Why build with Frontier?
Understand the fundamental differences between Frontier and traditional reverse proxies or message queues, and see why it excels in edge computing scenarios.
Structural Differences
| Capability | Frontier | Reverse Proxy (Nginx/Envoy) | Message Queue (Kafka/RabbitMQ) |
|---|---|---|---|
| Full-Duplex Native | Yes (Bi-directional RPC) | Client to Server only | Pub/Sub only, no RPC |
| Edge Node Presence | Built-in Online State | Stateless | Topic-based only |
| Point-to-Point Stream | Direct multiplexed streams | Proxy forwarding only | Not supported |
| Control Plane API | Query online nodes & state | Config management only | Consumer group metrics only |
Purpose-built for Edge Scenarios
IoT & Edge Devices
Maintain massive numbers of long connections. Services need to push commands to specific devices (RPC) and devices need to upload telemetry (Messaging) efficiently.
Remote Access & Tunneling
Need to expose a service behind NAT? Devices connect out to Frontier, and services can open a point-to-point stream to the device safely behind the firewall.
Realtime Chat & Sync
Clients maintain persistent presence. Services can route messages to specific clients based on Edge ID, with guaranteed delivery acknowledgment built into the protocol.
Multi-tenant Agents
Deploying agents into customer VPCs? Use Frontier as the central control plane where all agents connect, allowing cloud services to manage them securely without exposing internal ports.