topic Re: Pass variable values from one task to another in Data Engineering

Pass variable values from one task to another

swzzzsw — Mon, 24 Jan 2022 19:34:29 GMT

I created a Databricks job with multiple tasks. Is there a way to pass variable values from one task to another. For example, if I have tasks A and B as Databricks notebooks. Can I create a variable (e.g. x) in notebook A and later use that value in notebook B?

Re: Pass variable values from one task to another

-werners- — Tue, 25 Jan 2022 12:31:45 GMT

Hi,

check this: notebook workflows

Re: Pass variable values from one task to another

swzzzsw — Tue, 25 Jan 2022 15:17:20 GMT

Thank you, Werner! From my understanding, if I put a notebook in %run command, I won't have it as a separate task in the job. However, I'd also like to keep the ability to see it in the task view for transparency. Is there a way to achieve that?

Re: Pass variable values from one task to another

-werners- — Tue, 25 Jan 2022 15:23:36 GMT

hm I don´t think so as actually the notebooks will run within the same task.

Re: Pass variable values from one task to another

-werners- — Tue, 25 Jan 2022 15:26:43 GMT

you could also consider using an orchestration tool like Data Factory (Azure) or Glue (AWS). there you can inject and use parameters from notebooks.

The job scheduling of databricks also has the possibility to add parameters, but I do not know if you can dynamically determine the input (based on another task).

Re: Pass variable values from one task to another

swzzzsw — Tue, 25 Jan 2022 15:43:45 GMT

Thank you! We're on Azure. Will explore Data Factory!