aws deployment_error ai_generated true

An error occurred (InvalidParameterValueException): Unzipped size must be smaller than 262144000 bytes

ID: aws/lambda-package-too-large

Also available as: JSON · Markdown
88%Fix Rate
90%Confidence
100Evidence
2019-01-01First Seen

Version Compatibility

VersionStatusIntroducedDeprecatedNotes
2 active — — —

Root Cause

Lambda deployment package exceeds the 250MB unzipped limit (or 50MB zipped for direct upload). Common when bundling large dependencies like NumPy, pandas, PyTorch, or native compiled libraries.

generic

Workarounds

  1. 85% success Use Lambda Layers to separate large dependencies 
    # Create a layer with large dependencies:
mkdir -p python/lib/python3.11/site-packages
pip install pandas numpy -t python/lib/python3.11/site-packages/
zip -r layer.zip python/
aws lambda publish-layer-version --layer-name my-deps --zip-file fileb://layer.zip
# Attach layer to function (up to 5 layers, 250MB total unzipped)

    Sources: https://docs.aws.amazon.com/lambda/latest/dg/chapter-layers.html

  2. 92% success Use container image packaging instead of zip (up to 10GB) 
    # Dockerfile:
FROM public.ecr.aws/lambda/python:3.11
COPY requirements.txt .
RUN pip install -r requirements.txt
COPY app.py .
CMD ["app.handler"]
# Build and push to ECR, then create Lambda from image

    Sources: https://docs.aws.amazon.com/lambda/latest/dg/images-create.html

  3. 78% success Strip unnecessary files from dependencies to reduce size 
    # Remove tests, docs, __pycache__, .dist-info:
find python/ -type d -name '__pycache__' -exec rm -rf {} +
find python/ -type d -name 'tests' -exec rm -rf {} +
find python/ -name '*.pyc' -delete
# For numpy/scipy: strip debug symbols from .so files:
find python/ -name '*.so' -exec strip {} +

Dead Ends

Common approaches that don't work:

  1. Compressing the zip file more aggressively 90% fail

    AWS checks the unzipped size, not the zipped size. Better compression reduces upload time but doesn't affect the unzipped limit.

  2. Splitting code into multiple Lambda functions 70% fail

    If the large dependencies are shared across all functions, each function still needs the same large package. Splitting code doesn't reduce dependency size.

Error Chain

Frequently confused with: