High Availability Deployment

Overview

The HA deployment extends the standard single-node docker-compose.yml with redundancy at every layer. It is designed for production environments where downtime and data loss are unacceptable.

Architecture

Nginx (reverse proxy, port 3000)
    |
    +--- API instance 1 (EntraGuard / Ravenscan / RedFox)
    +--- API instance 2 (EntraGuard / Ravenscan / RedFox)
    |
PostgreSQL primary  ---streaming replication---  PostgreSQL replica
    |
Redis Sentinel (3 nodes)
    |  master election + failover
Redis primary  ---replication---  Redis replica
    |
Neo4j (single instance)
    |
Loki  <---  Promtail (log collection from all containers)
    |
Grafana (dashboards, port 3001)

Components

Deployment

HA requires two servers (physical or virtual). The Setup Wizard guides you through choosing the role of each node, similar to Azure AD Connect's primary/secondary model.

Node 1 — Primary

1. Install on the primary server
   Install (primary)

   # Linux / macOS
   curl -fsSL https://install.coderaft.io | bash
   
   # Windows (PowerShell as Administrator)
   irm https://install.coderaft.io/win | iex

2. Activate and choose role
   Open http://localhost:3000 — in the Setup Wizard:

   • Activate your Enterprise licence
   • In the "High Availability" step, select "Primary node"
   • The wizard generates a join token — copy it for the secondary node
   • Configure database passwords and network settings
3. Start the HA stack
   Start HA (primary)

   docker compose -f docker-compose-ha.yml up -d --build

Node 2 — Secondary

1. Install on the secondary server
   Install (secondary)

   # Linux / macOS
   curl -fsSL https://install.coderaft.io | bash
   
   # Windows (PowerShell as Administrator)
   irm https://install.coderaft.io/win | iex

2. Join the primary
   Open http://localhost:3000 — in the Setup Wizard:

   • Select "Secondary node"
   • Enter the primary node IP/hostname (e.g., 192.168.1.10)
   • Paste the join token from the primary
   • The wizard validates connectivity and starts replication
3. Verify replication
   Verify replication

   # On the secondary node — check replication status
   docker compose -f docker-compose-ha.yml ps
   docker compose -f docker-compose-ha.yml logs postgres-replica --tail 20
   docker compose -f docker-compose-ha.yml exec redis-sentinel redis-cli -p 26379 SENTINEL master mymaster

Monitoring (Grafana)

Grafana is available at http://localhost:3001 (default credentials are set in the environment file). Pre-configured dashboards include:

• System overview — container health, resource usage, uptime
• Audit log stream — real-time view of WORM audit entries across all products
• Scan activity — EntraGuard and Ravenscan scan durations, success/failure rates
• Session monitoring — RedFox active sessions, connection counts, latency
• Database replication — PostgreSQL replication lag, Redis Sentinel status

SIEM integration

The Loki log pipeline can forward events to external SIEM systems. Configure log forwarding in Settings → Integrations or export from Grafana to:

Supported platforms

Both primary and secondary nodes can run on different operating systems. Docker is the only runtime dependency.

Limitations

• Neo4j is single-instance in the current HA setup (Neo4j clustering requires Enterprise Edition)
• PostgreSQL failover is manual — automatic promotion with Patroni is planned
• Kubernetes deployment (Helm charts) is planned but not yet available

Hardware requirements (HA)

Each node (primary and secondary) requires:

Need help with HA deployment? Contact [email protected]. For Enterprise licensing, reach out to [email protected].