Downtime Classification - Post Incident

I host a SaaS application for many different clients. Down and Up are just not enough.

I need the ability to accurately report on my uptime, and this also includes accurately reporting downtime. I need to be able to answer the questions from clients like “How much time have you taken the application offline for maintenance in the last year” or “How are your metrics against our SLA?”

I also need to address “What about when the site was broken 2 months ago for 2 hours” when it was the customer who changed settings on their own site and broke it… not my issue.

Our contracts allow for maintenance “off hours” and our SLA is typically 6am - 8pm local time, M-F.

What I’d like to see:

Customers should have (at least) 5 types of downtime classification.

Customers should be able to give these a custom name.

I might name them something like

• 1 - Application Issue

• 2 - Scheduled Maintenance

• 3 - Emergency Maintenance

• 4 - Provider Issue (Azure / AWS)

• 5 - Client Issue

All downtime defaults to Type 1

After an incident, customers can reassign the downtime classification for that incident. So, if the site was down for maintenance, I can change that downtime to Type 2 - Scheduled maintenance. I do not want to stop monitoring during maintenance, because I need to accurately account for the real-world time the site was down. A maintenance window might be 4 hours, but the actual downtime could be 15 minutes. I cant stop monitoring for 4 hours and not capture any downtime. The site WAS down.

Reporting should allow me to run a report and to choose which of the 5 types of downtime are ACTUALLY downtime for that purpose. So, when I run an SLA report for a client, I would not count Type 2, Type 3 (Maintenance) or Type 5 (Client Issue) as part of my SLA report. These do not count against my SLA.

I can also run a report to show only Type 2 and Type 3 (Maintenance) - Maintenance is excused, however I can’t be performing “maintenance” all the time either…. so with this I can run a maintenance only report and show that we had XX hours of maintenance for the year. This is also a report management at my company would like to see regularly.

I also need to keep an accurate record of outages caused by Azure / AWS and also outages that were the client’s fault. This is important information to have from an uptime monitoring tool.