Fetching contributors…
Cannot retrieve contributors at this time
188 lines (117 sloc) 6.22 KB


This section covers more detailed installation info for pywb.

These instructions apply to older versions of pywb but will still work with pywb 0.9.0, although the directory based configuration system and ``wb-manager`` utility remove some of these steps.


pywb has tested in python 2.6, 2.7. It runs best in python 2.7.3+

pywb tool suite provides several WSGI applications, which have been tested under wsgiref, waitress, and uWSGI.

For best results, the uWSGI container is recommended.

Support for Python 3 is planned but not yet implemented.

Sample Data

pywb comes with a a set of sample archived content, also used by the test suite.

The data can be found in sample_archive and contains warc and cdx files.

The sample archive contains recent captures from and

Runnable Apps

The pywb tool suite currently includes several runnable applications, installed as command-line scripts via setuptools, including:

  • wayback -- start the full wayback on port 8080
  • cdx-server -- start standalone cdx server on port 8090
  • wb-manager -- manages creation of collections, adding warcs, indexing, adding metadata, etc...
  • cdx-indexer -- a low-level tool specifically for creating .cdx and .cdxj indexes from web archive files (WARC and ARC).

Step-By-Step Installation

To start a pywb with bundled sample data:

  1. Clone this repo
  2. Install with python install
  3. Run wayback to start the pywb wayback server with reference WSGI implementation.

OR run or to start with uWSGI or gunicorn (see below for more info).

  1. Test pywb in your browser! (pywb is set to run on port 8080 by default).

If everything worked, the following pages should be loading (served from sample_archive dir):

Original Url Latest Capture List of All Captures http://localhost:8080/pywb/ http://localhost:8080/pywb/*/ http://localhost:8080/pywb/ http://localhost:8080/pywb/*/

uWSGI and gunicorn startup scripts

pywb includes sample uWSGI and gunicorn scripts and which pip install uwsgi and gunicorn and attempt to launch the wsgi app with those containers.

Please see uWSGI Installation and Gunicorn QuickStart for more details on installing these containers.


pywb comes with a Vagrantfile to help you set up a VM quickly for testing and deploy pywb with uWSGI.

If you have Vagrant and VirtualBox installed, then you can start a test instance of pywb like so:

git clone
cd pywb
vagrant up

After pywb and all its dependencies are installed, the uWSGI server will startup

spawned uWSGI worker 1 (and the only) (pid: 123, cores: 1)

At this point, you can open a web browser and navigate to the examples above for testing.

Test Suite

Currently pywb includes a full (and growing) suite of unit doctest and integration tests.

Top level integration tests can be found in the tests/ directory, and each subpackage also contains doctests and unit tests.

The full set of tests can be run by executing:

python test

which will run the tests using py.test.

The py.test coverage plugin is used to keep track of test coverage.

Sample Setup

pywb is optionally configurable via yaml.

The simplest config.yaml is roughly as follows:

   pywb: ./sample_archive/cdx/

archive_paths: ./sample_archive/warcs/

This sets up pywb with a single route for collection /pywb

(The the latest version of config.yaml contains additional documentation and specifies all the optional properties, such as ui filenames for Jinja2/html template files.)

For more advanced use, the pywb init path can be customized further:

  • The PYWB_CONFIG_FILE env can be used to set a different yaml file.
  • Custom init app (with or without yaml) can be created. See and for examples of existing initialization paths.

A note on CDX index files

The new wb-manager tool will automatically generate CDX index files for all WARCs and ARCs, so manual updating of CDX indexes is no longer required.

Running wb-manager convert-cdx <path/to/cdx> will also automatically convert any .cdx files to SURT, JSON based format. This is the recommended approach for pywb 0.9.0+

The cdx-indexer also creates files in the SURT format by default.

However, if you need to use existing/legacy .cdx files (and you are unable to convert them as explained above), you may need to set a special config option.

If you are using .cdx files where the key is not in SURT format (that is, the CDX line may start with instaed of com,example)/), simply add the following to the main config.yaml

surt_ordered: false

A SURT CDX key reverses the order of domain and subdomains and allows for improved searching.

Again, this is provided strictly for compatibility, when older cdx files can not be converted to the new format.