Wardengate
Guide

Deployment models

Wardengate runs on your infrastructure — from a single-node pilot to multi-region gateway fleets. Pick the topology that matches your latency, availability, and compliance requirements.

Standalone pilot

One control plane node and one gateway connector on a single VM or small Kubernetes cluster. Ideal for proof-of-concept, lab environments, and teams replacing a regional bastion. Helm or Docker Compose deploys in under an hour; first brokered session typically follows the same day.

Active / standby

Two control plane instances with shared PostgreSQL and object storage. The standby node takes over API and scheduler duties if the primary fails. Gateway connectors reconnect automatically. Fits regulated workloads that need RTO measured in minutes without full horizontal scale.

Horizontal gateway scale

Multiple gateway connectors behind a load balancer, one authoritative control plane. Sessions stick to a connector for their duration; new sessions distribute across healthy nodes. Use when concurrent operator count or protocol throughput exceeds a single connector's capacity — common in large SRE and DBA teams.

Multi-region gateways

Gateway connectors in each region or cloud account, centralized policy and audit. Operators connect to the nearest connector; recordings and metadata replicate to your compliance region. Avoids backhauling SSH and RDP through a single geography while keeping one identity model and one evidence store.

Hybrid and air-gapped

Control plane on-prem with cloud gateway connectors — or fully offline installs with periodic evidence export. Wardengate does not require outbound SaaS connectivity. Offline bundles and side-channel updates support estates that cannot phone home.

Operational docs

Ready to deploy? Continue in documentation

Ready to evaluate?

See the platform on your architecture

Walk through gateway brokering, recording, and audit exports in a working session — or start with the interactive demo.