HTTP/2 与 RPC 精讲教程 / 14 - 容器化部署
第 14 章：容器化部署
从开发到生产——Docker、Kubernetes 与 Service Mesh 中的 RPC 部署
14.1 Docker 化 gRPC 服务
14.1.1 多阶段构建
# Dockerfile（Go gRPC 服务）
# 阶段 1：编译
FROM golang:1.21-alpine AS builder
RUN apk add --no-cache git ca-certificates
WORKDIR /app
COPY go.mod go.sum ./
RUN go mod download
COPY . .
RUN CGO_ENABLED=0 GOOS=linux GOARCH=amd64 \
go build -ldflags="-s -w" -o /app/server ./cmd/server
# 阶段 2：运行镜像
FROM alpine:3.18
RUN apk add --no-cache ca-certificates tzdata
COPY --from=builder /app/server /usr/local/bin/server
# gRPC 默认端口
EXPOSE 50051
# 健康检查端口
EXPOSE 50052
ENTRYPOINT ["server"]
# 构建镜像
docker build -t my-grpc-service:v1 .
# 运行容器
docker run -d \
--name grpc-service \
-p 50051:50051 \
-p 50052:50052 \
my-grpc-service:v1
14.1.2 gRPC 健康检查
package main
import (
"log"
"net"
"net/http"
"google.golang.org/grpc"
"google.golang.org/grpc/health"
"google.golang.org/grpc/health/grpc_health_v1"
)
func main() {
// gRPC 服务
grpcServer := grpc.NewServer()
healthServer := health.NewServer()
grpc_health_v1.RegisterHealthServer(grpcServer, healthServer)
// 设置服务状态
healthServer.SetServingStatus("user.UserService",
grpc_health_v1.HealthCheckResponse_SERVING)
// 启动 gRPC 服务
lis, _ := net.Listen("tcp", ":50051")
go grpcServer.Serve(lis)
// HTTP 健康端点（供 Docker/K8s 使用）
http.HandleFunc("/healthz", func(w http.ResponseWriter, r *http.Request) {
w.WriteHeader(http.StatusOK)
w.Write([]byte("ok"))
})
http.HandleFunc("/readyz", func(w http.ResponseWriter, r *http.Request) {
w.WriteHeader(http.StatusOK)
w.Write([]byte("ready"))
})
log.Println("健康检查服务启动于 :50052")
http.ListenAndServe(":50052", nil)
}
# Docker 健康检查
HEALTHCHECK --interval=10s --timeout=3s --start-period=5s --retries=3 \
CMD grpc_health_probe -addr=:50051 || exit 1
# docker-compose.yml
version: "3.8"
services:
user-service:
build: .
ports:
- "50051:50051"
healthcheck:
test: ["CMD", "grpc_health_probe", "-addr=:50051"]
interval: 10s
timeout: 3s
retries: 3
deploy:
resources:
limits:
cpus: "1.0"
memory: 512M
reservations:
cpus: "0.5"
memory: 256M
14.2 Kubernetes 部署
14.2.1 基础部署配置
# k8s/deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: user-service
labels:
app: user-service
spec:
replicas: 3
selector:
matchLabels:
app: user-service
template:
metadata:
labels:
app: user-service
spec:
containers:
- name: user-service
image: my-grpc-service:v1
ports:
- containerPort: 50051
name: grpc
- containerPort: 50052
name: http
# gRPC 健康检查（K8s 1.24+ 支持 gRPC 探针）
livenessProbe:
grpc:
port: 50051
initialDelaySeconds: 10
periodSeconds: 10
readinessProbe:
grpc:
port: 50051
initialDelaySeconds: 5
periodSeconds: 5
resources:
requests:
cpu: 100m
memory: 128Mi
limits:
cpu: 500m
memory: 512Mi
---
apiVersion: v1
kind: Service
metadata:
name: user-service
spec:
selector:
app: user-service
ports:
- name: grpc
port: 50051
targetPort: 50051
protocol: TCP
type: ClusterIP
14.2.2 gRPC 负载均衡问题
Kubernetes 默认的 Service 负载均衡基于 L4（TCP 连接级别）：
