Hermes Agent: Deployment Modes

Q: How to monitor deployment status?

hermes status shows running state, /health API health check, Docker uses docker ps .

Last updated: 2026-08-31

Deployment mode determines where Hermes Agent runs — your local laptop, a Docker container, a cloud server, or a cluster. Choose the right mode for stability and cost-effectiveness.

💡 Tip: Hermes supports four deployment modes: local standalone, Docker container, cloud server, and cluster mode. Local mode is recommended for individuals; Docker or cloud server for teams.

📋 Prerequisites: Lesson 2 Installation, Lesson 3 Configuration File

1. What You Will Learn

#	Content
❶	Four deployment modes comparison
❷	Local mode configuration
❸	Docker deployment
❹	Cloud server deployment
❺	Cluster mode and scaling

2. Story

(1) Pain Point: Choosing the Wrong Deployment

Bob runs Hermes locally, but when his computer shuts down, the Agent goes offline. The team needs a 24/7 online Agent.

(2) Solution: Choose Deployment by Scenario

BASH

# Alice's choice:
# Development & debugging → Local mode (cost-effective, convenient)
# Team use → Docker on server (24/7 online)
# High concurrency → Cluster mode (horizontal scaling)

3. Four Deployment Modes Comparison

Dimension	Local Mode	Docker	Cloud Server	Cluster Mode
Use Case	Individual dev	Team/Small business	Medium business	Large scale
Uptime	When powered on	24/7	24/7	24/7
Scalability	❌	✅ Vertical	✅ Vertical	✅ Horizontal
Data Security	✅ Local	✅ Local	⚠️ Cloud	⚠️ Distributed
Cost	Free	Server cost	Cloud fees	High
Ops Difficulty	⭐	⭐⭐	⭐⭐⭐	⭐⭐⭐⭐

4. Local Mode

(1) Standard Local Installation

BASH

# Install
pip install hermes-agent[all]

# Start
hermes serve --host 0.0.0.0 --port 8080

# Background run (Linux/macOS)
nohup hermes serve &

# Using systemd (Linux)
cat > /etc/systemd/system/hermes.service << EOF
[Unit]
Description=Hermes Agent
After=network.target

[Service]
Type=simple
User=alice
WorkingDirectory=/home/alice
ExecStart=/usr/local/bin/hermes serve
Restart=always
RestartSec=10

[Install]
WantedBy=multi-user.target
EOF

systemctl enable hermes
systemctl start hermes

(2) Local Mode Configuration

YAML

# ~/.hermes/config.yaml
deployment:
  mode: "local"
  host: "0.0.0.0"
  port: 8080
  
  # Local model support (offline operation)
  local_model:
    enabled: true
    provider: "ollama"
    fallback_to_cloud: true

5. Docker Deployment

(1) Single Container

BASH

# Pull image
docker pull nousresearch/hermes-agent:latest

# Run
docker run -d \
  --name hermes \
  -p 8080:8080 \
  -v ~/.hermes:/root/.hermes \
  -e OPENAI_API_KEY="${OPENAI_API_KEY}" \
  --restart unless-stopped \
  nousresearch/hermes-agent:latest

(2) Docker Compose (Recommended)

YAML

# docker-compose.yml
version: "3.8"

services:
  hermes:
    image: nousresearch/hermes-agent:latest
    container_name: hermes-agent
    restart: unless-stopped
    ports:
      - "8080:8080"
    volumes:
      - ./data:/root/.hermes
      - ./projects:/root/projects
    environment:
      - OPENAI_API_KEY=${OPENAI_API_KEY}
      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}
      - HERMES_MODEL_DEFAULT=gpt-4o
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8080/health"]
      interval: 30s
      timeout: 10s
      retries: 3

  # Ollama local model
  ollama:
    image: ollama/ollama:latest
    container_name: ollama
    restart: unless-stopped
    ports:
      - "11434:11434"
    volumes:
      - ./ollama-data:/root/.ollama
    deploy:
      resources:
        reservations:
          devices:
            - capabilities: ["gpu"]

  # Web UI
  web-ui:
    image: nousresearch/hermes-web:latest
    container_name: hermes-web
    restart: unless-stopped
    ports:
      - "3000:3000"
    environment:
      - HERMES_API_URL=http://hermes:8080

6. Cloud Server Deployment

(1) AWS Deployment

BASH

# Using AWS EC2
# 1. Launch EC2 instance (t3.medium recommended)
# 2. Install Docker
# 3. Run Hermes

ssh ec2-user@your-instance

# Install Docker
sudo yum install -y docker
sudo systemctl start docker

# Run Hermes
docker run -d \
  -p 8080:8080 \
  -v /data/hermes:/root/.hermes \
  -e OPENAI_API_KEY="sk-..." \
  --restart unless-stopped \
  nousresearch/hermes-agent:latest

