aims to provide a new way of deploying the Datadog Agent
but it doesn't allow to configure and manage resources like dashboards or monitors. That's where this operator comes in.
make run
will run the controller against the current cluster in the kube context. If this is your first time executing
the controller, make sure to run make install
before.
The controller relies on the datadog api key and the datadog application key. You can either set the environment
variables DATADOG_API_KEY
and DATADOG_APPLICATION_KEY
, or create a file like the following:
{
"api_key": "<your-api-key>",
"application_key": "<your-application-key>"
}
The default location for this file is .secrets.json
, but can be overwritten with the controller flag --secrets-path
.
Run make deploy
to deploy the controller on the current cluster in the kube context.
Run the TestAPIs
tes in the controllers/datadog/suite_test.go
file with a debugger of
your choice (like IDEs or delv).
If you set a breakpoint somewhere and want to examine what the API server sees at this point in time,
you will need to connect to the local API server running on your local machine.
This API server was started by the testEnv.Start() function in the suite_test.go file. TestAPIs
will automatically writes a kubectl file at ./controllers/cluster/kubectl, you can access the local
API server during a debugging session using this kubectl file:
./controllers/cluster/kubectl get ns