What problem are you trying to solve?

Focus on the what and why of the need you have, not the how you'd like it solved.

In atlas metrics that have been sent to datadog, I am able to group by projectname, but NOT on clustername for search nodes. Search nodes always show up as having a N/A value for clustername.

You can see this with a query like:

avg:mongodb.atlas.system.cpu.norm.user{projectname:<project> ,host:*search*} by {clustername,projectname}

What would you like to see happen?

Describe the desired outcome or enhancement.

clustername to be populated for search nodes by the atlas->datadog integration.

Why is this important to you or your team?

Explain how the request adds value or solves a business need.

We typically have monitors set up to do a alerts once per cluster (but not per host), in order to not get bombed by a ton of individual per-host monitors. However without clustername as a tag, it is impossible to tell which cluster is firing, without looking at the host name itself.

What steps, if any, are you taking today to manage this problem?

We work around this by templating our search node CPU monitors by hostname filters. That is, since the cluster name is present in the host name itself, we will have a monitor per cluster, like:

avg:mongodb.atlas.system.cpu.norm.user{host:*search*, host:*cluster1name*} by {clustername,projectname}

and then separately another with:

avg:mongodb.atlas.system.cpu.norm.user{host:*search*, host:*cluster2name*} by {clustername,projectname}

But ideally we could use the normal datadog features to have all of this on the same monitor!