Parallelism

Concurrent activities

Limit the number of concurrent activities for this recipe (zero for "unlimited").

Container configuration

Variables & Properties

Recipe variables

This allows you to setup per-recipe overrides of variables. The main use for per-recipe overrides is customization of connection properties
These variables will not be taken into account for the dataiku.get_custom_variables() call.
These variables will not be taken into account for the dkuCustomVariable() call.
JSON syntax error:
{{variablesEditor.variables.dkuJSONError}}

Recipe properties

This allows you to configure advanced Dataiku-specific properties for this recipe. Usually used at the direction of Dataiku Support

Flow dependencies computation

This setting controls how Dataiku checks for changes before doing a smart rebuild of the Flow:
  • Inherit project default: This recipe behaviour is inherited from the setting at the project level.
  • Enabled: Dataiku automatically checks if the input datasets have changed since the last build and skips running this recipe if no change occurred.
  • Enabled (but ignore non-ready sources): Same as Enabled but allows the Flow to proceed even if some input datasets are not ready. This is particularly useful if this recipe has inputs that are expected to not be built, such as when there is a loop in the Flow.
  • Disabled: Dataiku skips checking if the input datasets have changed and are ready. This is useful when computing if the input datasets have changed is costly, such as with large datasets made of many files. However, it means that Dataiku will always run this recipe, even if the input datasets haven't changed.