Creating an incident via API doesn't affect component downtime?

Question

The API to create an incident has a section where you can define the components affected by the incident, and their individual states.

When I create an incident using this API, any component mentioned in the API data is correctly updated in StatusPage to reflect the new state.

However, when I mark the incident as resolved and mark the components as operational, the uptime for the components is *not* affected.

Is this a bug or by design? If the latter, I'm struggling to understand why. It seems like I've got to mark each individual component as (say) "major_outage" and (ultimately) "operational", even though I've done all that via the incident!

Mark Campbell-Vincent_ghir58 · Answer

@Philip Colmer -

Glad you were able to find the root cause! Let me know if I can help with anything else.

Philip Colmer · Answer

@Mark Campbell-Vincent

I was correct - the root cause was the fact that I was updating the affected components after the incident was initially created. Changing the steps in the xMatters workflow so that it is now a single call to create the incident with the affected components results in StatusPage displaying a more logical view:

Philip Colmer · Answer

@Mark Campbell-Vincent

The only thing that still seems a bit "off" about how the incident is being reported is the "inside" view:

Note how it says "No components were marked as affected by this incident"?

I *think* that this is being caused by how the incident is being created. I'm using xMatters to update StatusPage and the built-in step to create a StatusPage incident doesn't support specifying the component(s) at the time of *creating* the incident, so I've written a custom step that is run straight afterwards that updates the incident with the affected components.

I'll try replacing the built-in step with a modified one that does allow the components to be specified at the creation time.