PuppetDB 4.0: Query structure
Included in Puppet Enterprise 2016.1. A newer version is available; see the version menu above for details.
PuppetDB’s query API can retrieve data objects from PuppetDB for use in other applications. For example, the PuppetDB-termini for Puppet masters use this API to collect exported resources.
The query API is implemented as HTTP URLs on the PuppetDB server. By default, it can only be accessed over the network via host-verified HTTPS; see the jetty settings if you need to access the API over unencrypted HTTP.
A query consists of an HTTP GET request to an endpoint URL which may or may not contain:
queryURL parameter, whose value is a query string.
- Other URL parameters, to configure paging or other behavior.
That is, most queries will look like a GET request to a URL that resembles the following:
Alternatively, you can provide the entity context instead of using the
<ENDPOINT> suffix with the following:
Consult the root endpoint documentation for more details.
API URLs generally look like this:
/pdb/query/ prefix, the first part of an API URL is the
API version, written in the
v4 format. This section describes version
4 of the API, so every URL will begin with
After the version, URLs are organized into a number of endpoints that express the entity you wish to query for.
Conceptually, an entity endpoint represents a PuppetDB entity. Each version of the PuppetDB API defines a set number of endpoints.
See the entities documentation for a list of the available endpoints. Each endpoint may have additional sub-endpoints under it; these are generally just shortcuts for the most common types of query, so that you can write terser and simpler query strings.
Finally, the URL may include some URL parameters. Some endpoints require certain parameters; for others they’re optional or disallowed. Each endpoint’s page lists the parameters it accepts, and most endpoints also support the paging parameters.
A group of parameters begins with a question mark (
?). Each parameter is formatted as
<PARAMETER>=<VALUE>, and additional parameters are separated by ampersands (
&). All parameter values must be URL-encoded.
The most common URL parameter is
query, which lets you define the set of results returned by most endpoints.
There are two query languages available in PuppetDB, consult the documentation for each for more details.
- AST query language: a JSON based query language.
- [Puppet query language (experimental)][pql]: a new query language designed for human users to simplify querying over the legacy AST language.
A complete query string describes a comparison operation. When submitting a query, PuppetDB will check every possible result from the endpoint to see if it matches the comparison from the query string, and will only return those objects that match.
Note: Only the root endpoint supports PQL.
The next most common URL parameters are the paging parameters.
Most PuppetDB query endpoints support paged results via a set of shared URL parameters. For more information, please see the documentation on paging.
All queries return data with a content type of
application/json. Each endpoint’s page describes the format of its return data.