deepdata: delivery stack (docker compose, helm, CI, build scripts)
CodeQL Advanced / Analyze (actions) (push) Canceled after 0s
CodeQL Advanced / Analyze (java-kotlin) (push) Canceled after 0s
CodeQL Advanced / Analyze (javascript-typescript) (push) Canceled after 0s
CodeQL Advanced / Analyze (python) (push) Canceled after 0s
Docker Image Build & Push / manifest (push) Canceled after 0s
Docker Image Build & Push / manifest (push) Canceled after 0s
Docker Image Build & Push / manifest (push) Canceled after 0s
Docker Image Build & Push / build-and-push-arm (push) Canceled after 0s
Docker Image Build & Push / build-and-push-amd (push) Canceled after 0s
Docker Image Build & Push / build-and-push-arm (push) Canceled after 0s
Docker Image Build & Push / build-and-push-amd (push) Canceled after 0s
Docker Image Build & Push / build-and-push-arm (push) Canceled after 0s
Docker Image Build & Push / build-and-push-amd (push) Canceled after 0s
Deer Flow Docker Image CI / build-and-push-amd (push) Canceled after 0s
Deer Flow Docker Image CI / build-and-push-arm (push) Canceled after 0s
Deer Flow Docker Image CI / manifest (push) Canceled after 0s
Docker Image Build & Push / manifest (push) Canceled after 0s
Docker Image Build & Push / build-and-push-arm (push) Canceled after 0s
Docker Image Build & Push / build-and-push-amd (push) Canceled after 0s
Docker Image Build & Push / manifest (push) Canceled after 0s
Docker Image Build & Push / build-and-push-arm (push) Canceled after 0s
Docker Image Build & Push / build-and-push-amd (push) Canceled after 0s
Docker Image Build & Push / manifest (push) Canceled after 0s
Docker Image Build & Push / build-and-push-arm (push) Canceled after 0s
Docker Image Build & Push / build-and-push-amd (push) Canceled after 0s
Docker Image Build & Push / manifest (push) Canceled after 0s
Docker Image Build & Push / build-and-push-arm (push) Canceled after 0s
Docker Image Build & Push / build-and-push-amd (push) Canceled after 0s
Python Backend Docker Image CI / Build and Push Backend Python Docker Image (push) Canceled after 0s
Backend Docker Image CI / Build and Push Backend Docker Image (push) Canceled after 0s
Database Docker Image CI / Build and Push Database Docker Image (push) Canceled after 0s
Frontend Docker Image CI / Build and Push Frontend Docker Image (push) Canceled after 0s
Gateway Docker Image CI / Build and Push Gateway Docker Image (push) Canceled after 0s
Label Studio Docker Image CI / Build and Push Label Studio Docker Image (push) Canceled after 0s
Runtime Docker Image CI / Build and Push Runtime Docker Image (push) Canceled after 0s

This commit is contained in:
figmar
2026-08-30 19:23:42 +08:00
parent bd3db30be9
commit 7bbc9d4ef0
852 changed files with 184030 additions and 0 deletions
+80
View File
@@ -0,0 +1,80 @@
# syntax=docker/dockerfile:1
FROM maven:3-eclipse-temurin-8 AS datax-builder
# 仅安装必要的 git
RUN apt-get update && \
apt-get install -y git && \
rm -rf /var/lib/apt/lists/*
# 🟢 修改 1:恢复 git clone 下载完整源码,本地先git clone DataX
# RUN git clone https://github.com/alibaba/DataX.git
COPY DataX/ /DataX/
# 🟢 修改 2:注释掉 COPY,因为它覆盖了完整代码且不完整
COPY runtime/datax/ DataX/
# 🟢 优化 1:挂载 Maven 缓存 + 执行修复与编译
# 注意:这里恢复了之前讨论的 MySQL 驱动修复(sed),因为源码完整了,路径也就对了
RUN --mount=type=cache,target=/root/.m2 \
cd DataX && \
sed -i "s/com.mysql.jdbc.Driver/com.mysql.cj.jdbc.Driver/g" \
plugin-rdbms-util/src/main/java/com/alibaba/datax/plugin/rdbms/util/DataBaseType.java && \
mvn -U clean package assembly:assembly -Dmaven.test.skip=true
FROM python:3.12-slim
# 🟢 优化 2:挂载 APT 缓存,加速 apt-get update
RUN --mount=type=cache,target=/var/cache/apt \
--mount=type=cache,target=/var/lib/apt \
apt-get update && \
apt-get install -y --no-install-recommends vim openjdk-21-jre nfs-common glusterfs-client rsync dos2unix && \
rm -rf /var/lib/apt/lists/*
ENV PYTHONDONTWRITEBYTECODE=1 \
PYTHONUNBUFFERED=1 \
POETRY_VERSION=2.2.1 \
POETRY_NO_INTERACTION=1 \
POETRY_VIRTUALENVS_CREATE=false \
POETRY_CACHE_DIR=/tmp/poetry_cache
ENV JAVA_HOME=/usr/lib/jvm/java-21-openjdk
ENV PATH="/root/.local/bin:$JAVA_HOME/bin:$PATH"
WORKDIR /app
# 安装 Poetry
RUN pip install --upgrade --root-user-action=ignore pip \
&& pip install --root-user-action=ignore pipx \
&& pipx install "poetry==$POETRY_VERSION"
# 复制依赖文件
COPY runtime/deepdata-python/pyproject.toml runtime/deepdata-python/poetry.lock* /app/
# 🟢 优化 3:挂载 Poetry 缓存,避免重复下载 Python 包
RUN --mount=type=cache,target=$POETRY_CACHE_DIR \
poetry install --no-root --only main
# 🆕 新增:安装 Ray SDK (用于提交任务到 Ray 集群)
# 注意:使用 [default] 确保包含 job submission client
RUN pip install "ray[default]>=2.9.0"
# 下载 NLTK 数据
RUN python -c "import nltk; nltk.download(['punkt_tab','averaged_perceptron_tagger_eng'], download_dir='/usr/local/nltk_data')"
ENV NLTK_DATA=/usr/local/nltk_data
