Incorect handling "inProgress" cases

Version 2019.4.0-CE.22.

How I achieve the effect as on the screenshot:

  1. Kill the job during the processing (transaction goes to “inProgress” → OK)
  2. Run the same job on the same robot:
    • Robot takes an inProgress transaction and… without any processing I see third row: “Action” = “Edit”, “Status” = “Successfull”. It is imo incorrect.
    • after that robot takes other transactions in the “New” state.

It is not deterministic. Sometimes robot edited the status from “inProgress” to “Successful”, sometimes leave cases in “inProgress”. I don’t know the reason for each scenario.

Is it some known bug?

Hi @koala

It is actually quite simple. After you start on a transaction, you need to set it to either Successful or Failed within the process. If you don’t, it will be stuck at In Progress status and you won’t be able to do anything with it until it is abandoned within a default timespan for your Orchestrator (for Orchestrator CE it is 24h).

Can you track your process to make sure that even the transactions that were stopped (due to exceptions) are getting marked as Failed instead of letting the process finish without handling that change of status? That should fix the issue.

The problem is that it has NOT been blocked in inProgress. Case should stay in inProgress state.

Second problem is that the same process (the same version) sometimes leave cases in the inProgress state after kill and sometimes “edit” the status to the Success.

Case should be in inProgress state when we kill this process from the Orchestrator.

It is hard to say without specifics. My first thought would be that there is a condition in the process that simply skips some transactions (this way you would get this behaviour of an instant edit with the Successful status).

As to the other thing, are you by chance using the ReFramework?

I am encountering the same problem, and i am using REF. Has this been resolved or is there more information on it somewhere ?