问题：
客户端 ──── TCP 连接 ────→ Service ────→ Pod A
（所有请求走同一条连接）
由于 HTTP/2 使用单一长连接，K8s Service 会将所有请求路由到同一个 Pod！
解决方案：
1. 客户端负载均衡（推荐）
2. Envoy/Istio sidecar（L7 负载均衡）
3. gRPC-LB 协议
4. Headless Service + 客户端解析
14.2.3 Headless Service + 客户端负载均衡
# Headless Service（不分配 ClusterIP）
apiVersion: v1
kind: Service
metadata:
name: user-service
spec:
clusterIP: None # 关键：Headless
selector:
app: user-service
ports:
- name: grpc
port: 50051
// 客户端：使用 DNS 解析实现负载均衡
package main
import (
"log"
"google.golang.org/grpc"
"google.golang.org/grpc/credentials/insecure"
"google.golang.org/grpc/resolver"
_ "google.golang.org/grpc/health" // 注册健康检查
)
func main() {
// 使用 DNS 解析 + round-robin 负载均衡
conn, err := grpc.Dial(
"dns:///user-service.default.svc.cluster.local:50051",
grpc.WithTransportCredentials(insecure.NewCredentials()),
grpc.WithDefaultServiceConfig(`{
"loadBalancingConfig": [{"round_robin":{}}],
"healthCheckConfig": {
"serviceName": "user.UserService"
}
}`),
)
if err != nil {
log.Fatalf("连接失败: %v", err)
}
defer conn.Close()
}
14.3 Service Mesh 集成
14.3.1 Istio gRPC 配置
# Istio DestinationRule - gRPC 负载均衡
apiVersion: networking.istio.io/v1beta1
kind: DestinationRule
metadata:
name: user-service
spec:
host: user-service
trafficPolicy:
connectionPool:
tcp:
maxConnections: 1000
http:
h2UpgradePolicy: DEFAULT
maxRequestsPerConnection: 1000
loadBalancer:
simple: ROUND_ROBIN
outlierDetection:
consecutive5xxErrors: 3
interval: 10s
baseEjectionTime: 30s
maxEjectionPercent: 50
---
# Istio VirtualService - gRPC 路由
apiVersion: networking.istio.io/v1beta1
kind: VirtualService
metadata:
name: user-service
spec:
hosts:
- user-service
http:
- match:
- method:
exact: POST
uri:
prefix: /user.v1.UserService
route:
- destination:
host: user-service
subset: v1
weight: 90
- destination:
host: user-service
subset: v2
weight: 10 # 金丝雀发布
retries:
attempts: 3
retryOn: unavailable,deadline-exceeded
timeout: 10s
14.3.2 Istio gRPC 可观测性
# gRPC 访问日志配置
apiVersion: telemetry.istio.io/v1alpha1
kind: Telemetry
metadata:
name: grpc-logging
spec:
accessLogging:
- providers:
- name: envoy
metrics:
- providers:
- name: prometheus
tracing:
- providers:
- name: jaeger
randomSamplingPercentage: 10
# Grafana gRPC 监控 Dashboard 查询
# 请求速率
sum(rate(istio_requests_total{
connection_security_policy!="mutual_tls",
destination_service="user-service"
}[5m])) by (grpc_service, grpc_method)
# P99 延迟
histogram_quantile(0.99,
sum(rate(istio_request_duration_milliseconds_bucket{
destination_service="user-service"
}[5m])) by (le)
)
# 错误率
sum(rate(istio_requests_total{
destination_service="user-service",
response_code!="200",
grpc_response_status!="0"
}[5m])) / sum(rate(istio_requests_total{
destination_service="user-service"
}[5m]))
14.4 gRPC-Web 代理
14.4.1 Envoy 代理配置
# envoy-grpc-web.yaml
static_resources:
listeners:
- name: listener_0
address:
socket_address:
address: 0.0.0.0
port_value: 8080
filter_chains:
