Fluentd-kubernetes-daemonset: [error]: config error file="/fluentd/etc/fluent.conf" error="Invalid Kubernetes API v1 endpoint https://172.28.0.1:443/api: SSL_connect returned=1 errno=0 state=SSLv3 read server certificate B: certificate verify failed"

[Jacky2]

Why did I create fluentd startup times wrong?

[Jacky2]

I have solved it.

add env into your yaml

    env:
      - name: KUBERNETES_SERVICE_HOST
        value: "192.168.251.187"
      - name: KUBERNETES_SERVICE_PORT
        value: "6443"

[TanTan-TT]

I have solved it.

add env into your yaml

    env:
      - name: KUBERNETES_SERVICE_HOST
        value: "192.168.251.187"
      - name: KUBERNETES_SERVICE_PORT
        value: "6443"

Hi, I have this problem too, but it does not work by using your solution. Do you have any other config? Or could you share your yaml?

[heidemn]

@TanTan-TT we had a similar problem with the example fluentd-daemonset-elasticsearch-rbac.yaml from this repo.
Maybe it helps in your case.

The example DaemonSet YAML worked out of the box, but after transforming it to Terraform code, it no longer worked.
The problem was that the token for the fluentd ServiceAccount was not mounted into the Pods.

Using the Terraform Kubernetes provider, kubernetes_service_account.automount_service_account_token was false by default.
Using kubectl, the same property was true by default.

Setting the property to true solved the problem.

[zanoubia]

kubernetes_service_account.automount_service_account_token

Same problem, translating fluentd-daemonset-elasticsearch-rbac.yaml into terraform, but your fix didn't work for me :(

[zanoubia]

Worked after all :)
I added automount_service_account_token = true in kubernetes_daemonset.spec.spec

[heidemn]

@zanoubia Thanks - suddenly, this additional one was also required for us.
Probably a Terraform provider version change.

For the record, I think you meant kubernetes_daemonset.spec.template.spec ("template" was missing).