Skip to main content
Version: 0.16.16

Data Docs Store

A Data Docs Store is a connector to store and retrieve information pertaining to human-readable documentation generated from Great Expectations metadata detailing Expectations, Validation Results, etc.

The Data Docs Store provides an easy way to configure where and how to have your Data DocsHuman readable documentation generated from Great Expectations metadata detailing Expectations, Validation Results, etc. rendered.

Relationship to other objects​

Your Data Docs Store will be used behind the scenes by any Action that updates your Data Docs. Great Expectations includes the UpdateDataDocsAction subclass of the ValidationAction class for this express purpose. Including the UpdateDataDocsAction in the action_list of a CheckpointThe primary means for validating data in a production deployment of Great Expectations. will cause your Data Docs to be updated accordingly.

Use cases​

When you configure your Data Docs, one of the items you will need to include is a data_docs_store_backend for each entry in the data_docs_sites key located in the great_expectations.yml file. This will provide Great Expectations with the necessary information to access the Data Docs being rendered for each given site. That process will generally be handled behind the scenes; after configuring your Data Docs sites to include a Data Docs Store, you will generally not have to directly interact with the Data Docs Store yourself.

When ProfilersGenerates Metrics and candidate Expectations from data. are run to create ExpectationsA verifiable assertion about data. their results are made available through Data Docs. Additionally, most Profiler workflows include a step where the Profiler's generated Expectation SuiteA collection of verifiable assertions about data. is ValidatedThe act of applying an Expectation Suite to a Batch. against a BatchA selection of records from a Data Asset. of data. These processes use a Data Docs Store behind the scenes.

When CheckpointsThe primary means for validating data in a production deployment of Great Expectations. are run to Validate data, their results are generally updated into Data Docs. Any Checkpoint that is configured to do so will use a Data Docs Store behind the scenes. Including the UpdateDataDocsAction in the action_list of a Checkpoint will cause your Data Docs to be updated accordingly. This update will only include the Validation ResultsGenerated when data is Validated against an Expectation or Expectation Suite. from the Checkpoint itself and will not re-render all of your existing Data Docs. If a Checkpoint does not include an UpdateDataDocsAction the Validation Results will not be rendered into Data Docs.

Ease of use​

Once your Data Docs sites are configured with a Data Docs Store backend, the Great Expectations Objects that access Data Docs will handle the rest: you won't have to directly interact with the Data Docs Store.

Access​

Your Data Context and other objects that interact with Data Docs will access your Data Docs Stores behind the scenes without you needing to directly manage them yourself beyond configuring your Data Docs sites during Setup.

In the Validate Data step, including UpdateDataDocsAction in the action_list of a Checkpoint will cause your Data Docs to be updated with the Checkpoint's Validation Results; this process will use your Data Docs Stores behind the scenes.

note
  • To ensure that the Validation Results are included in the updated Data Docs, UpdateDataDocsAction should be present after StoreValidationResultAction in the Checkpoint's action_list.

You can also render your Data Docs outside a Checkpoint by utilizing the build_data_docs() method of your Data Context. This will re-render all of your Data Docs according to the contents of your Data Docs Store.

Configure​

Unlike other Stores, Data Docs stores are configured in the store_backend section under data_doc_sites in the great_expectations.yml config file. For detailed information on how to configure a Data Docs Store for a given backend, please see the corresponding how-to guide: