Skip to content
Learning Outcomes
  • Run a single process without building a pipeline around it first.
  • Know what the Run button does, and what it does not do.

Run a process on its own

Testing one process used to mean building a pipeline around it by hand: create a pipeline, drop the process in, wire it up, then run that. For a quick check of a single step, most of that work had nothing to do with what you were actually trying to find out.

A process can now be run on its own.

How to do it

  1. Open the process in the process editor.
  2. Click Run.

That creates a run attempt for just that process and takes you to the ordinary run detail page.

From there it behaves like any other run. You choose a run environment and an execution directory, provide the inputs, and launch it. Logs, outputs and reports all appear exactly where you expect, because it is an ordinary run: see Run Page.

What it is good for

  • Checking that a process works before committing it to a pipeline.
  • Re-running one step against different inputs without disturbing a pipeline.
  • Reading the logs of a single step without hunting for it inside a larger run.

Worth knowing

  • The process is not modified. Running it this way does not change the process or create a revision of it.
  • It is a real run. It consumes the same resources on the same run environment as any other run, and it appears in your run lists alongside the rest.
  • Inputs still apply. A process that requires inputs still requires them here; the Run button creates the attempt, it does not guess what to feed it.

See also