- filters:
- name: envoy.filters.network.http_connection_manager
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.network.http_connection_manager.v3.HttpConnectionManager
codec_type: AUTO
stat_prefix: ingress_http
route_config:
virtual_hosts:
- name: local_service
domains: ["*"]
routes:
- match:
prefix: "/"
route:
cluster: grpc_backend
timeout: 60s
cors:
allow_origin_string_match:
- prefix: "*"
allow_methods: GET, PUT, DELETE, POST, OPTIONS
allow_headers: keep-alive,user-agent,cache-control,content-type,content-transfer-encoding,x-custom-header,x-accept-content-transfer-encoding,x-accept-response-streaming,x-user-agent,x-grpc-web,grpc-timeout
max_age: "1728000"
expose_headers: grpc-status,grpc-message
http_filters:
- name: envoy.filters.http.grpc_web
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.http.grpc_web.v3.GrpcWeb
- name: envoy.filters.http.cors
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.http.cors.v3.Cors
- name: envoy.filters.http.router
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.http.router.v3.Router
clusters:
- name: grpc_backend
connect_timeout: 1s
type: STRICT_DNS
lb_policy: ROUND_ROBIN
typed_extension_protocol_options:
envoy.extensions.upstreams.http.v3.HttpProtocolOptions:
"@type": type.googleapis.com/envoy.extensions.upstreams.http.v3.HttpProtocolOptions
explicit_http_version_config:
http2_protocol_options: {}
load_assignment:
cluster_name: grpc_backend
endpoints:
- lb_endpoints:
- endpoint:
address:
socket_address:
address: user-service
port_value: 50051
14.5 CI/CD 流水线
# .github/workflows/grpc-service.yml
name: Build and Deploy gRPC Service
on:
push:
branches: [main]
jobs:
build:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Setup Go
uses: actions/setup-go@v5
with:
go-version: "1.21"
- name: Install Protobuf
run: |
sudo apt-get install -y protobuf-compiler
go install google.golang.org/protobuf/cmd/protoc-gen-go@latest
go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@latest
- name: Generate proto
run: buf generate
- name: Run tests
run: go test -v -race ./...
- name: gRPC 集成测试
run: |
go run cmd/server/main.go &
sleep 2
grpcurl -plaintext localhost:50051 grpc.health.v1.Health/Check
- name: Build Docker image
run: docker build -t $REGISTRY/user-service:$GITHUB_SHA .
- name: Push to Registry
run: docker push $REGISTRY/user-service:$GITHUB_SHA
- name: Deploy to K8s
run: |
kubectl set image deployment/user-service \
user-service=$REGISTRY/user-service:$GITHUB_SHA
14.6 注意事项
⚠️ HTTP/2 连接复用问题：
- K8s 默认 L4 负载均衡对 HTTP/2 无效
- 必须使用 L7 负载均衡（Istio/Envoy）或客户端负载均衡
⚠️ 容器网络：
- gRPC 长连接可能穿越 NAT/防火墙，注意超时设置
- 启用 keepalive 探测连接活性
grpc.keepalive_time_ms和grpc.keepalive_timeout_ms
⚠️ 滚动更新：
- gRPC 长连接在 Pod 终止时不会自动断开
- 需要实现优雅关闭（接收 SIGTERM 后发送 GOAWAY）
- 设置
terminationGracePeriodSeconds足够长
💡 最佳实践：
- 始终实现 gRPC 健康检查
- 使用 Headless Service 或 Service Mesh 实现 L7 负载均衡
- 配置合理的资源限制和 HPA
- 监控 gRPC 特定指标（状态码分布、延迟分布）