For pipeline Execute snap…When to select “Reuse Executions to process the documents” and when not to select?
For pipeline Execute snap...When to select "Reuse Executions to process the documents" and when not to select?
Based on what I’ve been told, the Reuse checkbox is if the content of the executions will be the same. Meaning all of values are repeated for each pipeline execution. If they are different, you will get an error back and the pipeline will fail. I’ve experienced this as well from a failure perspective.
One way to think about it is when Reuse is selected, it’s like copying-and-pasting all of the snaps from the child pipeline into the parent pipeline where the Pipeline Execute snap is. When reuse is not selected, the child pipeline will be executed to completion for every document that goes into the Pipeline Execute snap in the parent.
As an example, given a child pipeline that consisted solely of an Aggregate snap and the parent pipeline fed 100 documents into the Pipeline Execute snap, you would see the following:
- With Reuse Enabled: The Aggregate snap would perform the operation on all 100 documents and output a single document that would be sent out the Pipeline Execute.
- With Reuse Disabled: There would be a separate Aggregate snap execution for every document, so the Aggregate would perform its operation on only one document and the Pipeline Execute snap would have 100 output documents. One output for every execution.
As to when to use it or not, it really depends on what the child is doing. As in the above example, if there is a snap in the child that collects documents for processing (e.g. Aggregate, Sort), you’ll need to enable reuse. In contrast, if you need to change the pipeline parameters for the child execution based on the input to the Pipeline Execute, you’ll need to disable reuse since parameters values cannot be changed after the first child has been started.
Can you describe your use case?
no requirement for now. I will get in that soon. Till then I am analyzing the snaps conceptually. I come across this and raised a flag here. BTW thanks for your explanation.