# 复制应用代码
COPY runtime/deepdata-python /app
COPY runtime/ops/examples/test_operator/test_operator.tar /app/test_operator.tar
COPY --from=datax-builder /DataX/target/datax/datax /opt/datax
RUN cp /opt/datax/plugin/reader/mysqlreader/libs/mysql* /opt/datax/plugin/reader/starrocksreader/libs/
COPY runtime/deepdata-python/deploy/docker-entrypoint.sh /docker-entrypoint.sh
RUN chmod +x /docker-entrypoint.sh \
&& dos2unix /docker-entrypoint.sh || true \
&& ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime
# Expose the application port
EXPOSE 18000
ENTRYPOINT ["/docker-entrypoint.sh"]
+31
View File
@@ -0,0 +1,31 @@
FROM maven:3-eclipse-temurin-21 AS builder
COPY backend/ /opt/backend
#RUN cd /opt/backend/ && mvn -U clean package -Dmaven.test.skip=true
RUN --mount=type=cache,target=/root/.m2 \
cd /opt/backend/ && mvn -U clean package -Dmaven.test.skip=true
FROM eclipse-temurin:21-jdk
RUN apt-get update && \
apt-get install -y vim wget curl rsync python3 python3-pip python-is-python3 dos2unix && \
apt-get clean && \
rm -rf /var/lib/apt/lists/*
COPY --from=builder /opt/backend/services/main-application/target/deepdata.jar /opt/backend/deepdata.jar
COPY scripts/images/backend/start.sh /opt/backend/start.sh
COPY runtime/ops/examples/test_operator/test_operator.tar /opt/backend/test_operator.tar
RUN dos2unix /opt/backend/start.sh \
&& chmod +x /opt/backend/start.sh \
&& ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime
EXPOSE 8080
ENTRYPOINT ["/opt/backend/start.sh"]
CMD ["java", "-Duser.timezone=Asia/Shanghai", "-jar", "/opt/backend/deepdata.jar"]
+8
View File
@@ -0,0 +1,8 @@
#!/bin/bash
set -e
umask 0022
echo "Starting main application..."
exec "$@"
@@ -0,0 +1,97 @@
#!/bin/bash
set -euo pipefail
run_once=false
if [ "${1:-}" = "--once" ]; then
run_once=true
shift
fi
LOG_ROOT="${1:-${LOG_ROTATION_PATH:-/tmp/ray}}"
LOG_ROTATION_MAX_SIZE="${LOG_ROTATION_MAX_SIZE:-100MB}"
LOG_ROTATION_BACKUP_COUNT="${LOG_ROTATION_BACKUP_COUNT:-30}"
LOG_ROTATION_INTERVAL_SECONDS="${LOG_ROTATION_INTERVAL_SECONDS:-300}"
parse_size_bytes() {
local value
value="$(printf '%s' "$1" | tr '[:lower:]' '[:upper:]')"
value="${value// /}"
case "$value" in
*GB) echo $((${value%GB} * 1024 * 1024 * 1024)) ;;
*G) echo $((${value%G} * 1024 * 1024 * 1024)) ;;
*MB) echo $((${value%MB} * 1024 * 1024)) ;;
*M) echo $((${value%M} * 1024 * 1024)) ;;
*KB) echo $((${value%KB} * 1024)) ;;
*K) echo $((${value%K} * 1024)) ;;
*B) echo "${value%B}" ;;
*) echo "$value" ;;
esac
}
file_size_bytes() {
wc -c < "$1" | tr -d ' '
}
backup_path() {
local file="$1"
local index="$2"
local ext="${file##*.}"
local stem="${file%.$ext}"
echo "${stem}.${index}.${ext}"
}
is_backup_file() {
local name
name="$(basename "$1")"
[[ "$name" =~ \.[0-9]+\.(log|out|err)$ ]]
}
rotate_file() {
local file="$1"
local max_bytes="$2"
local backup_count="$3"
[ -f "$file" ] || return 0
is_backup_file "$file" && return 0
local current_size
current_size="$(file_size_bytes "$file")"
[ "$current_size" -lt "$max_bytes" ] && return 0
rm -f "$(backup_path "$file" "$backup_count")"
local index
for ((index = backup_count - 1; index >= 1; index--)); do
local src dst
src="$(backup_path "$file" "$index")"
dst="$(backup_path "$file" "$((index + 1))")"
[ -f "$src" ] && mv "$src" "$dst"
done
cp -p "$file" "$(backup_path "$file" 1)"
: > "$file"
}
rotate_once() {
[ -d "$LOG_ROOT" ] || return 0
local max_bytes
max_bytes="$(parse_size_bytes "$LOG_ROTATION_MAX_SIZE")"
while IFS= read -r -d '' file; do
rotate_file "$file" "$max_bytes" "$LOG_ROTATION_BACKUP_COUNT"
done < <(
find "$LOG_ROOT" -type f \( -name '*.log' -o -name '*.out' -o -name '*.err' \) -print0 2>/dev/null || true
)
}
if [ "$run_once" = true ]; then
rotate_once
exit 0
fi
while true; do
rotate_once
sleep "$LOG_ROTATION_INTERVAL_SECONDS"
done
+22
View File
@@ -0,0 +1,22 @@
FROM postgres:15-trixie
# 创建日志目录并设置权限
USER root
RUN mkdir -p /var/log/deepdata/database && \
chown -R postgres:postgres /var/log/deepdata && \
chmod -R 755 /var/log/deepdata
ENV POSTGRES_INITDB_ARGS="--encoding=UTF8 --lc-collate=en_US.UTF-8 --lc-ctype=en_US.UTF-8"
# 创建配置目录
RUN mkdir -p /etc/postgresql/
# 复制 PostgreSQL 主配置文件
COPY scripts/images/database/postgresql.cnf /etc/postgresql/postgresql.conf
COPY scripts/images/database/prune-postgres-logs.sh /usr/local/bin/prune-postgres-logs.sh
COPY scripts/images/database/start.sh /usr/local/bin/deepdata-database-start.sh
COPY scripts/db/ /docker-entrypoint-initdb.d/
RUN chmod +x /usr/local/bin/prune-postgres-logs.sh /usr/local/bin/deepdata-database-start.sh
ENTRYPOINT ["/usr/local/bin/deepdata-database-start.sh"]
CMD ["postgres", "-c", "config_file=/etc/postgresql/postgresql.conf"]
+25
View File
@@ -0,0 +1,25 @@
# 1. 字符集设置
client_encoding = 'UTF8'
