This is a series of videos on how we built the mesh project. In this first video, we talk about the mesh balancing system. If every node has 20 connections, how to select what connections to make to the total graph is balanced. This is the basic fundamentals on the mesh network.

In this second video, we talk about information distribution and security. Basically, if you have many nodes on a network, how can you get information about all of the other nodes by talking to only a few? Signed node information will hop from one node to another with each node verifying that the information was not changed along the way. Each node will have its certificate for authentication and while the certificate is not an indication that you can trust a node (it’s not signed by a root), it guaranties that nodes can’t be impersonated and allows for the creation of trust triangles. Depending on the application, different trust models may apply, but the mesh lays out the ground works for authenticating and securing communication.

Now we talk about inter-agent communication. How to agent communicate with each other and do it securely. We show that mutually-authenticated TLS is used to compute the node identifier every node. We also share keys so that most of the traffic is UDP based.

In this last video we look at one of the most important usages of the mesh, scalable and reliable communication with a web service. Only one node needs to keep a permanent connection to the web service for each group of nodes in a mesh network. We scale better and keep the over all system robust and secure.