当前位置：首页 > news >正文

Qwen3.5-9B企业级部署教程：Nginx反向代理+HTTPS+负载均衡配置

news 2026/8/26 8:24:02

Qwen3.5-9B企业级部署教程：Nginx反向代理+HTTPS+负载均衡配置

1. 前言

Qwen3.5-9B作为新一代多模态大模型，在企业级应用中展现出卓越的性能表现。本教程将手把手指导您完成从基础部署到企业级生产环境的完整配置流程，涵盖Nginx反向代理、HTTPS安全加密以及负载均衡等高可用方案。

2. 环境准备与基础部署

2.1 系统要求

操作系统：Ubuntu 20.04/22.04 LTS
GPU配置：NVIDIA显卡（建议RTX 3090及以上）
CUDA版本：11.7或更高
内存要求：至少32GB RAM
存储空间：50GB可用空间

2.2 基础服务安装

# 安装Python环境 sudo apt update sudo apt install -y python3.10 python3.10-venv python3.10-dev # 安装CUDA Toolkit wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-ubuntu2204.pin sudo mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600 sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/3bf863cc.pub sudo add-apt-repository "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/ /" sudo apt-get update sudo apt-get -y install cuda # 验证安装 nvidia-smi

2.3 模型服务启动

# 创建虚拟环境 python3.10 -m venv qwen-env source qwen-env/bin/activate # 安装依赖 pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118 pip install gradio transformers accelerate # 启动基础服务 python /root/Qwen3.5-9B/app.py

3. Nginx反向代理配置

3.1 安装Nginx

sudo apt install -y nginx sudo systemctl start nginx sudo systemctl enable nginx

3.2 配置反向代理

创建配置文件/etc/nginx/conf.d/qwen.conf：

server { listen 80; server_name your-domain.com; location / { proxy_pass http://localhost:7860; 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_set_header X-Forwarded-Proto $scheme; # WebSocket支持 proxy_http_version 1.1; proxy_set_header Upgrade $http_upgrade; proxy_set_header Connection "upgrade"; } # 静态文件缓存 location /static/ { alias /path/to/static/files; expires 30d; access_log off; } }

3.3 测试并重载配置

sudo nginx -t sudo systemctl reload nginx

4. HTTPS安全配置

4.1 获取SSL证书

sudo apt install -y certbot python3-certbot-nginx sudo certbot --nginx -d your-domain.com

4.2 自动续期配置

sudo certbot renew --dry-run

4.3 强化安全配置

更新Nginx配置：

server { listen 443 ssl http2; server_name your-domain.com; ssl_certificate /etc/letsencrypt/live/your-domain.com/fullchain.pem; ssl_certificate_key /etc/letsencrypt/live/your-domain.com/privkey.pem; # 安全协议配置 ssl_protocols TLSv1.2 TLSv1.3; ssl_ciphers 'ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES128-GCM-SHA256...'; ssl_prefer_server_ciphers on; ssl_session_cache shared:SSL:10m; ssl_session_timeout 10m; # 其他配置同前... }

5. 负载均衡配置

5.1 多实例部署

在不同服务器或容器中启动多个模型实例：

# 实例1 CUDA_VISIBLE_DEVICES=0 python /root/Qwen3.5-9B/app.py --port 7860 # 实例2 CUDA_VISIBLE_DEVICES=1 python /root/Qwen3.5-9B/app.py --port 7861

5.2 Nginx负载均衡配置

更新/etc/nginx/conf.d/qwen.conf：

upstream qwen_servers { server 127.0.0.1:7860; server 127.0.0.1:7861; # 可添加更多服务器 keepalive 32; } server { listen 443 ssl http2; server_name your-domain.com; # SSL配置同前... location / { proxy_pass http://qwen_servers; # 其他代理配置同前... # 负载均衡策略 proxy_next_upstream error timeout http_500 http_502 http_503 http_504; } }

5.3 健康检查配置

location /health-check { proxy_pass http://qwen_servers/health; health_check interval=10 fails=3 passes=2; }

6. 性能优化与监控

6.1 模型推理优化

# 使用vLLM加速 pip install vllm python -m vllm.entrypoints.api_server --model unsloth/Qwen3.5-9B --tensor-parallel-size 2

6.2 Nginx性能调优

# 在http块中添加 worker_processes auto; worker_rlimit_nofile 100000; events { worker_connections 4000; use epoll; multi_accept on; } http { # 其他配置... keepalive_timeout 30; keepalive_requests 100000; sendfile on; tcp_nopush on; tcp_nodelay on; }

6.3 监控配置

# 安装Prometheus监控 wget https://github.com/prometheus/prometheus/releases/download/v2.47.0/prometheus-2.47.0.linux-amd64.tar.gz tar xvfz prometheus-*.tar.gz cd prometheus-*/ # 配置监控目标 echo " - job_name: 'qwen' static_configs: - targets: ['localhost:7860', 'localhost:7861'] " >> prometheus.yml # 启动Prometheus ./prometheus --config.file=prometheus.yml