Complete guide to choosing and implementing the ideal deployment method for your Cadence installation.

Quick Comparison

Method	Best For	Setup Time	Scalability	Expertise	Cost
Binary	Single server, testing, quick setup	5 min	Single host	Minimal	Free
Docker	Cloud-native, local development, CI/CD	10 min	Multi-container	Basic	Free (+ infra)
systemd	Linux servers, single instance, production	15 min	Single host	Medium	Free (+ CPU)
Kubernetes	Enterprise, HA, auto-scaling, multi-region	30 min	Unlimited	Advanced	Free (+ infra)

Decision Tree

What scale do you need?

├─ Testing/Development
│  ├─ Quick start? → Binary ✅
│  └─ Local containers? → Docker Compose ✅
│
├─ Production - Single Server
│  ├─ Max simplicity? → Binary
│  └─ Full management? → systemd ✅
│
├─ Production - Multiple Servers
│  ├─ Simple setup? → Docker + Load Balancer
│  └─ Enterprise grade? → Kubernetes ✅
│
└─ High Availability / Global
   └─ Multi-region? → Kubernetes + Helm ✅

Method Overview

1. Binary Installation

When: Testing, single server, simple deployments

Pros:

Minimal dependencies
Fast startup (~1 second)
Direct OS integration
Easy debugging

Cons:

Single process only
Manual restart management
Limited to one server

Setup: Binary Installation

2. Docker

When: Cloud deployments, containerized workflow, development

Pros:

Environment isolation
Easy versioning
Multi-container orchestration
Reproducible across machines

Cons:

Slight overhead (~50MB RAM)
Requires Docker knowledge
Port management complexity

Setup: Docker Deployment

3. systemd Service

When: Linux production servers, single instance, traditional ops

Pros:

Deep Linux integration
Automatic restarts
Resource limits built-in
Simple monitoring

Cons:

Linux-only
Single instance
Manual scaling

Setup: systemd Service

4. Kubernetes

When: Enterprise, High Availability, auto-scaling, multiple data centers

Pros:

Unlimited horizontal scaling
Automatic failover
Self-healing
Multi-region support

Cons:

Steep learning curve
Infrastructure overhead
Complexity for small deployments

Setup: Kubernetes Deployment

Implementation Guides

Binary Installation - Quick 5-minute setup
Docker Deployment - Containerized deployment
systemd Service - Linux service management
Kubernetes - Enterprise-grade deployment

Deployment Scenarios

Scenario 1: Quick Testing

Bash

# Just want to test Cadence?
→ Use Binary Installation
→ Download precompiled binary
→ Run locally or on test server
→ Time: 5 minutes

Scenario 2: Single Production Server

Bash

# One server, production workload?
→ Use systemd Service
→ Install as system service
→ Full monitoring and restarts
→ Resources: 2+ cores, 2GB RAM
→ Time: 15 minutes

Scenario 3: Small Team, Development

Bash

# Multiple developers, local testing?
→ Use Docker Compose
→ Shared docker-compose.yml
→ Consistent environment
→ Time: 10 minutes

Scenario 4: Cloud Native (AWS/GCP/Azure)

Bash

# Scalable cloud deployment?
→ Use Docker + Kubernetes
→ Container registry (ECR/GCR/ACR)
→ Auto-scaling
→ Time: 30-60 minutes

Scenario 5: Global High Availability

Bash

# Multiple regions, 99.9% uptime?
→ Use Kubernetes with Helm
→ Multi-region clusters
→ Automatic failover
→ Time: 2-3 hours

Next Steps

Choose a deployment method based on your needs
Follow the implementation guide
Review Operations Guide for production setup
Set up Monitoring

Deployment Methods