Following a Rhazes workflow run
Watch a Rhazes workflow run live, read what each step produced, and cancel, retry or download the result when it finishes.
Updated
Watching it live
A run reports progress as it happens rather than only when it finishes. You can watch it move through the steps, and open a step to see the model's output as it is being written rather than waiting for the whole run.
That is worth doing the first few times you run something new. Watching where a workflow spends its time, and what a step actually produced, tells you more about whether it is right than the final output does.
What a finished run keeps
Every run is kept. Opening one shows each step, what it produced, how many attempts it took, and where it stopped if it did.
Runs record whether they were scheduled or started by hand, so an automated run is distinguishable from one you ran yourself.
What the states mean
- Queued. Accepted, waiting to start.
- Running. Working through the steps.
- Waiting for input. Stopped at an Ask step. It resumes when you answer. See Asking the clinician mid-run.
- Interrupted. Stopped before finishing, without completing or failing outright.
- Succeeded. Ran to the end.
- Failed. Stopped on an error, showing the step it stopped on.
- Cancelled. Stopped deliberately.
Cancelling a run
A run that is queued, running or waiting on you can be cancelled. Retrieval already under way is stopped rather than left to finish, so a cancelled run does not carry on quietly pulling records in the background.
Retrying a failed run
A failed or cancelled run can be retried, which starts a new run rather than resuming the old one. Both stay in your history, so you can compare them and see whether the failure was transient or is going to happen every time.
Downloading the result
What a run produced can be downloaded as a branded PDF, a Word document, or plain text. Word is usually what you want if the output is going into another system or needs editing; PDF if it is going into a record as it stands.
When a run fails
Open it and look at the step it stopped on. A failure partway means earlier steps already ran and their side effects already happened, so check what was produced before re-running rather than assuming nothing did.