# server_encoding = 'UTF8' # 只能启动时设置,无法通过修改配置文件设置
# 2. 排序规则
# lc_collate = 'en_US.UTF-8' # 只能启动时设置,无法通过修改配置文件设置
# lc_ctype = 'en_US.UTF-8' # 只能启动时设置,无法通过修改配置文件设置
# 3. 时区设置
timezone = 'Asia/Shanghai'
log_timezone = 'Asia/Shanghai'
# 4. 日志设置
log_destination = 'stderr'
logging_collector = on
log_directory = '/var/log/deepdata/database'
log_filename = 'postgresql-%Y-%m-%d_%H%M%S.log'
log_file_mode = 0644
log_rotation_age = 1d
log_rotation_size = 100MB
log_truncate_on_rotation = off
log_error_verbosity = default
log_statement = 'ddl'
log_duration = off
log_lock_waits = on
log_min_duration_statement = 1000
# 5. 监听配置
listen_addresses = '*'
@@ -0,0 +1,32 @@
#!/bin/bash
set -euo pipefail
LOG_DIR="${POSTGRES_LOG_DIR:-/var/log/deepdata/database}"
LOG_ROTATION_BACKUP_COUNT="${LOG_ROTATION_BACKUP_COUNT:-30}"
LOG_ROTATION_INTERVAL_SECONDS="${LOG_ROTATION_INTERVAL_SECONDS:-300}"
prune_once() {
[ -d "$LOG_DIR" ] || return 0
mapfile -t files < <(
find "$LOG_DIR" -maxdepth 1 -type f -name 'postgresql-*.log' -printf '%T@ %p\n' 2>/dev/null \
| sort -rn \
| sed 's/^[^ ]* //'
)
local index
for ((index = LOG_ROTATION_BACKUP_COUNT; index < ${#files[@]}; index++)); do
rm -f "${files[$index]}"
done
}
if [ "${1:-}" = "--once" ]; then
prune_once
exit 0
fi
while true; do
prune_once
sleep "$LOG_ROTATION_INTERVAL_SECONDS"
done
+7
View File
@@ -0,0 +1,7 @@
#!/bin/bash
set -e
/usr/local/bin/prune-postgres-logs.sh &
exec docker-entrypoint.sh "$@"
@@ -0,0 +1,29 @@
FROM ghcr.io/astral-sh/uv:python3.12-bookworm
# Install uv.
COPY --from=ghcr.io/astral-sh/uv:latest /uv /bin/uv
# Install system dependencies including libpq
RUN apt-get update && apt-get install -y \
libpq-dev git \
&& rm -rf /var/lib/apt/lists/*
WORKDIR /app
# Copy the application into the container.
COPY deepflow /app/
COPY runtime/deepflow/.env /app/.env
COPY runtime/deepflow/conf.yaml /app/conf.yaml
# Pre-cache the application dependencies.
RUN --mount=type=cache,target=/root/.cache/uv \
uv sync --locked --no-install-project
# Install the application dependencies.
RUN --mount=type=cache,target=/root/.cache/uv \
uv sync --locked
EXPOSE 8000
# Run the application.
CMD ["uv", "run", "--no-sync", "python", "server.py", "--host", "0.0.0.0", "--port", "8000"]
@@ -0,0 +1,66 @@
##### DEPENDENCIES
FROM node:20-alpine AS deps
RUN apk add --no-cache libc6-compat openssl
WORKDIR /app
# 原git拉取逻辑注释,改用本地COPY,本地上下文目录存在 deepflow 文件夹
# && mv /deepflow/web/* /app \
# && rm -rf /deepflow
# 将本地 deepflow/web 下全部内容复制到镜像 /app
COPY deepflow/web/ ./
# Install dependencies based on the preferred package manager
RUN if [ -f yarn.lock ]; then yarn --frozen-lockfile; \
elif [ -f package-lock.json ]; then npm ci; \
elif [ -f pnpm-lock.yaml ]; then npm install -g pnpm && pnpm i; \
else echo "Lockfile not found." && exit 1; \
fi
##### BUILDER
FROM node:20-alpine AS builder
RUN apk add --no-cache git
WORKDIR /app
ARG NEXT_PUBLIC_API_URL="/deepflow-backend"
# 原git拉取逻辑注释
# RUN cd /deepflow \
# && mv /deepflow/web/* /app \
# && rm -rf /deepflow
# 本地复制web目录源码
COPY deepflow/web/ ./
COPY --from=deps /app/node_modules ./node_modules
ENV NEXT_TELEMETRY_DISABLED=1
RUN if [ -f yarn.lock ]; then SKIP_ENV_VALIDATION=1 yarn build; \
elif [ -f package-lock.json ]; then SKIP_ENV_VALIDATION=1 npm run build; \
elif [ -f pnpm-lock.yaml ]; then npm install -g pnpm && SKIP_ENV_VALIDATION=1 pnpm run build; \
else echo "Lockfile not found." && exit 1; \
fi
##### RUNNER
FROM node:20-alpine AS runner
WORKDIR /app
ENV NODE_ENV=production
ENV NEXT_TELEMETRY_DISABLED=1
COPY --from=builder /app/next.config.js ./
COPY --from=builder /app/public ./public
COPY --from=builder /app/package.json ./package.json
COPY --from=builder /app/.next/standalone ./
COPY --from=builder /app/.next/static ./.next/static
EXPOSE 3000
ENV PORT=3000
CMD ["server.js"]
+41
View File
@@ -0,0 +1,41 @@
FROM node:20-alpine AS builder
WORKDIR /app
COPY frontend ./
RUN if [ -f package-lock.json ]; then npm ci; else npm install; fi && \
npm run build
FROM nginx:1.29 AS runner
RUN --mount=type=cache,target=/var/cache/apt \
--mount=type=cache,target=/var/lib/apt \
apt update \
&& apt install -y dos2unix python3 python3-dev python3-venv libaugeas-dev gcc vim cron \
&& python3 -m venv /opt/certbot/ \
&& /opt/certbot/bin/pip install --upgrade pip \
&& /opt/certbot/bin/pip install certbot certbot-nginx \
&& ln -s /opt/certbot/bin/certbot /usr/local/bin/certbot
COPY --from=builder /app/dist /opt/frontend/statics
COPY scripts/images/frontend/routes.inc /opt/frontend/routes.inc
COPY scripts/images/frontend/http_backend.conf /opt/frontend/http_backend.conf
COPY scripts/images/frontend/https_backend.conf /opt/frontend/https_backend.conf
COPY scripts/images/frontend/nginx.conf /opt/frontend/nginx.conf
COPY scripts/images/frontend/rotate-nginx-logs.sh /usr/local/bin/rotate-nginx-logs.sh
