Preparing Meshcore for a grid or network outage¶
MeshCore can be valuable when the infrastructure you normally rely on is unavailable. During a power grid or network outage, there may be no opportunity to install software, configure devices, or troubleshoot network issues. Prepare your nodes, repeaters, and region scopes ahead of time, then test your mesh regularly under realistic conditions. Verifying that devices can communicate before an emergency ensures the network is ready when you need it most, rather than discovering problems after traditional communications have already failed.
Here we will discuss the following:
- Coordinating your repeaters with other repeaters in the area
- Redundant equipment
- Regularly scheduled testing
Coordinate your repeaters with others in the area¶
All Repeaters need region scopes¶
Mesh networks become gradually less effective as network traffic increases. During an outage, mesh traffic may spike considerably, and it may become difficult to use the mesh at all.
In our area, it is possible to communicate over the mesh with people in several nearby states. The long range messages are fun and exciting, but during an outage the extra traffic from other states may become a detriment to local communications.
Meshcore regions are designed for this scenario. If we have pre-installed regions in all our local repeaters, then we can temporarily limit outside traffic on a few key repeaters (think of them as 'transit repeaters'), and restore local mesh communications.
Coordination between repeater owners is key, every repeater should have the regions installed or the mesh will be degraded when switching to the local region.
Regularly testing region scopes with a scoped channel¶
Meshcore channels can have a region scope applied to them, so any messages in that channel are sent to that region. This is a great way to test how well our region scope is working with our repeaters. As more repeaters get the local regions configured, the scoped channel will begin to function better and better. If we ever need to switch to a local region, the region scoped channel will already be working and ready to use.
You can add the #va-scoped test channel from our Meshcore channels page.
Planning to limit unscoped traffic¶
In an outage scenario, it may be necessary to limit unscoped traffic on transit repeaters. These are the mountain top repeaters that bring outside traffic into the area. Since these are the only repeaters receiving traffic from outside the area, this will effectively limit all outside traffic from our region.
Having backup transit repeater admins is probably a good idea in case some of them are not available during the outage. Transit admins should share repeater passwords ahead of time and discuss what conditions should be required before limiting unscoped traffic.
Transit admins should try to notify everyone in the area to switch to the scoped channel as discussed above, or any other scoped channel, until the change is reverted.
This scenario should be tested from time to time as well. Limitting unscoped traffic will also limit unscoped channels, which may confuse some users. Only the channels with a region scope applied will work until the change is reverted.
To limit unscoped flood messages on a repeater, decide how many hops away to accept unscoped messages, perhaps 4-6 hops, and set that via the command line:
Be sure the repeater firmware is at least 1.16 for the flood.max.unscoped feature.
To revert, and allow unscoped messages from anywhere:
Redundant equipment¶
Repeaters are usually outdoors in locations vulnerable to bad weather. Some repeaters may be damaged during extreme weather events, so having extra repeaters covering the same area is a good idea.
If the internet is out, it may not be easy to flash companion nodes for other people. Have a few extra companion nodes already setup and configured to share with others.
Install the meshcore app on family devices, even if they don't use it. The app may not be available for installation during a power outage. (other ideas here?)
What else?
Regularly scheduled testing¶
Regularly scheduled check-ins can help make sure everything is working.
Things you can mention at the check-in:
- The website used for coordination
- Reminder to send a flood advert so others can save it as a contact
- Mention the scoped channel and send a test message there
- Reply to users checking-in so they know they are received