Skip to main content

Library for working with the PuppetDB REST API.

Project description


pypuppetdtb is a library to work with PuppetDB’s REST API. It is implemented using the requests library. .. _requests:

pypuppetdb >= 0.2.0 requires PuppetDB 3.0.0 or later. There is no support for previous versions beyond 0.1.1

pypuppetdb >= 0.2.2 supports PuppetDB 4.0.0. Backwards compatibility with 3.x is available.

This library is a thin wrapper around the REST API providing some convenience functions and objects to request and hold data from PuppetDB.

To use this library you will need:
  • Python 3.6, 3.7, 3.8


You can install this package from source or from PyPi.

$ pip install pypuppetdb
$ git clone
$ python install

If you wish to hack on it clone the repository but after that run:

$ pip install -r requirements.txt

This will install all the runtime requirements of pypuppetdb and the dependencies for the test suite and generation of documentation.


Native packages for your operating system will be provided in the near future.

OS Status  
Debian 6/Squeeze planned Requires Backports
Debian 7/Wheezy planned  
Ubuntu 13.04 planned  
Ubuntu 13.10 planned  
CentOS/RHEL 5 n/a Python 2.4
CentOS/RHEL 6 planned  
CentOS/RHEL 7 planned  
ArchLinux available Maintained by Tim Meusel
OpenBSD available Maintained by Jasper Lievisse Adriaanse


Once you have pypuppetdb installed you can configure it to connect to PuppetDB and take it from there.


The first thing you need to do is to connect with PuppetDB:

>>> from pypuppetdb import connect
>>> db = connect()


The following will return a generator object yielding Node objects for every returned node from PuppetDB.

>>> nodes = db.nodes()
>>> for node in nodes:
>>>   print(node)

To query a single node the singular node() can be used:

>>> node = db.node('hostname')
>>> print(node)
Node scope

The Node objects are a bit more special in that they can query for facts and resources themselves. Using those methods from a node object will automatically add a query to the request scoping the request to the node.

>>> node = db.node('hostname')
>>> print(node.fact('osfamily'))


>>> facts = db.facts('osfamily')
>>> for fact in facts:
>>>   print(fact)

That queries PuppetDB for the ‘osfamily’ fact and will yield Fact objects, one per node this fact is known for.


>>> resources = db.resources('file')

Will return a generator object containing all file resources you’re managing across your infrastructure. This is probably a bad idea if you have a big number of nodes as the response will be huge.


>>> catalog = db.catalog('hostname')
>>> for res in catalog.get_resources():
>>>     print(res)

Will return a Catalog object with the latest Catalog of the definded host. This catalog contains the defined Resources and Edges.

>>> catalog = db.catalog('hostname')
>>> resource = catalog.get_resource('Service','ntp')
>>> for rel in resource.relationships:
>>>     print(rel)
Class[Ntp] - contains - Service[ntp]
File[/etc/ntp.conf] - notifies - Service[ntp]
File[/etc/ntp.conf] - required-by - Service[ntp]

Will return all Relationships of a given Resource defined by type and title. This will list all linked other Resources and the type of relationship.

Query Builder

Starting with version 0.3.0 pypuppetdb comes shipped with a QueryBuilder component that, as the name suggests, allows users to build PuppetDB AST queries in an Object-Oriented fashion. Vastly superior to constructing long strings than adding additional clauses to fulfill new requirements.

The following code will build a query for the Nodes endpoint to find all nodes belonging to the production environment.

>>> from pypuppetdb.QueryBuilder import *
>>> op = AndOperator()
>>> op.add(EqualsOperator('catalog_environment', 'production'))
>>> op.add(EqualsOperator('facts_environment', 'production'))
>>> print(op)
["and",["=", "catalog_environment", "production"],["=", "facts_environment", "production"]]

This functionality is based on the PuppetDB AST query string syntax documented here.

Subqueries are implemented using corresponding operators (like documented).

  • SubqueryOperator
  • InOperator
  • ExtractOperator
>>> from pypuppetdb.QueryBuilder import *
>>> op = InOperator('certname')
>>> ex = ExtractOperator()
>>> ex.add_field(str('certname'))
>>> sub = SubqueryOperator('events')
>>> sub.add_query(EqualsOperator('status', 'noop'))
>>> ex.add_query(sub)
>>> op.add_query(ex)
>>> print(op)
["in","certname",["extract",["certname"],["select_events",["=", "status", "noop"]]]]

Or using [in <array>] querying:

>>> from pypuppetdb.QueryBuilder import *
>>> op = InOperator('certname')
>>> op.add_array(["", ""])
>>> print(op)
["in","certname",["array", ['', '']]]

You can also access different entities from a single query on the root endpoint with the FromOperator:

>>> op = InOperator('certname')
>>> ex = ExtractOperator()
>>> ex.add_field('certname')
>>> fr = FromOperator('fact_contents')
>>> nd = AndOperator()
>>> nd.add(EqualsOperator("path", ["networking", "eth0", "macaddresses", 0]))
>>> nd.add(EqualsOperator("value", "aa:bb:cc:dd:ee:00"))
>>> ex.add_query(nd)
>>> fr.add_query(ex)
>>> op.add_query(fr)
>>> print(op)
["in","certname",["from","fact_contents",["extract",["certname"],["and",["=", "path", ['networking', 'eth0', 'macaddresses', 0]],["=", "value", "aa:bb:cc:dd:ee:00"]]]]]

Getting Help

This project is still very new so it’s not inconceivable you’ll run into issues.