(2) Security Configuration

YAML

# Cloud deployment security hardening
security:
  # API authentication
  api:
    auth_enabled: true
    auth_type: "jwt"           # jwt / api_key / oauth
    jwt_secret: "${JWT_SECRET}"
    rate_limit: 100            # RPM

  # TLS
  tls:
    enabled: true
    cert: "/etc/ssl/hermes.crt"
    key: "/etc/ssl/hermes.key"

  # Network isolation
  network:
    allowed_origins:
      - "https://your-domain.com"
    blocked_ips: []

7. Cluster Mode

(1) Architecture Diagram

graph TB
    LB[Load Balancer] --> W1[Worker 1]
    LB --> W2[Worker 2]
    LB --> W3[Worker 3]
    
    W1 --> Redis[(Redis<br/>Shared Memory)]
    W2 --> Redis
    W3 --> Redis
    
    W1 --> DB[(PostgreSQL<br/>Persistent Storage)]
    W2 --> DB
    W3 --> DB
    
    W1 --> MQ[RabbitMQ<br/>Task Queue]
    W2 --> MQ
    W3 --> MQ

(2) Cluster Configuration

YAML

# docker-compose.cluster.yml
version: "3.8"

services:
  hermes-master:
    image: nousresearch/hermes-agent:latest
    environment:
      - HERMES_ROLE=master
      - HERMES_WORKERS=3
      - REDIS_URL=redis://redis:6379
      - DATABASE_URL=postgresql://hermes:pass@postgres:5432/hermes
    depends_on:
      - redis
      - postgres

  hermes-worker:
    image: nousresearch/hermes-agent:latest
    environment:
      - HERMES_ROLE=worker
      - REDIS_URL=redis://redis:6379
    deploy:
      replicas: 3
    depends_on:
      - redis

  redis:
    image: redis:7-alpine
    volumes:
      - redis-data:/data

  postgres:
    image: postgres:16-alpine
    environment:
      - POSTGRES_DB=hermes
      - POSTGRES_USER=hermes
      - POSTGRES_PASSWORD=pass
    volumes:
      - pg-data:/var/lib/postgresql/data

❓ FAQ

Q Which mode for individual users?

A Local mode. Simple, free, best privacy. Consider Docker when you need 24/7 availability.

Q Docker vs direct install?

A Docker offers better isolation, easier migration and upgrades. Direct install is lighter with lower latency. Choose based on your needs.

Q Is cloud deployment data secure?

A API Keys are encrypted, TLS encrypts transmission, optional encrypted disks. But data is on the cloud — for high compliance requirements, local deployment is recommended.

Q When do I need cluster mode?

A When concurrent users >50, requests per minute >100, or high availability is required. Small teams don't need it.

Q How to monitor deployment status?

A hermes status shows running state, /health API health check, Docker uses docker ps.

Q How to upgrade smoothly?

A Docker is easiest — pull new image and restart. Local install uses pip install --upgrade. Cluster mode uses rolling upgrades.

📖 Summary

Four modes: Local (individual) → Docker (team) → Cloud server (enterprise) → Cluster (large scale)
Local mode is zero cost, most private, suitable for development and testing
Docker Compose one-click deployment, includes Ollama and Web UI
Cloud server requires security hardening: authentication, TLS, network isolation
Cluster mode uses Redis + PostgreSQL for shared memory and high availability

📝 Exercises

Basic (⭐): Start Hermes in local mode, connect to Telegram platform, verify stable operation for 24 hours.
Intermediate (⭐⭐): Use Docker Compose to deploy Hermes + Ollama for local offline inference.
Advanced (⭐⭐⭐): Deploy Hermes on a cloud server, configure TLS and JWT authentication for secure remote access.

Web-Tutorial Tech Team

A team of developers maintaining programming tutorials. Each tutorial is written and reviewed by developers with expertise in that field. We work to keep our content accurate and reliable — if you spot an issue, please let us know.

About Us

Hermes Agent: Deployment Modes

1. What You Will Learn

2. Story

(1) Pain Point: Choosing the Wrong Deployment

(2) Solution: Choose Deployment by Scenario

3. Four Deployment Modes Comparison

4. Local Mode

(1) Standard Local Installation

(2) Local Mode Configuration

5. Docker Deployment

(1) Single Container

(2) Docker Compose (Recommended)

6. Cloud Server Deployment

(1) AWS Deployment

(2) Security Configuration

7. Cluster Mode

(1) Architecture Diagram

(2) Cluster Configuration

❓ FAQ

📖 Summary

📝 Exercises

🙏 帮我们做得更好