Files
DataMate/Makefile.offline.mk
Jerry Yan da5b18e423 feat(scripts): 添加 APT 缓存预装功能解决离线构建问题
- 新增 APT 缓存目录和相关构建脚本 export-cache.sh
- 添加 build-base-images.sh 脚本用于构建预装 APT 包的基础镜像
- 增加 build-offline-final.sh 最终版离线构建脚本
- 更新 Makefile.offline.mk 添加新的离线构建目标
- 扩展 README.md 文档详细说明 APT 缓存问题解决方案
- 为多个服务添加使用预装基础镜像的离线 Dockerfile
- 修改打包脚本包含 APT 缓存到最终压缩包中
2026-02-03 13:16:17 +08:00

305 lines
13 KiB
Makefile

# ============================================================================
# Makefile 离线构建扩展
# 将此文件内容追加到主 Makefile 末尾,或单独包含使用
# ============================================================================
# 离线构建配置
CACHE_DIR ?= ./build-cache
OFFLINE_VERSION ?= latest
# 创建 buildx 构建器(如果不存在)
.PHONY: ensure-buildx
ensure-buildx:
@if ! docker buildx inspect offline-builder > /dev/null 2>&1; then \
echo "创建 buildx 构建器..."; \
docker buildx create --name offline-builder --driver docker-container --use 2>/dev/null || docker buildx use offline-builder; \
else \
docker buildx use offline-builder 2>/dev/null || true; \
fi
# ========== 离线缓存导出(有网环境) ==========
.PHONY: offline-export
offline-export: ensure-buildx
@echo "======================================"
@echo "导出离线构建缓存..."
@echo "======================================"
@mkdir -p $(CACHE_DIR)/buildkit $(CACHE_DIR)/images $(CACHE_DIR)/resources
@$(MAKE) _offline-export-base-images
@$(MAKE) _offline-export-cache
@$(MAKE) _offline-export-resources
@$(MAKE) _offline-package
.PHONY: _offline-export-base-images
_offline-export-base-images:
@echo ""
@echo "1. 导出基础镜像..."
@bash -c 'images=( \
"maven:3-eclipse-temurin-21" \
"maven:3-eclipse-temurin-8" \
"eclipse-temurin:21-jdk" \
"mysql:8" \
"node:20-alpine" \
"nginx:1.29" \
"ghcr.nju.edu.cn/astral-sh/uv:python3.11-bookworm" \
"ghcr.nju.edu.cn/astral-sh/uv:python3.12-bookworm" \
"ghcr.nju.edu.cn/astral-sh/uv:latest" \
"python:3.12-slim" \
"python:3.11-slim" \
"gcr.nju.edu.cn/distroless/nodejs20-debian12" \
); for img in "$${images[@]}"; do echo " Pulling $$img..."; docker pull "$$img" 2>/dev/null || true; done'
@echo " Saving base images..."
@docker save -o $(CACHE_DIR)/images/base-images.tar \
maven:3-eclipse-temurin-21 \
maven:3-eclipse-temurin-8 \
eclipse-temurin:21-jdk \
mysql:8 \
node:20-alpine \
nginx:1.29 \
ghcr.nju.edu.cn/astral-sh/uv:python3.11-bookworm \
ghcr.nju.edu.cn/astral-sh/uv:python3.12-bookworm \
ghcr.nju.edu.cn/astral-sh/uv:latest \
python:3.12-slim \
python:3.11-slim \
gcr.nju.edu.cn/distroless/nodejs20-debian12 2>/dev/null || echo " Warning: Some images may not exist"
.PHONY: _offline-export-cache
_offline-export-cache:
@echo ""
@echo "2. 导出 BuildKit 缓存..."
@echo " backend..."
@docker buildx build --cache-to type=local,dest=$(CACHE_DIR)/buildkit/backend-cache,mode=max -f scripts/images/backend/Dockerfile -t datamate-backend:cache . 2>/dev/null || echo " Warning: backend cache export failed"
@echo " backend-python..."
