Handling location data for servers in CMDB

KB15
Giga Guru

What's the best way to handle locations for servers? We have SCCM integration currently and I'm aware that for users who are logged in the most get assigned an asset in both SCCM and ServiceNow. Should that be turned off for servers specifically so that any manually assigned locations are not affected?

Is there a better way to handle location information on the asset side that's more automatic?

2 REPLIES 2

Ashutosh Munot1
Kilo Patron
Kilo Patron

HI,

 

See what usually happens is what ever SCCM supply us then we update it according to our use cases in integration.

But if you want to do modification in this then you may have to either right a BR or you will have to handle it in OnBefore Transform script.


Here you have to check if the CI or asset which we are updating is server then don't update the location of this asset. Now you will need identifier to classify if this is a server or other machine.

In you case we create a custom field which says that this is a server or functional system and not a user workstation.


Thanks,
Ashutosh Munot

Daniel Underwoo
Tera Contributor

We use Discovery to set the location for all CI's that are installed at a fixed location. We have broken down our discovery schedules by subnets in a location, so all CI's found by that schedule are assigned to a specific location. The downside of this approach is an exploding number of schedules to maintain (we have over 600 discovery schedules). We also work closely with our network and corporate real estate teams to keep discovery in sync with any subnet changes and new locations.

As we mature our location data in ServiceNow, this approach may become harder to maintain. Currently we go down to the physical address level with our location data and our larger buildings are further segmented by floor. As we start expanding our use of ServiceNow beyond ITSM and into facilities management and physical access control, we will need much more detailed location data down to the room or cube level. That would obviously not be management using our discovery schedule approach.