.13, you can use the --init flag to wrap your Node.js process with a lightweight init system that properly handles running as PID 1.

docker run -it --init node

You can also include Tini directly in your Dockerfile, ensuring your process is always started with an init wrapper.

Non-root User

By default, Docker runs container as root which inside of the container can pose as a security issue. You would want to run the container as an unprivileged user wherever possible. The node images provide the node user for such purpose. The Docker Image can then be run with the node user in the following way:

-u "node"

Alternatively, the user can be activated in the Dockerfile:

FROM node:6.10.3
...
# At the end, set the user to use when running this image
USER node

Note that the node user is neither a build-time nor a run-time dependency and it can be removed or altered, as long as the functionality of the application you want to add to the container does not depend on it.

If you do not want nor need the user created in this image you can remove it with the following:

# For debian based images use:
RUN userdel -r node

# For alpine based images use:
RUN deluser --remove-home node

If you need to change the uid/gid of the user you can use:

RUN groupmod -g 999 node && usermod -u 999 -g 999 node

If you need another name for the user (ex. myapp) execute:

RUN usermod -d /home/myapp -l myapp node

For alpine based images, you do not have groupmod nor usermod, so to change the uid/gid you have to delete the previous user:

RUN deluser --remove-home node \
  && addgroup -S node -g 999 \
  && adduser -S -G node -u 999 node

Memory

By default, any Docker Container may consume as much of the hardware such as CPU and RAM. If you are running multiple containers on the same host you should limit how much memory they can consume.

-m "300M" --memory-swap "1G"

CMD

When creating an image, you can bypass the package.json's start command and bake it directly into the image itself. First off this reduces the number of processes running inside of your container. Secondly it causes exit signals such as SIGTERM and SIGINT to be received by the Node.js process instead of npm swallowing them.

CMD ["node","index.js"]

Docker Run

Here is an example of how you would run a default Node.JS Docker Containerized application:

$ docker run \
  -e "NODE_ENV=production" \
  -u "node" \
  -m "300M" --memory-swap "1G" \
  -w "/home/node/app" \
  --name "my-nodejs-app" \
  node [script]

Security

The Docker team has provided a tool to analyze your running containers for potential security issues. You can download and run this tool from here: https://github.com/docker/docker-bench-security

node-gyp alpine

Here is an example of how you would install dependencies for packages that require node-gyp support on the alpine variant:

FROM node:alpine

RUN apk add --no-cache --virtual .gyp python make g++ \
    && npm install [ your npm dependencies here ] \
    && apk del .gyp

And Here's a multistage build example

FROM node:alpine as builder

## Install build toolchain, install node deps and compile native add-ons
RUN apk add --no-cache --virtual .gyp python make g++
RUN npm install [ your npm dependencies here ]

FROM node:alpine as app

## Copy built node modules and binaries without including the toolchain
COPY --from=builder node_modules .