@docker buildx build --cache-to type=local,dest=$(CACHE_DIR)/buildkit/backend-python-cache,mode=max -f scripts/images/backend-python/Dockerfile -t datamate-backend-python:cache . 2>/dev/null || echo " Warning: backend-python cache export failed"
@echo " database..."
@docker buildx build --cache-to type=local,dest=$(CACHE_DIR)/buildkit/database-cache,mode=max -f scripts/images/database/Dockerfile -t datamate-database:cache . 2>/dev/null || echo " Warning: database cache export failed"
@echo " frontend..."
@docker buildx build --cache-to type=local,dest=$(CACHE_DIR)/buildkit/frontend-cache,mode=max -f scripts/images/frontend/Dockerfile -t datamate-frontend:cache . 2>/dev/null || echo " Warning: frontend cache export failed"
@echo " gateway..."
@docker buildx build --cache-to type=local,dest=$(CACHE_DIR)/buildkit/gateway-cache,mode=max -f scripts/images/gateway/Dockerfile -t datamate-gateway:cache . 2>/dev/null || echo " Warning: gateway cache export failed"
@echo " runtime..."
@docker buildx build --cache-to type=local,dest=$(CACHE_DIR)/buildkit/runtime-cache,mode=max -f scripts/images/runtime/Dockerfile -t datamate-runtime:cache . 2>/dev/null || echo " Warning: runtime cache export failed"
@echo " deer-flow-backend..."
@docker buildx build --cache-to type=local,dest=$(CACHE_DIR)/buildkit/deer-flow-backend-cache,mode=max -f scripts/images/deer-flow-backend/Dockerfile -t deer-flow-backend:cache . 2>/dev/null || echo " Warning: deer-flow-backend cache export failed"
@echo " deer-flow-frontend..."
@docker buildx build --cache-to type=local,dest=$(CACHE_DIR)/buildkit/deer-flow-frontend-cache,mode=max -f scripts/images/deer-flow-frontend/Dockerfile -t deer-flow-frontend:cache . 2>/dev/null || echo " Warning: deer-flow-frontend cache export failed"
@echo " mineru..."
@docker buildx build --cache-to type=local,dest=$(CACHE_DIR)/buildkit/mineru-cache,mode=max -f scripts/images/mineru/Dockerfile -t datamate-mineru:cache . 2>/dev/null || echo " Warning: mineru cache export failed"
.PHONY: _offline-export-resources
_offline-export-resources:
@echo ""
@echo "3. 预下载外部资源..."
@mkdir -p $(CACHE_DIR)/resources/models
@echo " PaddleOCR model..."
@wget -q -O $(CACHE_DIR)/resources/models/ch_ppocr_mobile_v2.0_cls_infer.tar \
https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar 2>/dev/null || echo " Warning: PaddleOCR model download failed"
@echo " spaCy model..."
@wget -q -O $(CACHE_DIR)/resources/models/zh_core_web_sm-3.8.0-py3-none-any.whl \
https://ghproxy.net/https://github.com/explosion/spacy-models/releases/download/zh_core_web_sm-3.8.0/zh_core_web_sm-3.8.0-py3-none-any.whl 2>/dev/null || echo " Warning: spaCy model download failed"
@echo " DataX source..."
@if [ ! -d "$(CACHE_DIR)/resources/DataX" ]; then \
git clone --depth 1 https://gitee.com/alibaba/DataX.git $(CACHE_DIR)/resources/DataX 2>/dev/null || echo " Warning: DataX clone failed"; \
fi
@echo " deer-flow source..."
@if [ ! -d "$(CACHE_DIR)/resources/deer-flow" ]; then \
git clone --depth 1 https://ghproxy.net/https://github.com/ModelEngine-Group/deer-flow.git $(CACHE_DIR)/resources/deer-flow 2>/dev/null || echo " Warning: deer-flow clone failed"; \
fi
.PHONY: _offline-package
_offline-package:
@echo ""
@echo "4. 打包缓存..."
