Bug #8023
closed
command_exection_result defaulting to 'error' state is not documented
Added by Florian Heigl over 8 years ago.
Updated over 2 years ago.
Description
This class can define a repair and a kept outcome.
It cannot define an error.
This is kind of, umm, problematic.
I have failing commands that are interpreted as happy commands.
(If possible, make it so that one is able to configure it so that any unexpected state means an error.
Doesn't need to be default, but that was the one big PITA from start with the normal scheduler.
We use script wrappers for this now.)
I will assume that you didn't mean scheduler here, but just command_exection_result
There is no error state in the command_exection_result genric method since any state not in success or repaired is considered as an error.
The acceptance test for this methods says it is the case.
Do you want tro have e default other than error or have I just missed something that happen to be your case ?
Hi,
then it should be all as I wanted it. I could not see the text saying it's defaulting to error. might be fine print ;)
I'll re-test tomorrow and close if it was just PBKAC
Ah, yeah its really just in the method documentation. Not obvious for mr. clickyclicky ;-)
- Subject changed from Scheduler command_exection_result is missing 'error' state to command_exection_result defaulting to 'error' state is not documented
- Target version set to 0.x
Sorry, changing the subject to patch documentation then
- Status changed from New to In progress
- Assignee set to Alexis Mousset
- Status changed from In progress to Rejected
Handled in #7879. Closing this one.
- Is duplicate of User story #7879: Add documentation for ncf generic methods added
- Target version changed from 0.x to ncf-0.x
- Priority set to 0
- Project changed from 41 to Rudder
- Category set to Generic methods
Also available in: Atom
PDF
Updated by 
Updated by 
Updated by