[Datadog Operator Pod] Proxy support

[rhallier]

Hi,
In the Datadog Operator pod, I got some errors such as :
{"level":"ERROR","ts":"2024-11-28T17:02:28.156Z","logger":"DatadogMetricForwarders","msg":"cannot retrieve Datadog metrics forwarder to send deployment metrics, will retry later...","CustomResource.Namespace":"datadog","CustomResource.Name":"datadog","error":"cannot validate datadog credentials: Get \"https://api.datadoghq.eu/api/v1/validate\": dial tcp 34.107.236.155:443: i/o timeout"}

It's because the k8s infra is behind a proxy. Except the pod operator itself, all the other pods (ClusterAgent, NodeAgent, ..) are correctly configured to use a proxy by adding env var DD_PROXY_HTTPS
I tried to add the same env var for the operator pod but without success
Any tips?

[khewonc]

Hi @rhallier, could you try using HTTPS_PROXY and HTTP_PROXY in the operator deployment? Env vars can be added using env if you're using the public helm chart for the operator

[Sillychilly91]

How would we pass those to the operator? Would it be via an override the same as the nodeagent? Like the below:

spec: override: operator: env: - name: HTTPS_PROXY value: <Proxy address>

[levan-m]

Hello @Sillychilly91, the HTTP_PROXY/HTTPS_PROXY env var should be added to the Operator container using the env config.
These env vars aren't specific to Operator, they are used by underlying Go and k8s client libraries. See this and this for details.