@cd $(CACHE_DIR) && tar -czf "build-cache-$$(date +%Y%m%d).tar.gz" buildkit images resources 2>/dev/null && cd - > /dev/null
@echo ""
@echo "======================================"
@echo "✓ 缓存导出完成!"
@echo "======================================"
@echo "传输文件: $(CACHE_DIR)/build-cache-$$(date +%Y%m%d).tar.gz"
# ========== 离线构建(无网环境) ==========
.PHONY: offline-setup
offline-setup:
@echo "======================================"
@echo "设置离线构建环境..."
@echo "======================================"
@if [ ! -d "$(CACHE_DIR)" ]; then \
echo "查找并解压缓存包..."; \
cache_file=$$(ls -t build-cache-*.tar.gz 2>/dev/null | head -1); \
if [ -z "$$cache_file" ]; then \
echo "错误: 未找到缓存压缩包 (build-cache-*.tar.gz)"; \
exit 1; \
fi; \
echo "解压 $$cache_file..."; \
tar -xzf "$$cache_file"; \
else \
echo "缓存目录已存在: $(CACHE_DIR)"; \
fi
@echo ""
@echo "加载基础镜像..."
@if [ -f "$(CACHE_DIR)/images/base-images.tar" ]; then \
docker load -i $(CACHE_DIR)/images/base-images.tar; \
else \
echo "警告: 基础镜像文件不存在,假设已手动加载"; \
fi
@$(MAKE) ensure-buildx
@echo ""
@echo "✓ 离线环境准备完成"
.PHONY: offline-build
offline-build: offline-setup
@echo ""
@echo "======================================"
@echo "开始离线构建..."
@echo "======================================"
@$(MAKE) _offline-build-services
.PHONY: _offline-build-services
_offline-build-services: ensure-buildx
@echo ""
@echo "构建 datamate-database..."
@docker buildx build \
--cache-from type=local,src=$(CACHE_DIR)/buildkit/database-cache \
--pull=false \
-f scripts/images/database/Dockerfile \
-t datamate-database:$(OFFLINE_VERSION) \
--load . || echo " Failed"
@echo ""
@echo "构建 datamate-gateway..."
@docker buildx build \
--cache-from type=local,src=$(CACHE_DIR)/buildkit/gateway-cache \
--pull=false \
-f scripts/images/gateway/Dockerfile \
-t datamate-gateway:$(OFFLINE_VERSION) \
--load . || echo " Failed"
@echo ""
@echo "构建 datamate-backend..."
@docker buildx build \
--cache-from type=local,src=$(CACHE_DIR)/buildkit/backend-cache \
--pull=false \
-f scripts/images/backend/Dockerfile \
-t datamate-backend:$(OFFLINE_VERSION) \
--load . || echo " Failed"
@echo ""
@echo "构建 datamate-frontend..."
@docker buildx build \
--cache-from type=local,src=$(CACHE_DIR)/buildkit/frontend-cache \
--pull=false \
-f scripts/images/frontend/Dockerfile \
-t datamate-frontend:$(OFFLINE_VERSION) \
--load . || echo " Failed"
@echo ""
@echo "构建 datamate-runtime..."
@docker buildx build \
--cache-from type=local,src=$(CACHE_DIR)/buildkit/runtime-cache \
--pull=false \
--build-arg RESOURCES_DIR=$(CACHE_DIR)/resources \
-f scripts/images/runtime/Dockerfile \
-t datamate-runtime:$(OFFLINE_VERSION) \
--load . || echo " Failed"
@echo ""
@echo "构建 datamate-backend-python..."
@docker buildx build \
--cache-from type=local,src=$(CACHE_DIR)/buildkit/backend-python-cache \
--pull=false \
--build-arg RESOURCES_DIR=$(CACHE_DIR)/resources \
-f scripts/images/backend-python/Dockerfile \
-t datamate-backend-python:$(OFFLINE_VERSION) \
--load . || echo " Failed"
@echo ""
@echo "======================================"
@echo "✓ 离线构建完成"
@echo "======================================"
# 单个服务离线构建 (BuildKit)
.PHONY: %-offline-build
%-offline-build: offline-setup ensure-buildx
@echo "离线构建 $*..."
