Cannot analyze GKE environment

[hapham1701]

I'm facing this issue while running command: k8sgpt analyze on Macos arm64 and connecting to GKE.

Error initialising kubernetes client: no Auth Provider found for name "gcp"
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x2 addr=0x0 pc=0x101e6c600]

[thschue]

Hello @hapham1701!

Unfortunately, I could not reproduce your issue. Authenticated on a cluster using

gcloud container clusters get-credentials ******** --zone ********** --project ******** 

and could analyze without problems. Please provide a bit more information on your configuration, e.g., some obfuscated information about your kube context. Thank you!

[hapham1701]

this is full error:
Error initialising kubernetes client: no Auth Provider found for name "gcp"
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x2 addr=0x0 pc=0x103d0c600]

goroutine 1 [running]:
github.com/k8sgpt-ai/k8sgpt/pkg/analyzer.AnalyzePod({0x104311018, 0x14000134010}, 0x14000257bb8, 0x0, {0x10?, 0x103d31816?}, 0x14000257ba0)
/home/runner/work/k8sgpt/k8sgpt/pkg/analyzer/podAnalyzer.go:17 +0x40
github.com/k8sgpt-ai/k8sgpt/pkg/analyzer.RunAnalysis({0x104311018, 0x14000134010}, 0x33?, 0x103d1d9f0?, {0x1043001a8, 0x1400042cb28}, 0x1040e0020?)
/home/runner/work/k8sgpt/k8sgpt/pkg/analyzer/analyzer.go:18 +0x38
github.com/k8sgpt-ai/k8sgpt/cmd/analyze.glob..func1(0x104ead1a0?, {0x103d1bc51?, 0x0?, 0x0?})
/home/runner/work/k8sgpt/k8sgpt/cmd/analyze/analyze.go:80 +0x30c
github.com/spf13/cobra.(*Command).execute(0x104ead1a0, {0x104ef9f68, 0x0, 0x0})
/home/runner/go/pkg/mod/github.com/spf13/cobra@v1.6.1/command.go:920 +0x5ac
github.com/spf13/cobra.(*Command).ExecuteC(0x104eacbe0)
/home/runner/go/pkg/mod/github.com/spf13/cobra@v1.6.1/command.go:1044 +0x340
github.com/spf13/cobra.(*Command).Execute(...)
/home/runner/go/pkg/mod/github.com/spf13/cobra@v1.6.1/command.go:968
github.com/k8sgpt-ai/k8sgpt/cmd.Execute({0x103f40730?, 0x140000021a0?})
/home/runner/work/k8sgpt/k8sgpt/cmd/root.go:36 +0x58
main.main()
/home/runner/work/k8sgpt/k8sgpt/main.go:11 +0x2c

[thschue]

As written before, please add some more information about your configuration.

e.g. kubectl config view (for your current context).

[AlexsJones]

If you could print your ~/.k8sgpt.yaml ( omitting your key for openAI) we could see what's going on.. the auth backend shouldn't be giving that error at all.

[hapham1701]

result : cat ~/.k8sgpt.yaml

backend_type: openai
kubernetesclient: null
openai_key:xxxxxxxxxx

[matthisholleville]

@hapham1701 Do you have gcloud properly installed/configured on your environment ?

[hapham1701]

This is information about gcloud and kubectl, which i have already installed on my laptop. Please feel free to ask me if you need any further information:

gcloud version

Google Cloud SDK 423.0.0
alpha 2023.03.17
beta 2023.03.17
bq 2.0.88
config-connector 1.93.0
core 2023.03.17
gsutil 5.21
Updates are available for some Google Cloud CLI components. To install them,
please run:
$ gcloud components update

To take a quick anonymous survey, run:
$ gcloud survey

---

~ kubectl version
Client Version: version.Info{Major:"1", Minor:"23", GitVersion:"v1.23.0", GoVersion:"go1.17.3", Compiler:"gc", Platform:"darwin/arm64"}

Server Version: version.Info{Major:"1", Minor:"24", GitVersion:"v1.24.9-gke.3200", GoVersion:"go1.18.9b7", Compiler:"gc", Platform:"linux/amd64"}

[hapham1701]

Hi @thschue This is the result of the command: kubectl config view

• name: xxxxxxx
  user:
  auth-provider:
  config:
  access-token: xxxxx
  cmd-args: config config-helper --format=json
  cmd-path: /opt/homebrew/Caskroom/google-cloud-sdk/latest/google-cloud-sdk/bin/gcloud
  expiry: "xxxxxx"
  expiry-key: '{.credential.token_expiry}'
  token-key: '{.credential.access_token}'
  name: gcp

[matthisholleville]

@hapham1701 can you try to run : gcloud components install gke-gcloud-auth-plugin please ?

[matthisholleville]

I have set up a clean environment on Debian, and I had no issues.

gcloud container clusters get-credentials xxx --region xxx --project xxx

Here's my configuration:

apiVersion: v1
clusters:
- cluster:
    certificate-authority-data: DATA+OMITTED
    server: https://xxxxx
  name: gcp
contexts:
- context:
    cluster: gcp
    user: gcp
  name: gcp
current-context: gcp
kind: Config
preferences: {}
users:
- name: gcp
  user:
    exec:
      apiVersion: client.authentication.k8s.io/v1beta1
      args: null
      command: gke-gcloud-auth-plugin
      env: null
      installHint: Install gke-gcloud-auth-plugin for use with kubectl by following
        https://cloud.google.com/blog/products/containers-kubernetes/kubectl-auth-changes-in-gke
      interactiveMode: IfAvailable
      provideClusterInfo: true

Here's my version of gcloud:

root@xxxx:~/.kube# gcloud version
Google Cloud SDK 424.0.0
alpha 2023.03.24
beta 2023.03.24
bq 2.0.89
bundled-python3-unix 3.9.16
core 2023.03.24
gcloud-crc32c 1.0.0
gke-gcloud-auth-plugin 0.5.2
gsutil 5.21

I think you need to use gke-gcloud-auth-plugin as authentication method.

[hapham1701]

Here's my version of gcloud:

~ gcloud version
Google Cloud SDK 423.0.0
alpha 2023.03.17
beta 2023.03.17
bq 2.0.88
config-connector 1.93.0
core 2023.03.17
gke-gcloud-auth-plugin 0.5.2
gsutil 5.21

---

After install gke-gcloud-auth-plugin, The error still exists.

[hapham1701]

It worked after I upgraded ver chatgpt to 0.1.6 and ran the command 'gcloud container clusters get-credentials xxx --region xxx --project xxx'. I don't know why I needed to get credentials even though they already existed.
Thank you for your support.

[umiyosh]

I experienced the same problem, but I can confirm that it is recovered by installing and activating the gke-gcloud-auth-plugin described in the discussion here. It is important to note that the problem will not be solved without installing, activating and getting credentials for the gke-gcloud-auth-plugin, as I found out when I checked the operation. Therefore, to solve the problem, you need to execute all of the following commands.

% gcloud components install gke-gcloud-auth-plugin
-> gke-gcloud-auth-plugin will be installed
% export USE_GKE_GCLOUD_AUTH_PLUGIN=True
-> gke-gcloud-auth-plugin will be activated
% gcloud container clusters get-credentials xxx --region xxx --project xxx
-> Credentials will be obtained by authentication with gke-gcloud-auth-plugin
% k8sgpt analyse
-> You will be able to see that the problem has been resolved.

A change in the kubectl authentication method causes this problem. You can read the details in the following document.
https://cloud.google.com/blog/products/containers-kubernetes/kubectl-auth-changes-in-gke?hl=en