Building Redundancy Into a Regional Mesh

Redundancy means the network can lose a node, path or supporting service without losing its essential coverage. It is not achieved by buying duplicate radios and placing them together. Real resilience requires independent routes, power sources, sites, maintainers and recovery plans.

Find the points that can divide the network

Draw every fixed infrastructure node and every link that has been verified in both directions. Remove one node at a time on the drawing. If a town or large coverage area becomes isolated, that node is a critical articulation point. Do the same with links, power systems and hosts.

Start redundancy work at the points with the greatest consequence of failure, not at the easiest place to install another box.

Four kinds of redundancy

Type	What it protects against	Example
Path	Blocked or failed radio link	A second ridge route between two towns
Site	Loss of a building, tower or host	Coverage from separate properties
Power	Mains or solar failure	Battery-backed site plus independent solar site
Operational	Unavailable maintainer or lost configuration	Two authorized stewards and stored backups

Internet and MQTT redundancy is separate. A public Meshtastic network should retain local RF function when a broker or internet connection fails.

Prefer geographically independent paths

Two nodes on the same tower share wind, lightning, access and permission risks. Two towers on the same ridge may share wildfire, ice or power exposure. The strongest alternative route often crosses a different ridge or uses a chain of lower community sites.

Independent does not mean equal. A backup path may have less coverage or capacity, provided it preserves essential connectivity when the primary path fails.

Avoid redundancy that creates congestion

Meshtastic uses managed flooding, so alternate routes may already hear and rebroadcast the same packet. Duplicate suppression limits repetition, but dense groups of aggressively configured infrastructure can still consume airtime. Use normal client behavior for most nodes, assign special roles only where justified and keep hop limits controlled.

Test the mesh under realistic load. A path that works for occasional traceroute packets may perform differently when many nodes broadcast position and telemetry.

Provide fade margin

A link that works only in ideal weather is not a resilient link. Improve antenna placement, shorten lossy coax and choose a path with margin above the minimum decodable signal. Test during rain, foliage, temperature changes and busy periods when possible. Record packet delivery over time, not only the best RSSI screenshot.

Protect the supporting systems

Use batteries or an uninterruptible supply at critical mains-powered sites. Size solar for the worst season and monitor voltage. Protect outdoor cables and enclosures, provide grounding and surge protection, and keep vulnerable connectors accessible for inspection.

Operational resilience includes configuration backups, spare programmed radios, antenna and cable spares, documented access and more than one person able to maintain the system. A technically redundant mesh can still fail if the only administrator is unavailable.

Test failure deliberately

Do not wait for a storm to learn whether the backup route exists. During a planned maintenance window, switch off a critical node and verify:

which areas remain connected;
whether hop counts remain acceptable;
whether message delivery becomes slower or unreliable;
whether monitoring correctly reports the failure; and
whether users know the degraded network status.

Meshtastic’s Traceroute module can reveal the first returned route for a test, but packets can reach destinations by multiple paths. Combine traceroute with delivery testing and direct observation.

Use a simple resilience classification

Experimental: useful when available, no failure commitment.
Maintained: documented owner, monitoring and reasonable repair.
Redundant: important areas retain tested alternate RF paths.
Operationally resilient: alternate paths, independent support systems, spares and multiple maintainers.

This classification helps MeshAtlas users understand what a map line actually represents without pretending that a volunteer mesh has formal carrier availability.

Prioritize restoration

Define which sites should be restored first and keep the necessary equipment ready. If a remote summit cannot be accessed in winter, plan a temporary lower substitute or accept and publish the seasonal limitation. Honest limitations build more trust than nominal redundancy that has never been tested.