Should Impact be cleared on Resolving an Incident

Normann P_ Nielsen _Netic_

We are testing statuspage, and several Incidents are "Resoved", but Impact is not set to "none"

 

This seems to cause a notification to all site admins regularly, as this:

 

Hey Karsten,

Your component API (example) on your page Netic operational status has been set to Major Outage for over 6 hours. If you're still actively pursuing an issue, please disregard this email.

You can change the status of your components by heading to the management dashboard for your account .

 

So:

 

1. Does impact need to be set = None at resolved

2. Site-admin does not need such mails, can it be silenced...

 

BR,

 

Normann 

chare_51fvm8

After some experimentation yesterday, I don't think Impact actually causes those messages.  I created a test incident in our statuspage, set to resolved but didn't adjust the impact to None.  I don't have any emails.  I know that if I create an incident, resolve it and don't update the component status, it will appear like the component is experiencing a problem, even without the incident.  That will generate an email message for sure.

In a resolved incident, it appears the Impact value just gives the title a different color, even after it is resolved.

The first image has an Impact of Major, and the second has an impact of None.  

The way to avoid the emails is to make sure the operational status for the component gets changed back to operational.  I didn't check all of the incident stages, but I know I have seen where moving from one stage to another doesn't automatically update the component operational status.  And yes, it is very possible to have a component set to "partial outage" for 6 hours even without an incident (Incidents are not automatically created or managed.)

For example, this component had a major outage.  This means the component status was changed.  Notice, there was no incident created/associated with this outage.  The outage can be found in the uptime graph as the bar will have a slightly different color.

I have code which tests for API response time and sends automation messages to the component specific email address to change the status.  This means I can report changes to the component even if an outage is not created.  If the component stays in the non-operational status (except when maintenance is happening) for 3 minutes, then my code automatically creates an incident for that component.  If the component returns to service and there is "[AUTO]" in the title, then my code can close the incident.  Postmortems are not always done for these '[AUTO]" incidents. 

As for a way to stop the "log running operational status" emails, I don't think you can stop them.  But, if you are not experiencing an outage and there is a component in a non-operational state, I would think you would want to know so you can update the component status.  Otherwise, it appears on your uptime graph, even if there is no corresponding incident.

Chris