Flow Execution Failure After Assessment Metric Changes – Best Recovery Approach?

[satya30]

Hi All,

I’m looking for guidance on a flow execution issue related to Assessment metrics.

We have a Flow Designer process triggered from HR cases that retrieves survey metric results using metric sys_ids. While several cases were already in progress, the survey form was modified (metrics/questions were updated). This resulted in new metric sys_ids being generated.

When the in-flight flow executions reached the “Look Up Record” step (asmt_metric_result), they failed with a “No record found” error because they were still referencing the old metric sys_ids.

Current situation:

• ~15 flow executions are with the old flow where it has look up to older metric sysid for that particular instance generated in HR Task for HR Case

• few HR tasks were already completed before the failure.

• Remaining tasks did not generate.

• New cases are working correctly after updating the flow logic.

• The failed executions cannot be resumed (no Resume option available).

• Out of 15 flow executions (6 already in error state and remaining are yet to come to error state ) as they are following the older flow where the metric was part of the lookup and that metric is not part of the employee form that is generated during the HR Task instance

Since these are long-running HR lifecycle flows, fully restarting them may duplicate previously created tasks unless additional safeguards are added.

My questions:

1. What is the recommended approach for recovering hard-failed flow contexts where core actions error out?

2. Is restarting the flow (after adding idempotency checks) the correct safe method?

3. Is there any supported way to resume execution from a later step?

Appreciate any guidance or best practice recommendations

Thank you

#ITSM