COPY scripts/images/frontend/start.sh /opt/frontend/start.sh
RUN dos2unix /opt/frontend/start.sh \
&& dos2unix /usr/local/bin/rotate-nginx-logs.sh \
&& chmod +x /opt/frontend/start.sh \
&& chmod +x /usr/local/bin/rotate-nginx-logs.sh \
&& mkdir -p /etc/nginx/cert \
&& ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime \
&& rm -f /etc/nginx/nginx.conf \
&& rm -f /etc/nginx/conf.d/default.conf
EXPOSE 3000
CMD ["/opt/frontend/start.sh"]
+13
View File
@@ -0,0 +1,13 @@
server {
listen 3000;
server_name 0.0.0.0;
access_log /var/log/deepdata/frontend/access.log main;
error_log /var/log/deepdata/frontend/error.log notice;
client_max_body_size 1024M;
add_header Set-Cookie "NEXT_LOCALE=zh";
include /etc/nginx/conf.d/routes.inc;
}
@@ -0,0 +1,18 @@
server {
listen 3000 ssl;
server_name 0.0.0.0;
access_log /var/log/deepdata/frontend/access.log main;
error_log /var/log/deepdata/frontend/error.log notice;
client_max_body_size 1024M;
ssl_certificate /etc/nginx/cert/server.pem;
ssl_certificate_key /etc/nginx/cert/server.key;
error_page 497 https://$host$request_uri;
add_header Set-Cookie "NEXT_LOCALE=zh";
include /etc/nginx/conf.d/routes.inc;
}
+32
View File
@@ -0,0 +1,32 @@
user nginx;
worker_processes auto;
error_log /var/log/deepdata/frontend/error.log notice;
pid /run/nginx.pid;
events {
worker_connections 1024;
}
http {
include /etc/nginx/mime.types;
default_type application/octet-stream;
log_format main '$remote_addr - $remote_user [$time_local] "$request" '
'$status $body_bytes_sent "$http_referer" '
'"$http_user_agent" "$http_x_forwarded_for"';
access_log /var/log/deepdata/frontend/access.log main;
sendfile on;
#tcp_nopush on;
keepalive_timeout 65;
#gzip on;
include /etc/nginx/conf.d/*.conf;
}
@@ -0,0 +1,92 @@
#!/bin/bash
set -euo pipefail
LOG_DIR="${NGINX_LOG_DIR:-/var/log/deepdata/frontend}"
LOG_ROTATION_MAX_SIZE="${LOG_ROTATION_MAX_SIZE:-100MB}"
LOG_ROTATION_BACKUP_COUNT="${LOG_ROTATION_BACKUP_COUNT:-30}"
LOG_ROTATION_INTERVAL_SECONDS="${LOG_ROTATION_INTERVAL_SECONDS:-300}"
parse_size_bytes() {
local value
value="$(printf '%s' "$1" | tr '[:lower:]' '[:upper:]')"
value="${value// /}"
case "$value" in
*GB) echo $((${value%GB} * 1024 * 1024 * 1024)) ;;
*G) echo $((${value%G} * 1024 * 1024 * 1024)) ;;
*MB) echo $((${value%MB} * 1024 * 1024)) ;;
*M) echo $((${value%M} * 1024 * 1024)) ;;
*KB) echo $((${value%KB} * 1024)) ;;
*K) echo $((${value%K} * 1024)) ;;
*B) echo "${value%B}" ;;
*) echo "$value" ;;
esac
}
file_size_bytes() {
wc -c < "$1" | tr -d ' '
}
backup_path() {
local file="$1"
local index="$2"
local ext="${file##*.}"
local stem="${file%.$ext}"
echo "${stem}.${index}.${ext}"
}
rotate_file() {
local file="$1"
local max_bytes="$2"
local backup_count="$3"
[ -f "$file" ] || return 1
local current_size
current_size="$(file_size_bytes "$file")"
[ "$current_size" -lt "$max_bytes" ] && return 1
rm -f "$(backup_path "$file" "$backup_count")"
local index
for ((index = backup_count - 1; index >= 1; index--)); do
local src dst
src="$(backup_path "$file" "$index")"
dst="$(backup_path "$file" "$((index + 1))")"
[ -f "$src" ] && mv "$src" "$dst"
done
mv "$file" "$(backup_path "$file" 1)"
install -o nginx -g nginx -m 0644 /dev/null "$file"
return 0
}
rotate_once() {
mkdir -p "$LOG_DIR"
touch "$LOG_DIR/access.log" "$LOG_DIR/error.log"
chown -R nginx:nginx "$LOG_DIR" 2>/dev/null || true
local max_bytes
max_bytes="$(parse_size_bytes "$LOG_ROTATION_MAX_SIZE")"
local rotated=false
for file in "$LOG_DIR/access.log" "$LOG_DIR/error.log"; do
if rotate_file "$file" "$max_bytes" "$LOG_ROTATION_BACKUP_COUNT"; then
rotated=true
fi
done
if [ "$rotated" = true ]; then
nginx -s reopen || true
fi
}
if [ "${1:-}" = "--once" ]; then
rotate_once
exit 0
fi
while true; do
rotate_once
sleep "$LOG_ROTATION_INTERVAL_SECONDS"
done
+40
View File
@@ -0,0 +1,40 @@
location /api/ {
proxy_pass http://deepdata-gateway:8080/api/;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
}
location /chat {
proxy_pass http://deepdata-gateway:8080/chat;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_intercept_errors on;
error_page 500 501 502 503 504 =200 @static_fallback;
}
location /_next {
proxy_pass http://deepdata-gateway:8080/_next;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
}
location /deepflow-backend {
proxy_pass http://deepdata-gateway:8080/deepflow-backend;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
}
location @static_fallback {
root /opt/frontend/statics;
try_files $uri $uri/ /index.html;
}
location / {
root /opt/frontend/statics;
try_files $uri $uri/ /index.html;
}
+62
View File
@@ -0,0 +1,62 @@
#!/bin/bash
function set_proxy_protocol() {
if [ "${REAL_IP_MODE}" != "proxy_protocol" ]; then
echo "REAL_IP_MODE is ${REAL_IP_MODE}, no need to update nginx configuration file."
