Optimizing Dockerfiles for Layers

Think about moving to a new apartment.

Inefficient: Take 50 trips with small boxes, leaving packing material everywhere, bringing stuff you don’t need.

Efficient: Pack strategically, consolidate, leave behind what you don’t need, clean as you go.

Dockerfile optimization is similar: combine operations, clean up in the same layer, don’t include unnecessary files.

1. Combine RUN Commands

Bad (3 layers, cleanup doesn’t help):

RUN apt-get update
RUN apt-get install -y curl
RUN rm -rf /var/lib/apt/lists/*

Good (1 layer, cleanup works):

RUN apt-get update && \
    apt-get install -y curl && \
    rm -rf /var/lib/apt/lists/*

2. Clean Up in the Same Layer

Once a file is in a layer, it’s there forever — even if you delete it in a later layer. The delete just creates a whiteout; the space is still used.

Bad:

RUN wget huge-file.tar.gz
RUN tar xzf huge-file.tar.gz
RUN rm huge-file.tar.gz  # Still in previous layer!

Good:

RUN wget huge-file.tar.gz && \
    tar xzf huge-file.tar.gz && \
    rm huge-file.tar.gz

3. Use Specific COPY

Bad:

COPY . .  # Includes everything, invalidates often

Good:

COPY requirements.txt .
RUN pip install -r requirements.txt
COPY src/ ./src/
COPY config/ ./config/

4. Order Layers by Change Frequency

# Rarely changes
FROM python:3.11-slim
RUN apt-get update && apt-get install -y ...

# Sometimes changes
COPY requirements.txt .
RUN pip install -r requirements.txt

# Frequently changes
COPY . .
CMD ["python", "app.py"]

5. Use Slim/Alpine Base Images

# 1.01 GB
FROM python:3.11

# 131 MB
FROM python:3.11-slim

# 51 MB (but may have compatibility issues)
FROM python:3.11-alpine

# See layer sizes
docker history myapp

# See total size
docker images myapp

# Detailed size breakdown
docker inspect myapp --format '{{.Size}}'

Use dive to analyze images:

dive myapp

Shows:

• Each layer’s contents
• Wasted space
• Efficiency score