Service Availability

[rkmullis]

I'm curious how other ServiceNow customers are measuring service availability, taking into consideration not only application availability but network availability, as well. The cmdb_ci_outage table does an OK job of capturing outage durations, but it's solely (?) dependent on someone actually creating the record and documenting an outage. Any ideas / options we might consider?

[Jace Benson]

The best way I've seen this is to manage it on a table like problem.

Here were the issues and how it was set up;
Added a field on the outage table referencing task. Added a field on problem that was a watchlist type referencing configuration items (not just business services, even though only business services show up on the availability home page). Added another field that indicated the type of outage if any applied to the configuration items, either "Down", "Degraded", or "Up". Added a field on cmdb_ci referencing business services, so who ever can pick whatever ci and the appropriator outages are recorded and available on the outages home page.

At this point all the cosmetic stuff is done, you just need to figure out how your business wants to handle these now.

The hardest part here I thought had to do with identifying changes to the watchlist. Say you have a problem that is making x, y, z services down. Someone updates it and add's a, and remove z. You should already have outages for x, y, and z. When that person removed z, you need to know so you can delete the outage (I assume z was added accidentally in these cases), and you need to create a new outage for a.

Hopefully this helps.