When you have an active incident with a task and you choose a incident model which does not have a task and you save the incident then you can't move the incident to another phase anymore. This leaves a bugged incident as you can't close it anymore, and you need to close it with a REST call. This has been tested by us and support. Information from support:
"We found the cause as of why a Task on the Incident disappears and causing the Incident to be unable to move to next phase. After you choose an incident model for an incident, the task in the incident will be replaced by the task defined in the model. The 2 OOTB models has no task defined, and when we choose the OOTB model, the task we defined for our incident will be replaced by the 'null' task from the model.
This has been stated as a working as designed behavior. If you need to choose a model, then you’d need to define the task in the classify phase for the model itself."
I personally think that the design is not correct, and I think it needs an adjustment. Examples for a fix:
- When you have an incident with a task and you choose an incident model that has no task defined, then leave the current task and don't remove/bug it
- When you have an incident with a task and you choose an incident model that has no task defined, then remove the current task and don't bug it.
Tested and confirmed this is bugged in 2020.05.002