pip-accel: Accelerator for pip, the Python package manager

The pip-accel program is a wrapper for pip, the Python package manager. It accelerates the usage of pip to initialize Python virtual environments given one or more requirements files. It does so by combining the following two approaches:

1. Source distribution downloads are cached and used to generate a local index of source distribution archives. If all your dependencies are pinned to absolute versions whose source distribution downloads were previously cached, pip-accel won’t need a network connection at all! This is one of the reasons why pip can be so slow: given absolute pinned dependencies available in the download cache it will still scan PyPI and distribution websites.
2. Binary distributions are used to speed up the process of installing dependencies with binary components (like M2Crypto and LXML). Instead of recompiling these dependencies again for every virtual environment we compile them once and cache the result as a binary *.tar.gz distribution.

Usage

The pip-accel command supports all subcommands and options supported by pip, however it is of course only useful for the pip install subcommand. So for example:

pip-accel install -r requirements.txt

If you pass a -v option then pip and pip-accel will both use verbose output. To enable verbose output from pip-accel only, you can set the environment variable PIP_ACCEL_VERBOSE.

Based on the user running pip-accel the following file locations are used by default:

Root user	All other users	Purpose
/root/.pip/download-cache	~/.pip/download-cache	Assumed to be pip’s download cache
/var/cache/pip-accel	~/.pip-accel	Used to store the source/binary indexes

These defaults can be overridden by defining the environment variables PIP_DOWNLOAD_CACHE and/or PIP_ACCEL_CACHE.

How fast is it?

To give you an idea of how effective pip-accel is, below are the results of a test to build a virtual environment for one of the internal codebases of Paylogic. This code base requires more than 40 dependencies including several packages that need compilation with SWIG and a C compiler:

Program	Description	Duration	Percentage
pip	Default configuration	444 seconds	100% (baseline)
pip	With download cache (first run)	416 seconds	94%
pip	With download cache (second run)	318 seconds	72%
pip-accel	First run	397 seconds	89%
pip-accel	Second run	30 seconds	7%

Control flow of pip-accel

The way pip-accel works is not very intuitive but it is very effective. Below is an overview of the control flow. Once you take a look at the code you’ll notice that the steps below are all embedded in a loop that retries several times. This is mostly because of step 2 (downloading the source distributions).

1. Run pip install --no-index --no-install -r requirements.txt to unpack source distributions available in the local source index. This is the first step because pip-accel should accept requirements.txt files as input but it will manually install dependencies from cached binary distributions (without using pip or easy_install):

• If the command succeeds it means all dependencies are already available as downloaded source distributions. We’ll parse the verbose pip output of step 1 to find the direct and transitive dependencies (names and versions) defined in requirements.txt and use them as input for step 3. Go to step 3.
• If the command fails it probably means not all dependencies are available as local source distributions yet so we should download them. Go to step 2.

2. Run pip install --no-install -r requirements.txt to download missing source distributions to the download cache:

• If the command fails it means that pip encountered errors while scanning PyPI, scanning a distribution website, downloading a source distribution or unpacking a source distribution. Usually these kinds of errors are intermittent so retrying a few times is worth a shot. Go to step 2.
• If the command succeeds it means all dependencies are now available as local source distributions; we don’t need the network anymore! Go to step 1.

3. Run python setup.py bdist_dumb --format=gztar for each dependency that doesn’t have a cached binary distribution yet (taking version numbers into account). Go to step 4.
4. Install all dependencies from binary distributions based on the list of direct and transitive dependencies obtained in step 1. We have to do these installations manually because easy_install nor pip support binary *.tar.gz distributions.

Contact

If you have questions, bug reports, suggestions, etc. please create an issue on the GitHub project page. The latest version of pip-accel will always be available on GitHub. The internal API documentation is hosted on Read The Docs.

License

This software is licensed under the MIT license just like pip (on which pip-accel is based).

© 2013 Peter Odding and Paylogic International.