引言
在现代应用程序开发中，Kubernetes（简称K8s）已经成为了容器编排的标准。Minikube作为一个轻量级的本地Kubernetes集群环境，为开发和测试提供了极大的便利。今天，我将分享如何在MacBook上使用k8sGPT扫描Minikube集群，以确保我们的Kubernetes集群设置和应用程序的最佳实践。
$ minikube start --nodes 2 -p multinode-demo
😄 Darwin 14.5 (arm64) 上的 [multinode-demo] minikube v1.33.1
✨ 自动选择 docker 驱动。其他选项：parallels, ssh
📌 使用具有 root 权限的 Docker Desktop 驱动程序
👍 Starting "multinode-demo" primary control-plane node in "multinode-demo" cluster
🚜 Pulling base image v0.0.44 ...
> gcr.io/k8s-minikube/kicbase...: 435.76 MiB / 435.76 MiB 100.00% 6.65 Mi
🔥 Creating docker container (CPUs=2, Memory=7792MB) ...
🌐 找到的网络选项：
▪ HTTP_PROXY=192.168.21.101:7890
❗ You appear to be using a proxy, but your NO_PROXY environment does not include the minikube IP (192.168.49.2).
📘 Please see https://minikube.sigs.k8s.io/docs/handbook/vpn_and_proxy/ for more details
▪ HTTPS_PROXY=192.168.21.101:7890
🐳 正在 Docker 26.1.1 中准备 Kubernetes v1.30.0…
▪ env HTTP_PROXY=192.168.21.101:7890
▪ env HTTPS_PROXY=192.168.21.101:7890
▪ 正在生成证书和密钥...
▪ 正在启动控制平面...
▪ 配置 RBAC 规则 ...
🔗 配置 CNI (Container Networking Interface) ...
🔎 正在验证 Kubernetes 组件...
▪ 正在使用镜像 gcr.io/k8s-minikube/storage-provisioner:v5
🌟 启用插件： default-storageclass, storage-provisioner
👍 Starting "multinode-demo-m02" worker node in "multinode-demo" cluster
🚜 Pulling base image v0.0.44 ...
🔥 Creating docker container (CPUs=2, Memory=7792MB) ...
🌐 找到的网络选项：
▪ NO_PROXY=192.168.49.2
▪ HTTP_PROXY=192.168.21.101:7890
❗ You appear to be using a proxy, but your NO_PROXY environment does not include the minikube IP (192.168.49.3).
📘 Please see https://minikube.sigs.k8s.io/docs/handbook/vpn_and_proxy/ for more details
▪ HTTPS_PROXY=192.168.21.101:7890
🐳 正在 Docker 26.1.1 中准备 Kubernetes v1.30.0…
▪ env HTTP_PROXY=192.168.21.101:7890
▪ env HTTPS_PROXY=192.168.21.101:7890
▪ env NO_PROXY=192.168.49.2
🔎 正在验证 Kubernetes 组件...
🏄 完成！kubectl 现在已配置，默认使用"multinode-demo"集群和"default"命名空间
$ kubectl get node
NAME STATUS ROLES AGE VERSION
multinode-demo Ready control-plane 2m3s v1.30.0
multinode-demo-m02 Ready <none> 103s v1.30.0
$ minikube status -p multinode-demo
multinode-demo
type: Control Plane
host: Running
kubelet: Running
apiserver: Running
kubeconfig: Configured
multinode-demo-m02
type: Worker
host: Running
kubelet: Running
$ minikube addons list
|-----------------------------|--------------------------------|
| ADDON NAME | MAINTAINER |
|-----------------------------|--------------------------------|
| ambassador | 3rd party (Ambassador) |
| auto-pause | minikube |
| cloud-spanner | Google |
| csi-hostpath-driver | Kubernetes |
| dashboard | Kubernetes |
| default-storageclass | Kubernetes |
| efk | 3rd party (Elastic) |
| freshpod | Google |
| gcp-auth | Google |
| gvisor | minikube |
| headlamp | 3rd party (kinvolk.io) |
| helm-tiller | 3rd party (Helm) |
| inaccel | 3rd party (InAccel |
| | [info@inaccel.com]) |
| ingress | Kubernetes |
| ingress-dns | minikube |
| inspektor-gadget | 3rd party |
| | (inspektor-gadget.io) |
| istio | 3rd party (Istio) |
| istio-provisioner | 3rd party (Istio) |
| kong | 3rd party (Kong HQ) |
| kubeflow | 3rd party |
| kubevirt | 3rd party (KubeVirt) |
| logviewer | 3rd party (unknown) |
| metallb | 3rd party (MetalLB) |
| metrics-server | Kubernetes |
| nvidia-device-plugin | 3rd party (NVIDIA) |
| nvidia-driver-installer | 3rd party (Nvidia) |
| nvidia-gpu-device-plugin | 3rd party (Nvidia) |
| olm | 3rd party (Operator Framework) |
| pod-security-policy | 3rd party (unknown) |
| portainer | 3rd party (Portainer.io) |
| registry | minikube |
| registry-aliases | 3rd party (unknown) |
| registry-creds | 3rd party (UPMC Enterprises) |
| storage-provisioner | minikube |
| storage-provisioner-gluster | 3rd party (Gluster) |
| storage-provisioner-rancher | 3rd party (Rancher) |
| volumesnapshots | Kubernetes |
| yakd | 3rd party (marcnuri.com) |
|-----------------------------|--------------------------------|
$ kubectl create deployment hello-minikube1 --image=kicbase/echo-server:1.0
deployment.apps/hello-minikube1 created
$ kubectl expose deployment hello-minikube1 --type=LoadBalancer --port=8080
service/hello-minikube1 exposed
$ kubectl get pod -A
NAMESPACE NAME READY STATUS RESTARTS AGE
default hello-minikube1-67bf99b564-jstcc 0/1 ContainerCreating 0 12s
kube-system coredns-7db6d8ff4d-85s2n 1/1 Running 2 (3m16s ago) 3m43s
kube-system etcd-multinode-demo 1/1 Running 0 3m57s
kube-system kindnet-wtzgz 1/1 Running 0 3m44s
kube-system kindnet-zwhtz 1/1 Running 0 3m39s
kube-system kube-apiserver-multinode-demo 1/1 Running 0 3m57s
kube-system kube-controller-manager-multinode-demo 1/1 Running 0 3m57s
kube-system kube-proxy-8t4fd 1/1 Running 0 3m39s
kube-system kube-proxy-dxhnv 1/1 Running 0 3m44s
kube-system kube-scheduler-multinode-demo 1/1 Running 0 3m57s
kube-system storage-provisioner 1/1 Running 1 (3m32s ago) 3m55s
$ kubectl get pod -A
NAMESPACE NAME READY STATUS RESTARTS AGE
default hello-minikube1-67bf99b564-jstcc 0/1 ContainerCreating 0 16s
kube-system coredns-7db6d8ff4d-85s2n 1/1 Running 2 (3m20s ago) 3m47s
kube-system etcd-multinode-demo 1/1 Running 0 4m1s
kube-system kindnet-wtzgz 1/1 Running 0 3m48s
kube-system kindnet-zwhtz 1/1 Running 0 3m43s
kube-system kube-apiserver-multinode-demo 1/1 Running 0 4m1s
kube-system kube-controller-manager-multinode-demo 1/1 Running 0 4m1s
kube-system kube-proxy-8t4fd 1/1 Running 0 3m43s
kube-system kube-proxy-dxhnv 1/1 Running 0 3m48s
kube-system kube-scheduler-multinode-demo 1/1 Running 0 4m1s
kube-system storage-provisioner 1/1 Running 1 (3m36s ago) 3m59s
$ brew install k8sgpt
==> Downloading https://formulae.brew.sh/api/formula.jws.json
############################################################################################################################################################################################################################################## 100.0%
==> Downloading https://formulae.brew.sh/api/cask.jws.json
############################################################################################################################################################################################################################################## 100.0%
==> Downloading https://ghcr.io/v2/homebrew/core/k8sgpt/manifests/0.3.40
############################################################################################################################################################################################################################################## 100.0%
==> Fetching k8sgpt
==> Downloading https://ghcr.io/v2/homebrew/core/k8sgpt/blobs/sha256:e22d500e85a13ae94bce5be3471eb9c2fc10b343fc335adb9fd6c39a9adfc9bd
############################################################################################################################################################################################################################################## 100.0%
==> Pouring k8sgpt--0.3.40.arm64_sonoma.bottle.tar.gz
🍺 /opt/homebrew/Cellar/k8sgpt/0.3.40: 7 files, 89.3MB
==> Running `brew cleanup k8sgpt`...
Disable this behaviour by setting HOMEBREW_NO_INSTALL_CLEANUP.
Hide these hints with HOMEBREW_NO_ENV_HINTS (see `man brew`).
$ k8sgpt generate
Opening: https://beta.openai.com/account/api-keys to generate a key for openai
Please copy the generated key and run `k8sgpt auth add` to add it to your config file
$ k8sgpt auth add
Warning: backend input is empty, will use the default value: openai
Warning: model input is empty, will use the default value: gpt-3.5-turbo
Enter openai Key: openai added to the AI backend provider list
$ k8sgpt analyze --explain
AI Provider: openai
No problems detected
$ k8sgpt analyze --explain --with-doc
AI Provider: openai
No problems detected
$ k8sgpt analyze --explain --filter=Pod --namespace=default
AI Provider: openai
No problems detected
$ k8sgpt analyze --explain --filter=Service --output=json
{
"provider": "openai",
"errors": null,
"status": "OK",
"problems": 0,
"results": null
}
$ k8sgpt analyze --explain --filter=Service --output=json --anonymize
{
"provider": "openai",
"errors": null,
"status": "OK",
"problems": 0,
"results": null
}
$ k8sgpt filters list
Active:
> PersistentVolumeClaim
> Ingress
> Deployment
> ReplicaSet
> Service
> StatefulSet
> CronJob
> Node
> ValidatingWebhookConfiguration
> MutatingWebhookConfiguration
> Pod
Unused:
> Log
> GatewayClass
> Gateway
> HTTPRoute
> HorizontalPodAutoScaler
> PodDisruptionBudget
> NetworkPolicy
$ k8sgpt filters add Log
Warning: by enabling logs, you will be sending potentially sensitive data to the AI backend.
Filter Log added
$ k8sgpt filters list
Active:
> Service
> Node
> StatefulSet
> CronJob
> Log
> Ingress
> MutatingWebhookConfiguration
> ValidatingWebhookConfiguration
> Pod
> Deployment
> ReplicaSet
> PersistentVolumeClaim
Unused:
> HorizontalPodAutoScaler
> PodDisruptionBudget
> NetworkPolicy
> GatewayClass
> Gateway
> HTTPRoute
$ k8sgpt auth list
Default:
> openai
Active:
> openai
Unused:
> localai
> ollama
> azureopenai
> cohere
> amazonbedrock
> amazonsagemaker
> google
> noopai
> huggingface
> googlevertexai
> oci
> watsonxai
$ minikube stop -p multinode-demo
✋ 正在停止节点 "multinode-demo-m02" ...
🛑 正在通过 SSH 关闭“multinode-demo-m02”…
✋ 正在停止节点 "multinode-demo" ...
🛑 正在通过 SSH 关闭“multinode-demo”…
🛑 2 个节点已停止。
$ docker ps -a
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
19a65b1bbc03 gcr.io/k8s-minikube/kicbase:v0.0.44 "/usr/local/bin/entr…" 45 minutes ago Exited (130) 48 seconds ago multinode-demo-m02
0f2e98799a5a gcr.io/k8s-minikube/kicbase:v0.0.44 "/usr/local/bin/entr…" 45 minutes ago Exited (130) 37 seconds ago multinode-demo
b3c3c3c6fec6 ghcr.io/open-webui/open-webui:main "bash start.sh" 3 months ago Exited (0) 13 days ago open-webui
实践体验
在我的实践中，k8sGPT帮助我发现了一些常见的配置问题，如资源限制未设置、未使用最新的镜像版本等。这些问题可能看起来微不足道，但在生产环境中可能会导致性能下降或安全隐患。
此外，k8sGPT还建议了很多优化措施，例如使用ConfigMap和Secret来管理配置数据，避免将敏感信息直接硬编码到Pod中。通过这些建议，我的Minikube集群变得更加健壮和高效。
总结
通过在MacBook上使用k8sGPT扫描Minikube集群，我们可以快速发现和解决Kubernetes集群中的潜在问题。Minikube提供了一个方便的本地开发环境，而k8sGPT则为集群的健康和优化提供了强有力的支持。希望这篇文章能帮助你在使用Kubernetes时更加得心应手。如果你还没有尝试过k8sGPT，强烈推荐你在你的开发环境中试试它。
参考：
- https://github.com/k8sgpt-ai/k8sgpt
- https://minikube.sigs.k8s.io/docs/tutorials/multi_node/