For bug reports you can file an issue. If you need help with something feel free to pop by #voxpupuli on Freenode or the #puppetboard channel.


API documentation is automatically generated from the docstrings using Sphinx’s autodoc feature.

Documentation will automatically be rebuilt on every push thanks to the Read The Docs webhook. You can find it here.

You can build the documentation manually by doing:

$ cd docs
$ make html

Doing so will only work if you have Sphinx installed, which you can achieve through:

$ pip install -r requirements.txt


We welcome contributions to this library. However, there are a few ground rules contributors should be aware of.


This project is licensed under the Apache v2.0 License. As such, your contributions, once accepted, are automatically covered by this license.

Copyright (c) 2013-2014 Daniele Sluijters

Commit messages

Write decent commit messages. Don’t use swear words and refrain from uninformative commit messages as ‘fixed typo’.

The preferred format of a commit message:

docs/quickstart: Fixed a typo in the Nodes section.

If needed, elaborate further on this commit. Feel free to write a
complete blog post here if that helps us understand what this is
all about.

Fixes #4 and resolves #2.

If you’d like a more elaborate guide on how to write and format your commit messages have a look at this post by Tim Pope.


Commits are expected to contain tests or updates to tests if they add to or modify the current behavior.

The test suite is powered by pytest and requires pytest, pytest-pep8, httpretty and pytest-httpretty which will be installed for you if you run:

$ pip install -r requirements.txt

To run the unit tests (the ones that don’t require a live PuppetDB):

$ py.test -v -m unit

If the tests pass, you’re golden. If not we’ll have to figure out why and fix that. Feel free to ask for help on this.



  • Loosen requirements and drop six


  • Bugfix: Fixed metric() function to query the new v2 endpoint based on Jolokia (
  • Added a new parameter metric_api_version to the BaseAPI() constructor that allows changing the version of the metric API being queried. Valid values are ‘v1’ for PuppetDB <= 6.9.0, ‘v2’ for PuppetDB >= 6.9.1 or None which defaults to ‘v2’.
  • Added a mew parameter version to the metric() function that allows overriding the version of the metric API being queried for that individual call. If nothing is specified it will default to the self.metric_api_version of the class, else it expects a value of ‘v1’ or ‘v2’ same as the metric_api_version class parameter.
  • Added new payload parameter to _query() to allow users to send arbitrary payloads with their queries (useful for debugging).


  • Dropping old python 2.7/3.5 and ensuring 3 latest versions are supported
  • Adding mypy + cleanup + further removal of python2 code
  • Bugfix: Httpretty is not used outside tests and breaks install on newer python versions for some systems


  • Add option to get nodes without using event-counts
  • define the project status as stable
  • bundle requirements-test.txt in python package


  • deduplicate dependencylist
  • Use native data structures for internal representation
  • Add support for the Command API, /pdb/cmd/v1. Added _cmd alongside _query to minimise changes to original code


  • Bump dependencies
  • QueryBuilder: Added support for FromOperator, arrays and FromOperator
  • New endpoint: status
  • POST query in request body
  • Simplify JSON encoding for POST
  • Upload and publish is built in to setuptools


  • Add support for authentication with tokens
  • Fix bug with parsing results from inventory endpoint


  • Fixed noop puppet runs reporting unchanged instead of noop.
  • Fixed unreported nodes shown as ‘noop’ in puppetdb > 4.1.0.
  • Add Inventory API endpoint for PuppetDB 4.2.0.
  • Support for producer field on catalogs, facts and report types.


  • Fixed a datetime related bug in pypuppetdb.api.nodes() that caused all returned nodes to be an unreported status


  • New QueryBuilder module allows users to build PuppetDB queries in an Object-Oriented fashion.
  • Adding support for new fields provided in PuppetDB 4.1.0.


  • Removed deprecation of Expanded resource events data timestamps are not parseable.
  • Escaping additional path parameters passed to _url() with urllib.quote


  • Fixed URL Encoding found when querying the specific value of a macaddress fact.
  • Adding support for PuppetDB 4.0.0 information. Namely Adding a catalog_uuid attribute to the Catalog type object. Adding code_id, catalog_uuid and cached_catalog_status attributes to the Report type object.
  • Removing unneeded sudo option from .travis.yml, this gave unnecessary warning in the test environment.
  • Updating the files under docs/ so can be updated
  • Deprecating in favour of the new events list variable.
  • Renaming test-requirements.txt to requirements.txt


  • Adding a version comparison utility function using examples provided in
  • Adding a new variable latest_report_hash to the Node object. Default None but is given a real value from the field of the same name in the Nodes endpoint available in PuppetDB 3.2 or higher.
  • Allowing support for ‘GET’ AND ‘POST’ requests in the api _query() function. This will allow clients to send requests to the PuppetDB that are too long for a GEt request query string
  • Adding a node field, code_id, to the Catalog object using the field of the same name from the Catalogs endpoint (currently unused as of PuppetDB 3.2.2)
  • Adding test cases for new features EXCEPT the GET and POST update.