return 0
fi
sed -i 's/listen 3000;/listen 3000 proxy_protocol;/' /opt/frontend/http_backend.conf
sed -i 's/listen 3000 ssl;/listen 3000 ssl proxy_protocol;/' /opt/frontend/https_backend.conf
sed -i '/access_log.*main/a\ set_real_ip_from 0.0.0.0/0;' /opt/frontend/nginx.conf
sed -i '/access_log.*main/a\ real_ip_header proxy_protocol;' /opt/frontend/nginx.conf
echo "Nginx configuration file updated."
}
mkdir -p /var/log/deepdata/frontend
touch /var/log/deepdata/frontend/access.log /var/log/deepdata/frontend/error.log
chown -R nginx:nginx /var/log/deepdata/frontend 2>/dev/null || true
/usr/local/bin/rotate-nginx-logs.sh &
if [ -f "/cert/server.pem" ]; then
cp /cert/server.pem /etc/nginx/cert/server.pem
chown nginx:nginx /etc/nginx/cert/server.pem
fi
if [ -f "/cert/server.key" ]; then
# Private key MUST be encrypted. Plaintext keys are not accepted.
if grep -q "ENCRYPTED" /cert/server.key 2>/dev/null; then
if [ -z "$CERT_PASS" ]; then
echo "ERROR: CERT_PASS is required to decrypt the private key"
exit 1
fi
echo "$CERT_PASS" | openssl pkey -in /cert/server.key -out /etc/nginx/cert/server.key -passin stdin
chmod 600 /etc/nginx/cert/server.key
else
echo "ERROR: /cert/server.key is not encrypted. Please encrypt it before deployment."
exit 1
fi
chown nginx:nginx /etc/nginx/cert/server.key
fi
set_proxy_protocol
cp /opt/frontend/nginx.conf /etc/nginx/nginx.conf
if [ -f "/etc/nginx/cert/server.pem" ]; then
cp /opt/frontend/https_backend.conf /etc/nginx/conf.d/default.conf
cp /opt/frontend/routes.inc /etc/nginx/conf.d/routes.inc
echo "Switching to HTTPS config"
else
cp /opt/frontend/http_backend.conf /etc/nginx/conf.d/default.conf
cp /opt/frontend/routes.inc /etc/nginx/conf.d/routes.inc
if [ -n "$DOMAIN" ]; then
cron
certbot --nginx "-d ${DOMAIN//,/ -d }"
echo "Switching to HTTPS config, Domain: $DOMAIN"
echo "0 0 1 * * root /usr/local/bin/certbot renew --quiet" | tee /etc/cron.d/certbot-renew
chmod 0644 /etc/cron.d/certbot-renew
else
echo "Switching to HTTP config"
fi
fi
exec nginx -g "daemon off;"
+28
View File
@@ -0,0 +1,28 @@
FROM maven:3-eclipse-temurin-21 AS builder
COPY backend/ /opt/gateway
RUN cd /opt/gateway/ && \
mvn -U clean package -Dmaven.test.skip=true
FROM eclipse-temurin:21-jdk
RUN apt-get update && \
apt-get install -y vim wget curl dos2unix && \
apt-get clean && \
rm -rf /var/lib/apt/lists/*
COPY --from=builder /opt/gateway/api-gateway/target/gateway.jar /opt/gateway/gateway.jar
COPY scripts/images/gateway/start.sh /opt/gateway/start.sh
RUN dos2unix /opt/gateway/start.sh \
&& chmod +x /opt/gateway/start.sh \
&& ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime
EXPOSE 8080
ENTRYPOINT ["/opt/gateway/start.sh"]
CMD ["java", "-Duser.timezone=Asia/Shanghai", "-jar", "/opt/gateway/gateway.jar"]
+6
View File
@@ -0,0 +1,6 @@
#!/bin/bash
set -e
echo "Starting main application..."
exec "$@"
+188
View File
@@ -0,0 +1,188 @@
ARG NODE_VERSION=22
ARG PYTHON_VERSION=3.13
ARG POETRY_VERSION=2.1.4
ARG VERSION_OVERRIDE
ARG BRANCH_OVERRIDE
ARG GIT_REPO=https://github.com/ModelEngine-Group/Label-Studio.git
ARG GIT_BRANCH=ls-release/1.21.0.wsi
################################ Stage: downloader (Source Code)
FROM alpine/git AS downloader
ARG GIT_REPO
ARG GIT_BRANCH
WORKDIR /label-studio
RUN git clone --branch ${GIT_BRANCH} ${GIT_REPO} .
