It depends !
It depends on what you intend to do with the CI/CI-Relation data.
If you intend to assign Incidents and Problems to CIs, and thats all you don’t need such a high degree of accuracy of your CI/CI-Relation data. You might be quite happy with discovered only data.
If the goal is that you be able to identify configurations which are not authorized (==> no approval to be there) you have to make sure that you document every change as “authorized state” in your CMS (apply strict configuration control) so that later you are able to compare this state with what has been discovered.
If you want to trigger automation you have to be certain that the content of your CMS is accurate. If not, you might end up in serious troubles e.g. running scripts against instances which do not fit or re-using keys (TCP/IP-Address) of objects in reality not yet decommissioned.
If you want to execute capacity management for example, you should get accurate data from you CMS , otherwise you generate wrong predictions, and you might spend to much money or you are not well prepared to future demand.
So again, the level of configuration control you apply depends on what you want to do with the CIs / CI Relations. BUT over time, you can be sure that you need full control because on the long run you want to be able to rely 100% on your CMS data.
Tags: content accuracy, control