  • Version bump to 0.2.0
  • Adding support for v4 of the Query API
  • Removing v2 and v3 api functions as per changelog
  • pypuppetdb will no longer support multiple API versions, removing the api_version attribute from pypuppetdb.connect()
  • All clients must remove the api_version attribute from the connect function, or the starting number, since it is no longer supported
  • Removing all NotImplemented errors in the function of BaseAPI and filled them with the real code

New Features

New endpoints:

  • environments: environments()
  • factsets: factsets()
  • fact-paths: fact_paths()
  • fact-contents: fact_contents()
  • edges: edges()

Changes to Types:

  • pypupperdb.types.Report now requires api to be passed as the second argument, this allows to directly query for any events that occurred in this report object. This functionality was proposed and denied because of backward compatability reasons, since the previous versions are now removed this is no longer a problem.
  • All pypupperdb.types.* accept the v4 API information as optional parameters. These parameters are primarily environment related but may include additional information if provided from that endpoint.
  • Functions appearing inside pypuppetdb.types that run queries against the PuppetDB now accept and passing additional keyword arguments to the query.
  • All pypuppetdb.BaseAPI functions pass any received keyword arguments to the pypuppetdb.api.__init__._query() function. This allows for easy integration with paging functions and parameters.


  • Fix the license in our The license shouldn’t be longer than 200 characters. We were including the full license tripping up tools like bdist_rpm.


Significant changes have been made in this release. The complete v3 API is now supported except for query pagination.

Most changes are backwards compatible except for a change in the SSL configuration. The previous behaviour was buggy and slightly misleading in the names the options took:

  • ssl has been renamed to ssl_verify and now defaults to True.
  • Automatically use HTTPS if ssl_key and ssl_cert are provided.

For additional instructions about getting SSL to work see the Quickstart in the documentation.


Support for API v2 will be dropped in the 0.2.x release series.

New features

The following features are only supported for API v3.

The node() and nodes() function have gained the following options:

  • with_status=False
  • unreported=2

When with_status is set to True an additional query will be made using the events-count endpoint scoped to the latest report. This will result in an additional events and status keys on the node object. status will be either of changed, unchanged or failed depending on if events contains successes or failures or none.

By default unreported is set to 2. This is only in effect when with_status is set to True. It means that if a node hasn’t checked in for two hours it will get a status of unreported instead.

New endpoints:

  • events-count: events_count()
  • aggregate-event-counts: aggregate_event_counts()
  • server-time: server_time()
  • version: current_version()
  • catalog: catalog()

New types:

  • pypuppetdb.types.Catalog
  • pypuppetdb.types.Edge

Changes to types:

  • pypuppetdb.types.Node now has: * status defaulting to None * events defaulting to None * unreported_time defaulting to None


Due to a fairly serious bug 0.0.3 was pulled from PyPi minutes after release.

When a bug was fixed to be able to query for all facts we accidentally introduced a different bug that caused the facts() call on a node to query for all facts because we were resetting the query.

  • Fix a bug where node.facts() was causing us to query all facts because the query to scope our request was being reset.


With the introduction of PuppetDB 1.5 a new API version, v3, was also introduced. In that same release the old /experimental endpoints were removed, meaning that as of PuppetDB 1.5 with the v2 API you can no longer get access to reports or events.

In light of this the support for the experimental endpoints has been completely removed from pypuppetdb. As of this release you can only get to reports and/or events through v3 of the API.

This release includes preliminary support for the v3 API. Everything that could be done with v2 plus the experimental endpoints is now possible on v3. However, more advanced funtionality has not yet been implemented. That will be the focus of the next release.

  • Removed dependency on pytz.
  • Fixed the behaviour of facts() and resources(). We can now correctly query for all facts or resources.
  • Fixed an issue with catalog timestampless nodes.
  • Pass along the timeout option to connect().
  • Added preliminary PuppetDB API v3 support.
  • Removed support for the experimental endpoints.
  • The connect() method defaults to API v3 now.


  • Fix a bug in preventing successful installation.


Initial release. Implements most of the v2 API.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for pypuppetdb, version 2.2.0
Filename, size File type Python version Upload date Hashes
Filename, size pypuppetdb-2.2.0-py2.py3-none-any.whl (35.8 kB) File type Wheel Python version py2.py3 Upload date Hashes View
Filename, size pypuppetdb-2.2.0.tar.gz (45.7 kB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page