Could you explain the concept of layers in Docker and how they affect image creation?

A Docker image consists of a sequence of layers. Each layer is a set of filesystem changes, such as adding or modifying files. When you write a Dockerfile, each instruction (e.g., RUN, COPY, ADD) creates a new layer on top of the previous ones.

Layers have several important properties:

• Caching: If a layer hasn't changed, Docker uses the cached version, speeding up the image build.
• Reusability: Layers can be used by different images, saving disk space.
• Independence: Changes in one layer do not require rebuilding all previous layers.

For example, if you have a Dockerfile:

FROM python:3.9
COPY requirements.txt /app/
RUN pip install -r /app/requirements.txt
COPY . /app

• FROM python:3.9 — base layer
• COPY requirements.txt /app/ — new layer with added file
• RUN pip install ... — layer with installed packages
• COPY . /app — layer with source code

If you change only the source code, Docker will rebuild only the last layer, while the layers with dependencies installation will remain unchanged, saving time.

Thus, understanding layers helps optimize image build and management.