Carbon Impact app and the discovery mode for hosts

[marina_pollehn]

Hi all,

I am trying to gain more insights on the carbon impact app in combination with hosts in discovery mode. Do I get all the same functionalities in the carbon app as for full-stack or infra-only hosts?

I have an Azure VM in discovery mode for test purposes and it does appear in the host list of the carbon app and counts towards the calculations. Can I safely assume that no process data is needed and every calculation is performed on host or database level, and hence, there is no added benefit of moving to infra-only or full-stack if I only consider my carbon app metrics/dashboard?

[isidre · Accepted solution]

Yes, you are right.

For Carbon Impact app calculations it uses the data available from any oneAgent modes (Discovery, Infra-only and Full-stack): data center location/region, CPU information, and usage metrics (CPU, Memory, Network and Storage IO)

But in case you need to analyze individual components at software level (processes, services, applications), you will require Full-stack mode running DQL in Notebooks/Dashboards. Same applies to Kubernetes, as it requires also Full-Stack, in other words, Application only monitoring is not supported as environmental impact metrics are not calculated when host is unknown.