In this article
Overview
Cache Pipelines enable you to run pipelines that reference information from a main pipeline.
Prerequisites
Cache pipelines must have the following input and output views:
0 unconnected input views (Binary or Document)
1 unconnected Document output views
0 unconnected Binary output views
The output document must have a flattened document structure. None of the values for primary keys can have objects or arrays as part of their structure.
If a document contains any value that is an array or object, it fails to process, which causes the pipeline to fail. If a cache pipeline fails to run, it will result in the preparation of the main pipeline failing (because it specifies the cache pipelines).
Design the Main Pipeline
Main pipelines reference the information in the cache pipeline. You can reference multiple main pipelines.
In the Designer tab:
Click your user name in the top banner.
In the dropdown list, click User Settings.
In the Settings dialog, under Preview Document Count, set the number of records (1, 2, 10. 50,100, 500, 1000, 11500, 2000). This value is the same for the number of records allowed for the expression language lookup during validation.
Create or import a pipeline onto the Designer canvas.
Click to open the Edit Pipeline dialog.
Fill out the form:
Label. Enter a name for your main pipeline. We recommend that you choose a label that’s easily recognizable as the pipeline.
Project. Confirm that the pipeline is the correct project. If you need to change the Project, refer to Moving a Pipeline.
Resumable Mode. Leave deselected.
Error Pipeline. Leave blank.
Error Pipeline Parameters: Leave empty
Parameters: Enter Pipeline Parameters
Capture: Select to proliferate parameters downstream.
Expression Libraries: Add libraries as needed.
In Cache Pipelines properties, click to enter the alias and the pipeline to run for the data stored in that alias. This is the first argument in the Cache Pipelines expression language functions as a reference.
Click Save.
Build Cache Pipelines
In Designer, create a new pipeline that will generate data to be referenced in the main pipeline.
Use any Snaps you need to get the data to cache, ensuring the following:
The final output document needs to have a flattened structure.
The pipeline does not have an open input view.
Optionally, end with a Mapper Snap to restrict the data in your cache or apply transformations for different needs as desired.
Click to run pipeline validation.
Click to open the Data Preview to verify the data that is stored in the cache.