@if [ ! -d "$(CACHE_DIR)/buildkit/$*-cache" ]; then \
echo "错误: $* 的缓存不存在"; \
exit 1; \
fi
@$(eval IMAGE_NAME := $(if $(filter deer-flow%,$*),$*,datamate-$*))
@docker buildx build \
--cache-from type=local,src=$(CACHE_DIR)/buildkit/$*-cache \
--pull=false \
$(if $(filter runtime backend-python deer-flow%,$*),--build-arg RESOURCES_DIR=$(CACHE_DIR)/resources,) \
-f scripts/images/$*/Dockerfile \
-t $(IMAGE_NAME):$(OFFLINE_VERSION) \
--load .
# 传统 Docker 构建(不使用 BuildKit,更稳定)
.PHONY: offline-build-classic
offline-build-classic: offline-setup
@echo "使用传统 docker build 进行离线构建..."
@bash scripts/offline/build-offline-classic.sh $(CACHE_DIR) $(OFFLINE_VERSION)
# 诊断离线环境
.PHONY: offline-diagnose
offline-diagnose:
@bash scripts/offline/diagnose.sh $(CACHE_DIR)
# 构建 APT 预装基础镜像(有网环境)
.PHONY: offline-build-base-images
offline-build-base-images:
@echo "构建 APT 预装基础镜像..."
@bash scripts/offline/build-base-images.sh $(CACHE_DIR)
# 使用预装基础镜像进行离线构建(推荐)
.PHONY: offline-build-final
offline-build-final: offline-setup
@echo "使用预装 APT 包的基础镜像进行离线构建..."
@bash scripts/offline/build-offline-final.sh $(CACHE_DIR) $(OFFLINE_VERSION)
# 完整离线导出(包含 APT 预装基础镜像)
.PHONY: offline-export-full
offline-export-full:
@echo "======================================"
@echo "完整离线缓存导出(含 APT 预装基础镜像)"
@echo "======================================"
@$(MAKE) offline-build-base-images
@$(MAKE) offline-export
@echo ""
@echo "导出完成!传输时请包含以下文件:"
@echo " - build-cache/images/base-images-with-apt.tar"
@echo " - build-cache-YYYYMMDD.tar.gz"
# ========== 帮助 ==========
.PHONY: help-offline
help-offline:
@echo "离线构建命令:"
@echo ""
@echo "【有网环境】"
@echo " make offline-export [CACHE_DIR=./build-cache] - 导出构建缓存"
@echo " make offline-export-full - 导出完整缓存(含 APT 预装基础镜像)"
@echo " make offline-build-base-images - 构建 APT 预装基础镜像"
@echo ""
@echo "【无网环境】"
@echo " make offline-setup [CACHE_DIR=./build-cache] - 解压并准备离线缓存"
@echo " make offline-build-final - 使用预装基础镜像构建(推荐,解决 APT 问题)"
@echo " make offline-build-classic - 使用传统 docker build"
@echo " make offline-build - 使用 BuildKit 构建"
@echo " make offline-diagnose - 诊断离线构建环境"
@echo " make <service>-offline-build - 离线构建单个服务"
@echo ""
@echo "【完整工作流程(推荐)】"
@echo " # 1. 有网环境导出完整缓存"
@echo " make offline-export-full"
@echo ""
@echo " # 2. 传输到无网环境(需要传输两个文件)"
@echo " scp build-cache/images/base-images-with-apt.tar user@offline-server:/path/"
@echo " scp build-cache-*.tar.gz user@offline-server:/path/"
@echo ""
@echo " # 3. 无网环境构建"
@echo " tar -xzf build-cache-*.tar.gz"
@echo " docker load -i build-cache/images/base-images-with-apt.tar"
@echo " make offline-build-final"