Fibmesh Public IPsAvailable
Give supported devices and gateways a public address over your existing connection—for incoming access, outgoing identity, or both.
203.0.113.20Incoming and outgoing. Separate decisions.
A public address can receive connections, identify the source of outgoing traffic, or do both. Choose the scope before the route.
203.0.113.20Inbound access: permitted incoming connections and their replies use Fibmesh. Unrelated outgoing traffic keeps the existing ISP route.
Permitted incoming connections reach your service through Fibmesh. Their replies follow the matching return path. New, unrelated outgoing connections keep the existing ISP route.
Chosen destinations see the assigned Fibmesh source address. Other destinations keep the existing ISP route. Add separate inbound rules if the same server also needs to receive connections.
Eligible internet connections use the Fibmesh tunnel. Define local and platform exceptions, IPv4 and IPv6 handling, and what happens when the tunnel disconnects. Incoming access remains a separate choice.
Planning examples only; no routes or firewall settings are changed. 203.0.113.20 is a documentation address. Confirm the selected traffic policy, platform and gateway support during assisted setup.
Your access ISPStill carries the connection.
Your public addressSupplied through Fibmesh.
Your traffic policyIncoming and outgoing are separate.
Choose where the address belongs
Deliver an address to a supported endpoint, or reach selected LAN services through a gateway. Both work over supported existing connections.
Explore address delivery →A fully routed public address can be assigned to a supported endpoint without address translation. Keep the host firewall rules and its own authentication in place.
Explore Device IPs →Forward incoming connections to LAN services, or route eligible outgoing traffic through the gateway. Routing and NAT can keep equipment on its private addresses.
Shared IPv4 egress gives those LAN devices one public source identity; it does not assign a separate public IP to each device.
Explore Gateway IPs →Put the address to work
Start with the service, the people reaching it and the systems it calls.
Run an office server, local GPU or colocated workload where it makes sense. Separate private administration from public service traffic.
Your own infrastructure ↗Use an assigned source address for an agreed API or business-system allowlist. The destination’s login and access rules still apply.
Outbound allowlisting ↗Forward permitted connections to selected NVR, NAS and equipment services through a supported gateway. Keep the existing LAN addresses.
Equipment & IoT ↗Customer proof of concept
We supplied a preconfigured Fibmesh Edge device. The customer connected it to the existing network; we configured the targets remotely so NVRs and servers could be reached, without installing software on that equipment or replacing its router.
Proof-of-concept deployment. Fibmesh Edge hardware remains in development or pilot.
Use dual-stack IPv4 and IPv6, or IPv6-only, where the deployment supports it. Fully routed describes address delivery; it does not require full-tunnel internet routing. A NAT alternative is also available.
Managed inbound access starts closed. Limit permitted sources, protocols and service ports, and retain application authentication. Outgoing routes must select the intended source address. Gateway traffic must actually traverse that gateway.
Replies to incoming connections follow the Fibmesh return path. WireGuard protects the tunnel to the Fibmesh peer; application encryption remains a separate requirement. Translation alone does not choose a return route.
The reserved Fibmesh address is separate from the access ISP’s address. A supported tunnel reconnects over the new connection. Existing sessions may restart; this does not promise uninterrupted roaming or portability between serving regions.
Choose the application ports and permitted sources through the supported firewall configuration. Confirm protocol support and service or platform reservations. A routed IP is not an instruction to open every port.
No. The public address is delivered over your existing access connection. Its bandwidth, reliability and local conditions still matter. Throughput, traffic allowance, location and support are agreed separately.
The developer contract includes GET /v1/public-ip/assignments. Observed delivery and policy state need to be checked alongside an actual service test. Dedicated Public IP CLI commands remain to be implemented.
Your next connection
Tell us about the device or gateway, location, services and traffic you need. We’ll confirm the supported setup before arranging delivery.