Extend DD_APM_IGNORE_RESSOURCES to support Pod-level configuration

[stoufix]

Currently the DD_APM_IGNORE_RESSOURCES enviroment variable in DD agent is configured globaly at the node-level , meaning it applieds to all pods running on a node , this is insufficient for enviroments where different applications require distinct ressource filtring rules .

Is it possible to extend this mechanism to support pod-level configuration? In addition, the existing node-level configuration would allow defining resource exclusions directly in pod specifications.

Thanks

[ichinaski]

Hi @stoufix,

What is the Agent installation and configuration method you are using? The agent does support pod-level configuration, or at the container level in your deployment.

To better understand your issue, please report more information on how you are deploying the Agent, and how the configuration is applied.

[stoufix]

Hi @ichinaski ,
It's a single daemonSet that deploys identical agents across all nodes in the same cluster , since they are a part of DaemonSet so they share the same configuration .
We want to add an annotation to the applicative pods to override filtring APM traces .