KuberTENes Birthday Bash Guadalajara - K8sGPT first impressions
Download as PPTX, PDF0 likes79 views
K8sGPT is a tool that analyzes and diagnoses Kubernetes clusters. This presentation was used to share the requirements and dependencies to deploy K8sGPT in a local environment.
![LocalAI ¨C
Testing Service
$ kubectl exec -ti -n local-ai-system local-ai-77785cbc9-z6jq2 -- curl http://localhost:8080/v1/chat/completions -
H "Content-Type:application/json" -d '{ "model": "gpt-4", "messages": [{"role": "user","content": "How are you
doing?", "temperature": 0.1}] }'
Defaulted container "local-ai" out of: local-ai, download-model (init)
{"created":1717624727,"object":"chat.completion","id":"3b7e0870-764d-407d-9aab-
bffa15f790f6","model":"gpt-
4","choices":[{"index":0,"finish_reason":"stop","message":{"role":"assistant","content":"I'm doing well, thank
you for asking. How can I help you today?"}}],"usage":{"prompt_tokens":13,"completion_tokens":18,"](https://image.slidesharecdn.com/k8sgptfirstimpressions-240607175030-814f78cb/85/KuberTENes-Birthday-Bash-Guadalajara-K8sGPT-first-impressions-9-320.jpg)
More Related Content
Similar to KuberTENes Birthday Bash Guadalajara - K8sGPT first impressions (20)
More from Victor Morales (20)
Recently uploaded (20)
KuberTENes Birthday Bash Guadalajara - K8sGPT first impressions
- 1. k8sgpt first impressions Victor Morales
- 2. $ whoami ? +19 yrs as a Software Engineer ? .NET, Java, python, Go programmer ? OpenStack, OPNFV, ONAP, Matter, Nephio and CNCF contributor. https://about.me/electrocucaracha
- 3. Agenda ? Requirements ? Deployment steps ? Demo ? Q & A
- 4. k8sGPT It is a tool for scanning Kubernetes clusters, diagnosing and triaging issues in simple English. Requirements: ? Hardware: ? 8 vCPUs ? 16 GB ? 2 persistent local volumes (5 GB, 10 GB) ? Software: ? Kubernetes ? Helm 3 ? AI Backend ¨C OpenAI, AzureAI, LocalAI
- 6. LocalAI ¨C values.yaml https://github.com/go-skynet/helm-charts/blob/local-ai-3.2.0/charts/local-ai/values.yaml hipblas, sycl, nvidia-cuda, gpu-intel, cublas Image size 2.7 ¨C 21 GB The list of URLs to download models from
- 7. LocalAI ¨C Download Models https://github.com/go-skynet/helm-charts/blob/local-ai-3.2.0/charts/local-ai/templates/deployment.yaml#L129-L172
- 8. LocalAI ¨C Preloading Models
- 9. LocalAI ¨C Testing Service $ kubectl exec -ti -n local-ai-system local-ai-77785cbc9-z6jq2 -- curl http://localhost:8080/v1/chat/completions - H "Content-Type:application/json" -d '{ "model": "gpt-4", "messages": [{"role": "user","content": "How are you doing?", "temperature": 0.1}] }' Defaulted container "local-ai" out of: local-ai, download-model (init) {"created":1717624727,"object":"chat.completion","id":"3b7e0870-764d-407d-9aab- bffa15f790f6","model":"gpt- 4","choices":[{"index":0,"finish_reason":"stop","message":{"role":"assistant","content":"I'm doing well, thank you for asking. How can I help you today?"}}],"usage":{"prompt_tokens":13,"completion_tokens":18,"
- 10. k8sgpt ¨C Operator Deployment https://docs.k8sgpt.ai/getting-started/in-cluster-operator/ https://github.com/k8sgpt-ai/k8sgpt/issues/1107
- 11. Demo Deploy a non-existing Docker image
- 12. Workflow
- 13. $ kubectl describe results.core.k8sgpt.ai -n k8sgpt-operator-system defaultbrokenpod Name: defaultbrokenpod Namespace: k8sgpt-operator-system Labels: k8sgpts.k8sgpt.ai/backend=localai k8sgpts.k8sgpt.ai/name=k8sgpt-local k8sgpts.k8sgpt.ai/namespace=k8sgpt-operator-system Annotations: <none> API Version: core.k8sgpt.ai/v1alpha1 Kind: Result Metadata: Creation Timestamp: 2024-06-05T22:34:13Z Generation: 1 Resource Version: 16213 UID: 95f20ff9-babc-4249-a503-870a5f715409 Spec: Backend: localai Details: Error: Failed to pull and unpack image due to unresolved reference and not found Solution: 1. Check your internet connection and make sure Docker can access Docker Hub. 2. Verify the image name is correct and available on Docker Hub. 3. Try using the full image name including the repository, e.g., "nginx:latest" instead of just "nginx". 4. If the issue persists, consider using a different image or updating the image to the latest version. Error: Text: rpc error: code = NotFound desc = failed to pull and unpack image "docker.io/library/nginx:1.a.b.c": failed to resolve reference "docker.io/library/nginx:1.a.b.c": docker.io/library/nginx:1.a.b.c: not found Kind: Pod Name: default/broken-pod Parent Object: Status: Events: <none> Results resource
- 14. Q&A