[Bug]: Result CR not showing any details under Specs.Details section

[kodefoundry]

Checklist

• I've searched for similar issues and couldn't find anything matching
• I've included steps to reproduce the behavior

Affected Components

• K8sGPT (CLI)
• K8sGPT Operator

K8sGPT Version

v.0.3.29

Kubernetes Version

v1.29.3+rke2r1

Host OS and its Version

SLES15 SP 4

Steps to reproduce

1. Deploy K8sGPT following the Readme.md
2. Verify scan result using "kubectl describe result"

Manifest
apiVersion: core.k8sgpt.ai/v1alpha1 kind: K8sGPT metadata: name: k8sgpt-sample namespace: k8sgpt spec: ai: enabled: true model: gpt-3.5-turbo backend: openai secret: name: opemai-secret key: openai-api-key # anonymized: false # language: english noCache: false repository: ghcr.io/k8sgpt-ai/k8sgpt version: v0.3.29

Expected behaviour

Specs.Details section should have populated some details that were queried from OpenAI backend.

Actual behaviour

Specs.Detail section is empty
Output from kubectl describe result jarvis -n k8sgpt
Name: jarvis Namespace: k8sgpt Labels: k8sgpts.k8sgpt.ai/backend=openai k8sgpts.k8sgpt.ai/name=k8sgpt-sample k8sgpts.k8sgpt.ai/namespace=k8sgpt Annotations: <none> API Version: core.k8sgpt.ai/v1alpha1 Kind: Result Metadata: Creation Timestamp: 2024-04-03T12:49:47Z Generation: 1 Resource Version: 3369563 UID: 5345ff82-5e7b-4874-9fbe-218e8c6e1762 Spec: Backend: openai Details: Error: Sensitive: Masked: KlkvdlVI Unmasked: jarvis Text: jarvis has condition of type EtcdIsVoter, reason MemberNotLearner: Node is a voting member of the etcd cluster Kind: Node Name: jarvis Parent Object: Status: Lifecycle: historical Events: <none>

Additional Information

Logs from the k8sgpt POD

{"level":"info","ts":1712149239.235297,"caller":"server/log.go:50","msg":"request completed","duration_ms":1009,"method":"/schema.v1.ServerService/Analyze","request":"backend:"openai" anonymize:true language:"english" max_concurrency:10 output:"json"","remote_addr":"...:59408"}
{"level":"info","ts":1712149275.4206443,"caller":"server/log.go:50","msg":"request completed","duration_ms":1007,"method":"/schema.v1.ServerService/Analyze","request":"backend:"openai" anonymize:true language:"english" max_concurrency:10 output:"json"","remote_addr":".**..:49692"}
{"level":"info","ts":1712149311.6114714,"caller":"server/log.go:50","msg":"request completed","duration_ms":1007,"method":"/schema.v1.ServerService/Analyze","request":"backend:"openai" anonymize:true language:"english" max_concurrency:10 output:"json"","remote_addr":"..*.:59432"}

[qdrddr]

Same issue here.

Checklist

• I've searched for similar issues and couldn't find anything matching *
• I've included steps to reproduce the behavior *

Affected Components

• K8sGPT (CLI)
• K8sGPT Operator

K8sGPT Version
v.0.3.29

Kubernetes Version
v1.27.12+rke2r1

Host OS and its Version
Ubuntu 22.04.04

Local-Ai Version:
v2.11.0-aio-cpu

Description

My Result.example.yaml the Spec.Details field does not show details from my local-AI. Also, in the localai pod, the debug log I do not see any HTTP requests when I’m running k8sgpt:v0.3.29 with localai:v2.11.0-aio-cpu & the ggml-gpt4all-j model, both localai & k8sgpt are in the same namespace.

Expected behavior

Spec.Details from the Result CR section should have populated details with responses from Local-AI that were queried to the backend.

Steps to reproduce

I deployed k8sgpt as a helm-chart operator and added this CR:

apiVersion: core.k8sgpt.ai/v1alpha1
kind: K8sGPT
metadata:
  name: k8sgpt-resource
  namespace: k8sgpt-operator-system
spec:
  ai:
    enabled: true
    backend: localai
    model: ggml-gpt4all-j_f5d8f27287d3
    baseUrl: http://local-ai.k8sgpt-operator-system.svc.cluster.local/v1
    anonymized: true
    language: english
  noCache: false
  version: v0.3.29

Actual behavior

In the Result.example.yaml CR manifest, the Spec.Details field is empty.

Additional Information

However, when I curl to localai, it works just fine, and in localai pod, I can see debug with this request and I can see a response from my localai model ggml-gpt4all-j_f5d8f27287d3

curl http://local-ai.k8sgpt-operator-system.svc.cluster.local/v1/chat/completions \
    -H "Content-Type: application/json" \
    -d '{ "model": "ggml-gpt4all-j_f5d8f27287d3", "messages": [{"role": "user", "content": "How are you doing?", "temperature": 0.1}] }'

{"created":1712108557,"object":"chat.completion","id":"c264e40f-98a4-4ff3-9acf-9bad12aebd08","model":"ggml-gpt4all-j_f5d8f27287d3","choices":[{"index":0,"finish_reason":"stop","message":{"role":"assistant","content":"As an AI language model, I am doing well. Thank you for asking!"}}],"usage":{"prompt_tokens":0,"completion_tokens":0,"total_tokens":0}}

My log files are attached:
kube-rbac-proxy.log
k8sgpt.log
k8sgpt-manager.log

[qdrddr]

I just noticed k8sgpt-operator-controller-manager pod logs, from the manager container:

failed to call Analyze RPC: rpc error: code = Unknown desc = failed while calling AI provider localai: error, status code: 520, message: invalid character '<' looking for beginning of value"}

Finished Reconciling k8sGPT with error: failed to call Analyze RPC: rpc error: code = Unavailable desc = error reading from server: EOF

[qdrddr]

I have deployed k8sgpt-operator v0.1.3 and my Result manifests are no longer created