################################ Overview
# This Dockerfile builds a Label Studio environment.
# It consists of five main stages:
# 1. "frontend-builder" - Compiles the frontend assets using Node.
# 2. "frontend-version-generator" - Generates version files for frontend sources.
# 3. "venv-builder" - Prepares the virtualenv environment.
# 4. "py-version-generator" - Generates version files for python sources.
# 5. "prod" - Creates the final production image with the Label Studio, Nginx, and other dependencies.
################################ Stage: frontend-builder (build frontend assets)
FROM --platform=${BUILDPLATFORM} node:${NODE_VERSION}-trixie AS frontend-builder
ENV BUILD_NO_SERVER=true \
BUILD_NO_HASH=true \
BUILD_NO_CHUNKS=true \
BUILD_MODULE=true \
YARN_CACHE_FOLDER=/root/web/.yarn \
NX_CACHE_DIRECTORY=/root/web/.nx \
NODE_ENV=production
WORKDIR /label-studio/web
# Fix Docker Arm64 Build
RUN yarn config set registry https://registry.npmjs.org/
RUN yarn config set network-timeout 1200000 # HTTP timeout used when downloading packages, set to 20 minutes
# [修改] 从 downloader 阶段复制 package.json 和 lock 文件
COPY --from=downloader /label-studio/web/package.json .
COPY --from=downloader /label-studio/web/yarn.lock .
COPY --from=downloader /label-studio/web/tools tools
RUN --mount=type=cache,target=/root/web/.yarn,id=yarn-cache,sharing=locked \
--mount=type=cache,target=/root/web/.nx,id=nx-cache,sharing=locked \
yarn install --prefer-offline --no-progress --pure-lockfile --frozen-lockfile --ignore-engines --non-interactive --production=false
# [修改] 从 downloader 阶段复制前端源码
COPY --from=downloader /label-studio/web/ .
# [修改] 从 downloader 阶段复制 pyproject.toml
COPY --from=downloader /label-studio/pyproject.toml ../pyproject.toml
RUN --mount=type=cache,target=/root/web/.yarn,id=yarn-cache,sharing=locked \
--mount=type=cache,target=/root/web/.nx,id=nx-cache,sharing=locked \
yarn run build
################################ Stage: frontend-version-generator
FROM frontend-builder AS frontend-version-generator
# [修改] 之前是 bind mount 本地 .git,现在需要将 downloader 的 .git 目录复制进来
# 注意:WORKDIR 目前是 /label-studio/web,所以我们将 git 放在上级目录以匹配原逻辑
COPY --from=downloader /label-studio/.git ../.git
RUN --mount=type=cache,target=/root/web/.yarn,id=yarn-cache,sharing=locked \
--mount=type=cache,target=/root/web/.nx,id=nx-cache,sharing=locked \
yarn version:libs
################################ Stage: venv-builder (prepare the virtualenv)
FROM python:${PYTHON_VERSION}-slim-trixie AS venv-builder
ARG POETRY_VERSION
ENV PYTHONUNBUFFERED=1 \
PYTHONDONTWRITEBYTECODE=1 \
PIP_NO_CACHE_DIR=off \
PIP_DISABLE_PIP_VERSION_CHECK=on \
PIP_DEFAULT_TIMEOUT=100 \
PIP_CACHE_DIR="/.cache" \
POETRY_CACHE_DIR="/.poetry-cache" \
POETRY_HOME="/opt/poetry" \
POETRY_VIRTUALENVS_IN_PROJECT=true \
POETRY_VIRTUALENVS_PREFER_ACTIVE_PYTHON=true \
PATH="/opt/poetry/bin:$PATH"
ADD https://install.python-poetry.org /tmp/install-poetry.py
RUN python /tmp/install-poetry.py
RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
--mount=type=cache,target=/var/lib/apt/lists,sharing=locked \
set -eux; \
apt-get update; \
apt-get install --no-install-recommends -y \
build-essential git; \
apt-get autoremove -y
WORKDIR /label-studio
ENV VENV_PATH="/label-studio/.venv"
ENV PATH="$VENV_PATH/bin:$PATH"
## Starting from this line all packages will be installed in $VENV_PATH
COPY --from=downloader /label-studio/pyproject.toml /label-studio/poetry.lock /label-studio/README.md ./
# Set a default build argument for including dev dependencies
ARG INCLUDE_DEV=false
# Install dependencies
RUN --mount=type=cache,target=/.poetry-cache,id=poetry-cache,sharing=locked \
poetry check --lock && \
if [ "$INCLUDE_DEV" = "true" ]; then \
poetry install --no-root --extras uwsgi --with test; \
else \
poetry install --no-root --without test --extras uwsgi; \
fi
# Install LS
COPY --from=downloader /label-studio/label_studio label_studio
RUN --mount=type=cache,target=/.poetry-cache,id=poetry-cache,sharing=locked \
poetry install --only-root --extras uwsgi && \
python3 label_studio/manage.py collectstatic --no-input
################################ Stage: py-version-generator
FROM venv-builder AS py-version-generator
ARG VERSION_OVERRIDE
ARG BRANCH_OVERRIDE
COPY --from=downloader /label-studio/.git ./.git
# Create version_.py and ls-version_.py
RUN VERSION_OVERRIDE=${VERSION_OVERRIDE} BRANCH_OVERRIDE=${BRANCH_OVERRIDE} poetry run python label_studio/core/version.py
################################### Stage: prod
FROM python:${PYTHON_VERSION}-slim-trixie AS production
ENV LS_DIR=/label-studio \
HOME=/label-studio \
LABEL_STUDIO_BASE_DATA_DIR=/label-studio/data \
OPT_DIR=/opt/heartex/instance-data/etc \
PATH="/label-studio/.venv/bin:$PATH" \
DJANGO_SETTINGS_MODULE=core.settings.label_studio \
PYTHONUNBUFFERED=1 \
PYTHONDONTWRITEBYTECODE=1
WORKDIR $LS_DIR
# install prerequisites for app
RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
--mount=type=cache,target=/var/lib/apt/lists,sharing=locked \
set -eux; \
apt-get update; \
apt-get upgrade -y; \
apt-get install --no-install-recommends -y libexpat1 libgl1 libglx-mesa0 libglib2.0-0t64 \
gnupg2 curl nginx \
libopenjp2-7 libnuma1 libspeex1 libvdpau1 libgomp1; \
apt-get autoremove -y
RUN set -eux; \
mkdir -p $LS_DIR $LABEL_STUDIO_BASE_DATA_DIR $OPT_DIR && \
chown -R 1001:0 $LS_DIR $LABEL_STUDIO_BASE_DATA_DIR $OPT_DIR /var/log/nginx /etc/nginx
COPY --chown=1001:0 --from=downloader /label-studio/deploy/default.conf /etc/nginx/nginx.conf
# Copy essential files for installing Label Studio and its dependencies
COPY --chown=1001:0 --from=downloader /label-studio/pyproject.toml .
