Maintaining a Public Meshtastic Network

How a community turns installed nodes into a dependable, understandable and sustainable shared service.

Define what the network promises

A volunteer Meshtastic network should not pretend to be a mobile operator. It can still be well run. State the intended service area, supported modem preset, public channel policy, expected infrastructure availability and whether any internet bridges are part of the design. Users can then distinguish planned infrastructure from experimental, mobile and personal nodes.

Availability targets should match resources. “Best effort, monitored weekly” is more honest than claiming permanent coverage that nobody verifies. Critical use cases must retain alternative communications; Meshtastic infrastructure can support a community without becoming its only safety system.

Create an inventory

Maintain one authoritative inventory for permanent nodes. Give each installation a stable site identifier separate from the radio node number, because boards can be replaced. Record the owner, host, maintainer, hardware, antenna, power source, firmware, configuration version, public map record, commissioning evidence and last physical inspection.

Separate public and restricted fields. The public entry can show approximate location, coverage, status and contact route. Restricted records can include exact coordinates, keys, access procedure, gate codes, host details, risks and photographs that reveal the site.

Set a review rhythm

Interval	Typical work	Trigger for action
Daily or automatic	Telemetry, last-heard and alert review	Unexpected silence or unsafe power trend
Monthly	Compare stable links and map status	Persistent degradation or stale listing
Seasonal	Solar, battery, heat and weather review	Insufficient winter reserve or summer temperature
Annual	Physical inspection and inventory audit	Corrosion, movement, damage or unclear ownership
After severe weather	Targeted remote and visual checks	Changed pattern, enclosure damage or water ingress

Intervals are starting points, not rules. A difficult alpine site deserves more remote instrumentation and more conservative seasonal planning. An accessible indoor gateway can be serviced when evidence demands it.

Measure service, not only devices

A green power value does not prove useful coverage. Maintain several representative test paths: nearby access, a known edge sector, a link to the next infrastructure site and, where relevant, the path across a redundancy boundary. Compare delivery rate, direction, route and radio metrics with commissioning results.

Do not generate excessive automated test traffic on a shared channel. Use ordinary operational observations where possible and run controlled range or packet tests at agreed times. Disable test modules afterward.

Keep configuration disciplined

Public infrastructure should follow a documented baseline for region, preset, frequency slot, hop limit, role, rebroadcast mode, channels and broadcast intervals. Exceptions should name the problem they solve. “It seemed stronger” is not adequate justification for a router role, maximum hop limit or aggressive telemetry.

Before a change, export the current configuration and note the firmware. Test proposed settings on bench hardware. After a change, check local access, remote administration, public interoperability, telemetry and known links. Update the inventory so the next maintainer does not troubleshoot yesterday’s design.

Maintain the complete RF and power system

Radio boards are only one failure point. Inspect the antenna, connector, adapter, feed line, surge protection, grounding, enclosure, vent, glands, mount, battery, charge controller, panel and cable strain relief. Look for slow problems: corrosion, UV damage, water tracks, loosened clamps, animal damage, vegetation and condensation.

Trend battery voltage rather than reacting to a single report. A solar battery that reaches full charge every afternoon can still be losing capacity. Compare overnight decline, morning minimum, charging start and recovery after cloudy periods. Temperature matters because charge and discharge behavior changes with chemistry.

Respond proportionately to incidents

Classify incidents before dispatching anyone. A monitoring gap may be an observer problem. Degradation may allow a planned visit. A loose mast or damaged electrical installation may require the host to make the area safe immediately. Remote configuration mistakes may be recoverable through authorized administration; power, antenna and enclosure faults generally are not.

Record the symptom, evidence, actions and outcome. If a board is swapped, retain the failed unit for bench diagnosis when practical. Repeated unexplained replacements hide systemic faults such as poor sealing, incorrect charging or receiver overload.

Design for volunteer continuity

Use shared documentation and at least two authorized contacts for important sites. Keep recoverable configuration backups and a controlled record of administration authority. Agree what happens when a volunteer leaves. A node should not become inaccessible because one phone, laptop or personal account disappeared.

Budget for batteries, connectors, transport and replacement hardware. Donation of a radio does not fund years of operation. Where a host provides electricity or internet, document whether that support is indefinite, reimbursed or subject to review.

Know when to downgrade status

When evidence shows that an infrastructure node is unreliable, label it degraded. If it becomes silent, mark it offline while investigating. If access, ownership or repair is no longer realistic, retire the listing. Historical entries may remain useful, but they must be clearly distinguished from active infrastructure so planners do not build routes around ghosts.