# syntax=docker/dockerfile:1
# ============================================================
# AIOps Backend + Agent 多阶段 Dockerfile (BuildKit 加速版)
# Stage 1: 编译后端 (Go 1.24)
# Stage 2: 交叉编译 Agent (Go 1.24, linux-amd64/arm64 + darwin-arm64)
# Stage 3: 运行时 (Alpine 3.20 固定版本)
# 署名 Logan.Li
# ============================================================

# 构建参数 — 版本号通过 git tag 或环境变量注入
ARG VERSION=dev
ARG BUILD_TIME=unknown

# ==================== Stage 1: Backend Builder ====================
FROM golang:1.24-alpine AS builder

WORKDIR /app

# 使用国内 Alpine 镜像源加速 // Logan.Li
RUN sed -i 's|dl-cdn.alpinelinux.org|mirrors.aliyun.com|g' /etc/apk/repositories
RUN apk add --no-cache git ca-certificates tzdata

ENV TZ=Asia/Shanghai
ENV GOPROXY=https://goproxy.cn,direct
ENV GONOSUMCHECK=*
ENV GOMODCACHE=/go/pkg/mod
ENV GOCACHE=/root/.cache/go-build

# 复制后端 go.mod（利用 Docker 层缓存）
COPY backend/go.mod backend/go.sum ./
# --mount=type=cache 让 go mod download 的结果跨构建复用 // Logan.Li
RUN --mount=type=cache,target=/go/pkg/mod \
    --mount=type=cache,target=/root/.cache/go-build \
    go mod download

# 复制后端源码
COPY backend/ ./

# 构建后端二进制（注入版本号）— cache mount 让增量编译极快
RUN --mount=type=cache,target=/go/pkg/mod \
    --mount=type=cache,target=/root/.cache/go-build \
    CGO_ENABLED=0 GOOS=linux GOARCH=amd64 go build \
    -ldflags="-w -s -X main.Version=${VERSION} -X main.BuildTime=${BUILD_TIME}" \
    -o aiops ./cmd/

# ==================== Stage 2: Agent Builder ====================
FROM golang:1.24-alpine AS agent-builder

WORKDIR /agent

ENV GOPROXY=https://goproxy.cn,direct
ENV GOSUMDB=off
ENV GOMODCACHE=/go/pkg/mod
ENV GOCACHE=/root/.cache/go-build

# 复制 agent go.mod
COPY agent/go.mod agent/go.sum ./
RUN --mount=type=cache,target=/go/pkg/mod \
    --mount=type=cache,target=/root/.cache/go-build \
    go mod download

# 复制 agent 源码
COPY agent/ ./

# 交叉编译三个平台的 agent 二进制（注入版本号）
RUN --mount=type=cache,target=/go/pkg/mod \
    --mount=type=cache,target=/root/.cache/go-build \
    CGO_ENABLED=0 GOOS=linux  GOARCH=amd64 go build -ldflags="-s -w -X main.version=${VERSION} -X main.buildTime=${BUILD_TIME}" -o bin/aiops-agent-linux-amd64  . && \
    CGO_ENABLED=0 GOOS=linux  GOARCH=arm64 go build -ldflags="-s -w -X main.version=${VERSION} -X main.buildTime=${BUILD_TIME}" -o bin/aiops-agent-linux-arm64  . && \
    CGO_ENABLED=0 GOOS=darwin GOARCH=arm64 go build -ldflags="-s -w -X main.version=${VERSION} -X main.buildTime=${BUILD_TIME}" -o bin/aiops-agent-darwin-arm64 .

# ==================== Stage 3: Runtime ====================
# 固定版本避免 :latest 每次变动导致全量 cache 失效 // Logan.Li
FROM alpine:3.20

# 使用国内 Alpine 镜像源加速 // Logan.Li
RUN sed -i 's|dl-cdn.alpinelinux.org|mirrors.aliyun.com|g' /etc/apk/repositories

# 安装运行时依赖（apk cache mount 加速重装）
# bash=kubectl交互终端，curl=健康检查，openssh-client=SSH远程安装agent
# kubectl=集群命令行管理，ansible=自动化运维，nodejs+npm=前端构建，maven=Java构建
RUN --mount=type=cache,target=/var/cache/apk \
    apk add ca-certificates tzdata bash curl openssh-client ansible sshpass git docker-cli docker-cli-buildx nodejs npm maven && \
    curl -sLO "https://dl.k8s.io/release/$(curl -sL https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl" && \
    chmod +x kubectl && mv kubectl /usr/local/bin/

# Maven 阿里云镜像加速（避免 DNS 间歇性 SERVFAIL）
RUN mkdir -p /root/.m2 && \
    echo '<?xml version="1.0" encoding="UTF-8"?><settings><mirrors><mirror><id>aliyun</id><mirrorOf>central</mirrorOf><name>Aliyun Maven</name><url>https://maven.aliyun.com/repository/public</url></mirror></mirrors></settings>' > /root/.m2/settings.xml

# npm 淘宝镜像加速
RUN npm config set registry https://registry.npmmirror.com

ENV TZ=Asia/Shanghai

WORKDIR /app

# 复制后端二进制
COPY --from=builder /app/aiops .

# 创建运行时目录（agent 二进制 + 安装脚本）
RUN mkdir -p /app/agent/bin /app/agent/deploy

# 复制 agent 二进制（三个平台）
COPY --from=agent-builder /agent/bin/aiops-agent-linux-amd64  /app/agent/bin/
COPY --from=agent-builder /agent/bin/aiops-agent-linux-arm64  /app/agent/bin/
COPY --from=agent-builder /agent/bin/aiops-agent-darwin-arm64 /app/agent/bin/

# 复制 agent 安装脚本和 systemd service 模板
COPY agent/deploy/install.sh          /app/agent/deploy/
COPY agent/deploy/aiops-agent.service /app/agent/deploy/

# 启动脚本（注入 hosts DNS 缓存）
COPY backend/entrypoint.sh /app/
RUN chmod +x /app/entrypoint.sh

EXPOSE 8080

HEALTHCHECK --interval=30s --timeout=3s --start-period=10s --retries=3 \
    CMD curl -f http://localhost:8080/health || exit 1

CMD ["./entrypoint.sh"]
