[kubernetes_inventory] connecting to localhost instead of "url" parameter value


### Relevant telegraf.conf:

```toml
[[inputs.kube_inventory]]
  url = ["https://kubernetes.default.svc"]
  tls_ca = "/var/run/secrets/kubernetes.io/serviceaccount/ca.crt"
  namespace = ""
  bearer_token = "/var/run/secrets/kubernetes.io/serviceaccount/token"
  response_timeout = "10s"
  resource_include = ["daemonsets","deployments","statefulsets","persistentvolumes","persistentvolumeclaims","pods"]
  tagexclude = ["node_name","host"]
```

### System info:

1. Telegraf version: `1.18.3`
2. Operating system: CentOS Linux release 7.9.2009 (Core)

### Docker
1. Kubernetes Version: `v1.18.6`
2. Docker version: `19.03.15, build 99e3ed8919`

### Steps to reproduce:

1. test it using `telegraf --test`
```sh
bash-5.0# telegraf --config /telegraf.conf --test
2021-05-24T06:48:19Z I! Starting Telegraf 1.18.3
2021-05-24T06:48:19Z E! [inputs.kube_inventory] Error in plugin: Get "https://localhost/apis/apps/v1/daemonsets": dial tcp 127.0.0.1:443: connect: connection refused
2021-05-24T06:48:19Z E! [inputs.kube_inventory] Error in plugin: Get "https://localhost/apis/apps/v1/deployments": dial tcp 127.0.0.1:443: connect: connection refused
2021-05-24T06:48:19Z E! [inputs.kube_inventory] Error in plugin: Get "https://localhost/api/v1/pods": dial tcp 127.0.0.1:443: connect: connection refused
2021-05-24T06:48:19Z E! [inputs.kube_inventory] Error in plugin: Get "https://localhost/api/v1/persistentvolumes": dial tcp 127.0.0.1:443: connect: connection refused
2021-05-24T06:48:19Z E! [inputs.kube_inventory] Error in plugin: Get "https://localhost/api/v1/persistentvolumeclaims": dial tcp 127.0.0.1:443: connect: connection refused
2021-05-24T06:48:19Z E! [inputs.kube_inventory] Error in plugin: Get "https://localhost/apis/apps/v1/statefulsets": dial tcp 127.0.0.1:443: connect: connection refused
2021-05-24T06:48:19Z E! [telegraf] Error running agent: input plugins recorded 6 errors
```
### Expected behavior:

telegraf should read the value of `url` in `/telegraf.conf` and try connecting to mentioned URL for the Kubernetes API

### Actual behavior:

`kube_inventory` plugin always connecting to https://localhost, despite changing the `url` value in `/telegraf.conf` to  "https://$HOSTIP:6443" or anything else (another hostname or IP)

### Additional info:

1. this works fine in `1.18.2`
2. Not very much sure but is it because in `1.18.3` - Migrate from github.com/ericchiang/k8s to github.com/kubernetes/client-go (#8937) - new k8 client change?
```diff
func newClient(baseURL, namespace, bearerToken string, timeout time.Duration, tlsConfig tls.ClientConfig) (*client, error) {
-                       Server:                baseURL,
+                       ServerName: baseURL,
```

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[kubernetes_inventory] connecting to localhost instead of "url" parameter value #9293

Relevant telegraf.conf:

System info:

Docker

Steps to reproduce:

Expected behavior:

Actual behavior:

Additional info:

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

[kubernetes_inventory] connecting to localhost instead of "url" parameter value #9293

Description

Relevant telegraf.conf:

System info:

Docker

Steps to reproduce:

Expected behavior:

Actual behavior:

Additional info:

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions