在Kubernetes部署Blazor应用遇CrashLoopBackOff错误求助

我尝试用Minikube（hyperv驱动）在本地Kubernetes环境部署ASP.NET v6 Blazor Server应用，Pod始终处于CrashLoopBackOff状态，且应用无日志输出。相关文件及命令输出如下：

Dockerfile（基于VS2022默认模板） #

#See https://aka.ms/customizecontainer to learn how to customize your debug container and how Visual Studio uses this Dockerfile to build your images for faster debugging.

FROM mcr.microsoft.com/dotnet/aspnet:6.0 AS base
WORKDIR /app
EXPOSE 80
EXPOSE 443

FROM mcr.microsoft.com/dotnet/sdk:6.0 AS build
WORKDIR /src
COPY ["project.csproj", "."]
RUN dotnet restore "./project.csproj"
COPY . .
WORKDIR "/src/."
RUN dotnet build "project.csproj" -c Release -o /app/build

FROM build AS publish
RUN dotnet publish "project.csproj" -c Release -o /app/publish /p:UseAppHost=false

FROM base AS final
WORKDIR /app
COPY --from=publish /app/publish .
ENTRYPOINT ["dotnet", "project.dll"]

Helm配置文件 #

Chart.yml #

apiVersion: v2
name: project
description: A Helm chart for Kubernetes
type: application
version: 0.1.6
appVersion: "0.2.0"

dependencies:
# - name: postgresql
#   version: "13.4.2"
#   repository: "@bitnami"
# - name: pgadmin4
#   version: "1.23.1"
#   repository: "@runix"

templates/deployment.yml #

apiVersion: apps/v1
kind: Deployment
metadata:
  name: project-deployment
  labels:
    app: project-deployment
spec:
  # replicas: 1
  selector:
    matchLabels:
      app: project-deployment
  template:
    metadata:
      labels:
        app: project-deployment
    spec:
      imagePullSecrets:
      - name: regcred
      containers:
      - name: project-pod
        image: registry/project/app:0.2.3
        # resources:
        #   requests:
        #     cpu: "100m"
        #     memory: "256Mi"
        #   limits:
        #     cpu: "500m"
        #     memory: "512Mi"
        # env:
        # - name: RedisConfig__UseAsBackplane
        #   value: "false"
        # - name: RedisConfig__Settings__Host
        #   value: "sigredis-master.default.svc.cluster.local"
        # - name: RedisConfig__Settings__Password
        #   value: "none"
        #   valueFrom:
        #     secretKeyRef:
        #       name: sigredis
        #       key: redis-password
        # - name: PodName
        #   valueFrom:
        #     fieldRef:
        #       fieldPath: metadata.name
        ports:
        - containerPort: 80
        - containerPort: 443

templates/service.yml #

kind: Service
apiVersion: v1
metadata:
  name:  project-service
spec:
  selector:
    app:  project-deployment
  ports:
  - name: http
    port:  80
    targetPort:  80
  - name: https
    port: 443
    targetPort: 443
  sessionAffinity: ClientIP
  type: LoadBalancer

templates/ingress.yml #

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: project-ingress
  annotations:
    nginx.ingress.kubernetes.io/affinity: "cookie"
    nginx.ingress.kubernetes.io/session-cookie-name: "route"
    nginx.ingress.kubernetes.io/session-cookie-hash: "sha1"
spec:
  rules:
  - host: project.local
    http:
      paths:
        - path: /
          pathType: Prefix
          backend:
            service:
              name: project-service
              port:
                number: 80

Kubectl命令输出 #

kubectl get all #

NAME                                    READY   STATUS             RESTARTS       AGE
pod/project-deployment-5cfb9ccfbd-r7zgj   0/1     CrashLoopBackOff   7 (2m8s ago)   13m
pod/ubuntu-pod                          1/1     Running            0              13m

NAME                    TYPE           CLUSTER-IP     EXTERNAL-IP   PORT(S)                      AGE
service/project-service   LoadBalancer   10.99.32.145   <pending>     80:30883/TCP,443:31528/TCP   13m
service/kubernetes      ClusterIP      10.96.0.1      <none>        443/TCP                      8d

NAME                               READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/project-deployment   0/1     1            0           13m

NAME                                          DESIRED   CURRENT   READY   AGE
replicaset.apps/project-deployment-5cfb9ccfbd   1         1         0       13m

kubectl describe pod/project-deployment-5cfb9ccfbd-r7zgj #

Name:             project-deployment-5cfb9ccfbd-r7zgj
Namespace:        default
Priority:         0
Service Account:  default
Node:             minikube/172.28.249.124
Start Time:       Sun, 04 Feb 2024 17:23:43 -0700
Labels:           app=project-deployment
                  pod-template-hash=5cfb9ccfbd
Annotations:      <none>
Status:           Running
IP:               10.244.0.113
IPs:
  IP:           10.244.0.113
Controlled By:  ReplicaSet/project-deployment-5cfb9ccfbd
Containers:
  project-pod:
    Container ID:   docker://4d59b3007cd4c14e532510e9c20005b4a11c9ebdd4ccf609719f13befb107824
    Image:          registry/project/app:0.2.3
    Image ID:       docker-pullable://registry.gitlab.com/sdmines/sd/2023/16/project/app@sha256:196c4bf5d6df9feb768cf6276a745ec1dd3077b4fe4150b2c2d8ef6b95193b5b
    Ports:          80/TCP, 443/TCP
    Host Ports:     0/TCP, 0/TCP
    State:          Waiting
      Reason:       CrashLoopBackOff
    Last State:     Terminated
      Reason:       Completed
      Exit Code:    0
      Started:      Sun, 04 Feb 2024 17:34:51 -0700
      Finished:     Sun, 04 Feb 2024 17:34:51 -0700
    Ready:          False
    Restart Count:  7
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from kube-api-access-gg5tt (ro)
Conditions:
  Type              Status
  Initialized       True
  Ready             False
  ContainersReady   False
  PodScheduled      True
Volumes:
  kube-api-access-gg5tt:
    Type:                    Projected (a volume that contains injected data from multiple sources)
    TokenExpirationSeconds:  3607
    ConfigMapName:           kube-root-ca.crt
    ConfigMapOptional:       <nil>
    DownwardAPI:             true
QoS Class:                   BestEffort
Node-Selectors:              <none>
Tolerations:                 node.kubernetes.io/not-ready:NoExecute op=Exists for 300s
                             node.kubernetes.io/unreachable:NoExecute op=Exists for 300s
Events:
  Type     Reason     Age                   From               Message
  ----     ------     ----                  ----               -------
  Normal   Scheduled  14m                   default-scheduler  Successfully assigned default/project-deployment-5cfb9ccfbd-r7zgj to minikube
  Normal   Pulled     12m (x5 over 14m)     kubelet            Container image "registry/project/app:0.2.3" already present on machine
  Normal   Created    12m (x5 over 14m)     kubelet            Created container project-pod
  Normal   Started    12m (x5 over 14m)     kubelet            Started container project-pod
  Warning  BackOff    4m24s (x48 over 14m)  kubelet            Back-off restarting failed container project-pod in pod project-deployment-5cfb9ccfbd-r7zgj_default(d8dfcdcd-9699-4a58-bb66-108d60167bec)

日志命令输出 #

• kubectl logs pod/project-deployment-5cfb9ccfbd-r7zgj: 无输出
• kubectl logs pod/project-deployment-5cfb9ccfbd-r7zgj --previous: 无输出

kubectl describe deployments/project-deployment #

Name:                   project-deployment
Namespace:              default
CreationTimestamp:      Sun, 04 Feb 2024 17:23:42 -0700
Labels:                 app=project-deployment
                        app.kubernetes.io/managed-by=Helm
Annotations:            deployment.kubernetes.io/revision: 1
                        meta.helm.sh/release-name: project
                        meta.helm.sh/release-namespace: default
Selector:               app=project-deployment
Replicas:               1 desired | 1 updated | 1 total | 0 available | 1 unavailable
StrategyType:           RollingUpdate
MinReadySeconds:        0
RollingUpdateStrategy:  25% max unavailable, 25% max surge
Pod Template:
  Labels:  app=project-deployment
  Containers:
   project-pod:
    Image:        registry/project/app:0.2.3
    Ports:        80/TCP, 443/TCP
    Host Ports:   0/TCP, 0/TCP
    Environment:  <none>
    Mounts:       <none>
  Volumes:        <none>
Conditions:
  Type           Status  Reason
  ----           ------  ------
  Available      False   MinimumReplicasUnavailable
  Progressing    False   ProgressDeadlineExceeded
OldReplicaSets:  <none>
NewReplicaSet:   project-deployment-5cfb9ccfbd (1/1 replicas created)
Events:
  Type    Reason             Age   From                   Message
  ----    ------             ----  ----                   -------
  Normal  ScalingReplicaSet  16m   deployment-controller  Scaled up replica set project-deployment-5cfb9ccfbd to 1

安装命令 #

helm secrets install . project -f enc/secrets.yml

注：未附上enc/secrets.yml，原因：1. 包含敏感信息；2. 文件已加密，对排查无帮助。

从kubectl describe pod输出看，容器退出码为0（Exit Code: 0），说明应用正常启动后主动结束进程，而非崩溃，这是核心排查线索。结合ASP.NET应用特性，给出以下步骤：

1. 本地验证镜像可用性 #

先在Minikube节点上直接运行镜像，确认镜像本身是否能正常启动：

# 进入Minikube节点
minikube ssh
# 运行容器并查看实时输出
docker run -it --rm registry/project/app:0.2.3

若容器立即退出，问题出在镜像本身，与Kubernetes配置无关。

2. 核对应用程序集名称 #

Dockerfile的ENTRYPOINT指定为dotnet project.dll，需确认发布后的程序集名称是否匹配：

docker run -it --rm registry/project/app:0.2.3 /bin/bash
# 查看/app目录下的文件
ls /app

若程序集名称不符，修改Dockerfile的ENTRYPOINT为正确名称。

3. 启用控制台日志输出 #

ASP.NET 6默认可能未将日志输出到控制台，导致Kubectl无法获取日志。在应用Program.cs中添加控制台日志提供者：

var builder = WebApplication.CreateBuilder(args);

// 添加控制台日志
builder.Logging.AddConsole();

// 其他配置代码...

重新构建镜像后部署，再查看日志。

4. 补充必要环境变量 #

当前Deployment未设置任何环境变量，若应用依赖配置项（如数据库连接字符串、Redis地址等），缺失会导致应用启动后退出。可临时添加环境变量测试：

containers:
- name: project-pod
  image: registry/project/app:0.2.3
  env:
  - name: ASPNETCORE_ENVIRONMENT
    value: Development
  # 添加应用所需的其他配置变量

5. 确认应用监听端口 #

ASP.NET 6默认监听5000/5001端口，若未配置为80，需调整：

• 方式一：在Program.cs中指定端口

builder.WebHost.ConfigureKestrel(options =>
{
    options.ListenAnyIP(80);
});

• 方式二：通过环境变量设置

env:
- name: ASPNETCORE_URLS
  value: "http://+:80"

6. 查看容器退出详情 #

若以上步骤未解决问题，在Minikube节点上用docker inspect查看容器退出的详细信息：

docker inspect 4d59b3007cd4c14e532510e9c20005b4a11c9ebdd4ccf609719f13befb107824

重点查看State字段下的ExitCode、FinishedAt及Error信息，进一步定位原因。

内容的提问来源于stack exchange，提问作者Ethan Parsons