Telegraf-operator image, how can we apply this to openshift cluster

We are using telegraf operator image for capturing the metrics from prometheus endpoint that is exposed by the OCP cluster. metrics are scraped and we have /metrics endpoint containing the metrics we wanted to monitor

telegraf-operator/deploy/dev.yml at master · influxdata/telegraf-operator · GitHub we are using this in our env, renamed as telegraf-operator.yml, but the content is same.

We are getting this error, how can this be fixed,
MutatingWebHookConfiguration is somehow not allowing to perform action on other deployments irrespective of namespace, in case if telegraf is throwing error.

In one case, when telegraf-operator was failing in our test cluster, we faced issues in deleting one of the deployment in other namespace than telegraf-operator, and error looked like this

•	
[scb@scbbox haproxy-router-logging-2]$ oc delete po/fluentd-749f9dbf9d-dt72t

Error from server (InternalError): Internal error occurred: failed calling webhook "telegraf.influxdata.com": failed to call webhook: Post "https://telegraf-operator-test.telegraf-operator-test.svc:443/mutate-v1-pod?timeout=10s": service "telegraf-operator-test" not found

I have changed failurePolicy in deployment.yaml from ‘Fail’ to “Ignore’ which seems to have fixed it but I want to understand why telegraf is controlling other deployments as well, since this mutatingwebhookconfiguration is not limited to a namespace

Also there is a second question,

• In the official documentation, only method to fetch metrics are via providing pod-level annotations. But to provide annotation to all deployments is not feasible, so how to configure it to fetch from all existing and new deployments irrespective of annotations they are having?
I understand that when we provide influxdata annotation to a new deployment, telegraf creates a sidecar and then metrics are being sent to output defined in ‘telegraf-operator-classes.yaml’ file

apiVersion: v1
kind: Secret
metadata:
  name: telegraf-operator-classes
  namespace: telegraf-operator
stringData:
  basic: |+
    [[outputs.file]]
      files = ["stdout"]
    [[outputs.kafka]]
      brokers = ["abc.com:9101"]
      topic = "abctopic"
      sasl_username = "abc"
      sasl_password = "XXXXXXXX"
      sasl_mechanism = "SCRAM-SHA-512"

Can some one help on this?

Hi Community,

Can some one help on the query?

Hi Community,

Can some one help on this query or suggestions?

Hello @venudoddi,
Sorry for the delay.
This is a deeper telegraf question and I don’t know the answer.
I’m hoping that maybe @jpowers does though.

Unfortunately, both questions are related to the telegraf-operator project and some K8s usage, both of which I know very little about.

You might be better off asking each of these in the repo for telegraf-operator itself and see if any users directly comment on it there.

Sorry I cannot be of more help.