How to manage an application’s base image#

The base image of your application is determined by its Dockerfile.

Whatever base image is used in an application should include the runtime environment(s) you need for your application, such as an appropriate version of Python.

Specify the base image in the Dockerfile#

For example, to use Divio’s own 0.4-py3.7-slim-stretch base image:

FROM divio/base:0.4-py3.7-slim-stretch

Or to use the official Python 3.9 image from Docker:

FROM python:3.9.0

Choosing an appropriate base image#

Pin images to a particular version#

It is always wise to pin a base image to a particular version number. For example, FROM python doesn’t specify which version of Python will be installed, and a future update could introduce breaking changes. Specify the version, e.g. FROM python:3.9.0, to prevent this from happening.

Note also that the underlying operating system components could change. Choosing python:3.9.0 on a Linux-based architecture such as Divio’s will use Debian as the basis for the operating system components, but will not specify which version of Debian. Pinning it further with for example python:3.9.0-buster may be a good idea.

When you pin versions, you should periodically check that the version you have chosen is still well-supported, particularly for security updates.

Check included components#

Multiple different tagged versions can be provided for the same fundamental image. For example, for the official Python image:

  • python:3.9.0

  • python:3.9.0-buster

  • python:3.9.0-slim-buster

  • python:3.9.0-alpine3.12

The slim-buster and alpine variants might seem attractive, because they are lightweight. Be aware however that if as a result they lack components that you must instead install manually, you may lose much of the benefit. For example, python:3.9.0-slim-buster does not include binaries for Postgres or MySQL; if you need them, which is likely, python:3.9.0-buster would be a better choice.

Base images provided by Divio#

As well as the many images, official and unofficial, that are available and suitable for a vast variety of web applications, we also provide a number of Divio-optimised Docker base images for applications.

These base images provide an underlying Linux operating system layer, and other layers in the stack, upon which you can build your custom application. For example, our Python base images will include a particular version of Python, and any system-level components required to run it.

Manually upgrading Divio base applications#

Manually upgrading in the repository requires replacing the base image in the Dockerfile. Remove any existing wrapping tags, # <DOCKER_FROM> (if any) from the Dockerfile to make sure the Control Panel does not overwrite the changes. For example, if the existing base image section of the Dockerfile contains

# <DOCKER_FROM>
FROM divio/base:4.18-py3.6-slim-stretch
# </DOCKER_FROM>

and it has to be upgraded with one of the Divio-optimised Docker base images for applications, say divio/base:2.2-py3.9-slim-buster then the upgraded base image section of the Dockerfile should contain only

FROM divio/base:2.2-py3.9-slim-buster