Troubleshooting

This guide covers the most common issues you may encounter when using conda-pypi and how to resolve them.

Environment Issues

Missing Python

Problem: conda-pypi fails because the target environment has no Python interpreter.

Solution: Ensure your target environment has Python installed:

# Install Python in your environment
conda install -n myenv python>=3.10

# Or when creating a new environment
conda create -n myenv python>=3.10

Invalid environment

Problem: Commands fail when specifying a non-existent environment.

Error messages:

• environment does not exist

• python not found

Solution:

# Check if environment exists
conda env list

# Create the environment if it doesn't exist
conda create -n myenv python=3.10

# Use correct environment name or path
conda pypi install -n myenv package-name

Package Resolution Issues

conda search reports missing repodata.json

Problem: Checking the conda-pypi channel with conda search fails on conda versions older than 26.7.0.

Error messages:

UnavailableInvalidChannel: HTTP 404 Not Found for channel conda-pypi

Artifact noarch/repodata.json not found

Cause: conda search gained support for sharded repodata in conda 26.7.0 (conda/conda#16134). Earlier versions request classic repodata.json metadata, which the conda-pypi channel does not serve.

Solutions: Upgrade to conda 26.7.0 or later. On earlier versions, use a dry-run solve to check the channel instead:

# Make sure the supported solver path is enabled
conda config --set solver rattler
conda config --append channels conda-pypi

# Test with a dry-run solve instead of conda search
conda create --dry-run -n conda-pypi-test django-modern-rest

If a dry-run solve or install also fails while using conda 26.7 or newer and the Rattler solver, report the full error in the GitHub issue tracker.

Package not found on PyPI

Problem: Package doesn’t exist or has a different name on PyPI.

Error messages:

• No matching distribution found

• Could not find a version that satisfies the requirement

• 404 Client Error

Solutions:

# Check package name on PyPI
pip search package-name  # or visit pypi.org

# Try common name variations
conda pypi install python-package-name  # instead of package-name

Dependency resolution timeout

Problem: Complex dependency trees exceed the solver’s retry limit.

Error messages:

• Exceeded maximum of 20 attempts

• Could not resolve dependencies after 20 attempts

Solutions:

# Use --ignore-channels to simplify resolution
conda pypi install --ignore-channels package-name

# Install dependencies from conda first
conda install numpy pandas scipy
conda pypi install your-package

# Try installing packages individually
conda pypi install package1
conda pypi install package2

Conflicting dependencies

Problem: Package requirements conflict with existing environment.

Solutions:

# Preview what would be installed
conda pypi install --dry-run package-name

# Check existing packages
conda list

# Create a fresh environment for testing
conda create -n test python=3.10 pip
conda activate test
conda pypi install package-name

Network and Connectivity Issues

PyPI connectivity problems

Problem: Cannot connect to PyPI servers.

Error messages:

• Connection timeout

• Failed to establish connection

• Name resolution failed

Solutions:

# Test basic connectivity
ping pypi.org

# Try with verbose output to see connection details
conda -v pypi install package-name

# Check conda's network configuration
conda config --show channels

Getting Help

Enable verbose output

For detailed debugging information:

# Basic verbose output
conda -v pypi install package-name

# INFO level logging (repeat -v twice)
conda -vv pypi install package-name

# DEBUG level logging (repeat -v three times) - most useful for troubleshooting
conda -vvv pypi install package-name

# TRACE level logging (repeat -v four times) - maximum detail
conda -vvvv pypi install package-name

When to Seek Further Help

If you encounter issues not covered here:

1. Check the version: Ensure you’re using the latest version of conda-pypi

2. Search existing issues: Check the GitHub repository for similar problems

3. Report issues: When reporting issues, please include all the relevant details

Remember that conda-pypi is still in early development, so feedback about unexpected behavior is valuable for improving the tool.