COPY --chown=1001:0 --from=downloader /label-studio/poetry.lock .
COPY --chown=1001:0 --from=downloader /label-studio/README.md .
COPY --chown=1001:0 --from=downloader /label-studio/LICENSE LICENSE
COPY --chown=1001:0 --from=downloader /label-studio/licenses licenses
COPY --chown=1001:0 --from=downloader /label-studio/deploy deploy
# Copy files from build stages
COPY --chown=1001:0 --from=venv-builder $LS_DIR $LS_DIR
COPY --chown=1001:0 --from=py-version-generator $LS_DIR/label_studio/core/version_.py $LS_DIR/label_studio/core/version_.py
COPY --chown=1001:0 --from=frontend-builder $LS_DIR/web/dist $LS_DIR/web/dist
COPY --chown=1001:0 --from=frontend-version-generator $LS_DIR/web/dist/apps/labelstudio/version.json $LS_DIR/web/dist/apps/labelstudio/version.json
COPY --chown=1001:0 --from=frontend-version-generator $LS_DIR/web/dist/libs/editor/version.json $LS_DIR/web/dist/libs/editor/version.json
COPY --chown=1001:0 --from=frontend-version-generator $LS_DIR/web/dist/libs/datamanager/version.json $LS_DIR/web/dist/libs/datamanager/version.json
USER 1001
EXPOSE 8080
ENTRYPOINT ["./deploy/docker-entrypoint.sh"]
CMD ["label-studio"]
+107
View File
@@ -0,0 +1,107 @@
# ==============================================================================
# Stage 1: DataX Builder
# ==============================================================================
FROM maven:3-eclipse-temurin-8 AS datax-builder
# 安装 git
RUN apt-get update && apt-get install -y git && rm -rf /var/lib/apt/lists/*
COPY DataX/ /DataX/
COPY runtime/datax/ DataX/
# 创建阿里云 settings.xml
RUN mkdir -p /root/.m2 && \
cat > /root/.m2/settings.xml <<EOF
<settings>
<mirrors>
<mirror>
<id>aliyunmaven</id>
<mirrorOf>central</mirrorOf>
<name>Aliyun Maven</name>
<url>https://maven.aliyun.com/repository/public</url>
</mirror>
</mirrors>
</settings>
EOF
# 设置工作目录
WORKDIR /DataX
RUN find /DataX -type f -name "DataBaseType.java" -exec sed -i "s/com.mysql.jdbc.Driver/com.mysql.cj.jdbc.Driver/g" {} + || echo "⚠️ MySQL driver file not found, skipping." && \
sed -i 's/<maven-assembly-plugin.version>.*<\/maven-assembly-plugin.version>/<maven-assembly-plugin.version>2.6<\/maven-assembly-plugin.version>/' /DataX/pom.xml || echo "⚠️ Maven assembly version tag not found, skipping."
# 编译 DataX (精确挂载 repository 目录,不覆盖 settings.xml)
RUN --mount=type=cache,target=/root/.m2/repository \
mvn -s /root/.m2/settings.xml clean package assembly:assembly -Dmaven.test.skip=true
# ==============================================================================
# Stage 2: Ray Worker Runtime (保留原有的 AI 环境,增加 Java 支持)
# ==============================================================================
FROM ghcr.io/astral-sh/uv:python3.11-bookworm
# 1. 安装系统依赖
RUN sed -i 's/deb.debian.org/mirrors.aliyun.com/g' /etc/apt/sources.list.d/debian.sources && \
sed -i 's/security.debian.org/mirrors.aliyun.com/g' /etc/apt/sources.list.d/debian.sources && \
apt-get update && \
apt-get install -y --no-install-recommends \
libgl1 libglib2.0-0 vim libmagic1 dos2unix swig \
poppler-utils tesseract-ocr openjdk-17-jre-headless \
&& rm -rf /var/lib/apt/lists/*
# 2. 下载 AI 模型 (保持不变)
RUN mkdir -p /home/models \
&& wget -q --show-progress https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/PP-LCNet_x1_0_doc_ori_infer.tar \
&& tar -xf PP-LCNet_x1_0_doc_ori_infer.tar -C /home/models \
&& rm -f PP_*.tar
# 3. 复制业务代码 (保持不变)
COPY runtime/python-executor /opt/runtime
COPY runtime/ops /opt/runtime/deepdata/ops
COPY runtime/ops/user /opt/runtime/user
COPY scripts/images/runtime-ray/start.sh /opt/runtime/start.sh
COPY scripts/images/common/log-rotate-copytruncate.sh /usr/local/bin/log-rotate-copytruncate.sh
# 4. 复制编译好的 DataX (从 builder 阶段复制)
# ✅ 新增: 将编译好的 datax 复制到 /opt/datax
COPY --from=datax-builder /DataX/target/datax/datax /opt/datax
# ✅ 新增: 修复 MySQL 驱动路径 (参照 backend-python)
RUN cp /opt/datax/plugin/reader/mysqlreader/libs/mysql* /opt/datax/plugin/reader/starrocksreader/libs/ || true
# 5. 环境变量配置
ENV PYTHONPATH=/opt/runtime/deepdata/
ENV UV_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple
ENV UV_EXTRA_INDEX_URL=https://download.pytorch.org/whl/cpu
ENV UV_INDEX_STRATEGY=unsafe-best-match
ENV PYTHONDONTWRITEBYTECODE=1
ENV PYTHONUNBUFFERED=1
# ✅ 新增: 设置 JAVA_HOME,确保 datax.py 能正确调用 Java
ENV JAVA_HOME=/usr/lib/jvm/java-17-openjdk-amd64
ENV PATH="$PATH:$JAVA_HOME/bin"
WORKDIR /opt/runtime
COPY scripts/images/runtime-ray/zh_core_web_sm-3.8.0-py3-none-any.whl /tmp/
ENV UV_LINK_MODE=copy
# 6. 安装 Python 依赖
RUN --mount=type=cache,target=/root/.cache/uv \
uv pip install -e . --system \
&& uv pip install -r /opt/runtime/deepdata/ops/pyproject.toml --system \
&& uv pip install "ray[default]>=2.9.0" --system \
&& uv pip install /tmp/zh_core_web_sm-3.8.0-py3-none-any.whl --system \
&& python -c "import nltk; nltk.download('punkt_tab'); nltk.download('averaged_perceptron_tagger_eng')" \
&& echo "/usr/local/lib/ops/site-packages" > /usr/local/lib/python3.11/site-packages/ops.pth
# 7. 初始化与权限
RUN ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime \
&& chmod +x /opt/runtime/start.sh \
&& chmod +x /usr/local/bin/log-rotate-copytruncate.sh \
&& dos2unix /opt/runtime/start.sh
EXPOSE 8081
ENTRYPOINT ["/opt/runtime/start.sh"]
+35
View File
@@ -0,0 +1,35 @@
#!/bin/bash
set -e
# 1. 初始化环境 (原有的逻辑:注入用户自定义算子)
if [ -d "/opt/runtime/user" ]; then
cp -r /opt/runtime/user/* /opt/runtime/deepdata/ops/user 2>/dev/null || true
fi
# 2. 根据 RAY_MODE 环境变量决定启动模式
if [ "${RAY_MODE}" = "head" ]; then
echo "🚀 Starting Ray Head..."
# ✅ 修复 Head 分支:
# Ray 2.x 已废弃 CLI 参数 --redis-address,改为自动读取环境变量 RAY_REDIS_ADDRESS。
# 只要确保 docker-compose.yml 中配置了 RAY_REDIS_ADDRESS 环境变量即可。
exec ray start --head \
--port=${RAY_HEAD_PORT:-6379} \
--redis-password=${RAY_REDIS_PASSWORD} \
--dashboard-host=0.0.0.0 \
--block
elif [ "${RAY_MODE}" = "worker" ]; then
echo "🚀 Starting Ray Worker connecting to ${RAY_HEAD_ADDRESS}..."
# ✅ 修复 Worker 分支:
# Ray 2.x 移除了 --worker 参数。只要不传 --head,默认即为 Worker 模式。
exec ray start \
--address=${RAY_HEAD_ADDRESS} \
--redis-password=${RAY_REDIS_PASSWORD} \
--num-cpus=${NUM_CPUS:-4} \
--block
else
echo "🔧 Starting Runtime (Legacy Mode / Job Execution)..."
# 默认行为:执行 docker-compose.yml command 传入的指令,或 Ray Job 的入口点
exec "$@"
fi
+39
View File
@@ -0,0 +1,39 @@
FROM ghcr.io/astral-sh/uv:python3.11-bookworm
RUN --mount=type=cache,target=/var/cache/apt \
--mount=type=cache,target=/var/lib/apt \
apt update \
&& apt install -y libgl1 libglib2.0-0 vim libmagic1 libreoffice dos2unix swig poppler-utils tesseract-ocr
RUN mkdir -p /home/models \
&& wget https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/PP-LCNet_x1_0_doc_ori_infer.tar \
&& tar -xf PP-LCNet_x1_0_doc_ori_infer.tar -C /home/models \
&& rm -f PP_*.tar
COPY runtime/python-executor /opt/runtime
COPY runtime/ops /opt/runtime/deepdata/ops
COPY runtime/ops/user /opt/runtime/user
COPY scripts/images/runtime/start.sh /opt/runtime/start.sh
COPY scripts/images/common/log-rotate-copytruncate.sh /usr/local/bin/log-rotate-copytruncate.sh
ENV PYTHONPATH=/opt/runtime/deepdata/
ENV UV_EXTRA_INDEX_URL="https://download.pytorch.org/whl/cpu"
ENV UV_INDEX_STRATEGY=unsafe-best-match
WORKDIR /opt/runtime
RUN --mount=type=cache,target=/root/.cache/uv \
uv pip install -e . --system \
&& uv pip install -r /opt/runtime/deepdata/ops/pyproject.toml --system \
&& python -m spacy download zh_core_web_sm \
&& python -c "import nltk; nltk.download('punkt_tab'); nltk.download('averaged_perceptron_tagger_eng')" \
&& echo "/usr/local/lib/ops/site-packages" > /usr/local/lib/python3.11/site-packages/ops.pth
RUN ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime \
&& chmod +x /opt/runtime/start.sh \
&& chmod +x /usr/local/bin/log-rotate-copytruncate.sh \
&& dos2unix /opt/runtime/start.sh
EXPOSE 8081
ENTRYPOINT ["/opt/runtime/start.sh"]
+12
View File
@@ -0,0 +1,12 @@
#!/bin/bash
set -e
cp -r /opt/runtime/user/* /opt/runtime/deepdata/ops/user
if [ "${LOG_ROTATION_ENABLED:-true}" = "true" ]; then
LOG_ROTATION_PATH="${LOG_ROTATION_PATH:-/tmp/ray}" /usr/local/bin/log-rotate-copytruncate.sh "${LOG_ROTATION_PATH:-/tmp/ray}" &
fi
echo "Starting main application..."
exec "$@"