MeshNetworkX

Networkx Graphs synced across devices using Zenoh.
Explore the docs »

Report Bug · Request Feature

Quickstart

• run zenohd -c .zenoh/meshnetworkx.json5 to create a storage
• or docker compose up zenoh-host to create a storage.

then run the following pip install -e ".[examples]" to install the package.

Then run an example with python examples/main.py

(back to top)

Why?

• If you model your application domain as a graph, this package makes it easy to run on multiple devices and automatically sync the graph between them.
• It is a drop-in replacement for networkx, so you can use all the networkx methods on the graph.

(back to top)

How?

• It is an abstraction on top of the Zenoh protocol, which is a very efficient protocol for IoT applications.
• The graph is stored in a zenoh storage, for more information on how zenoh keeps storages alligned (even if the network becomes partitioned and then later reconnects), check here.

(back to top)

Example

Assuming:

• you have two devices, and you want to create a graph on one device and access it on the other device.
• you have a zenoh storage running on both devices.
• Both devices can find each other using zenoh discovery (local network or using a Zenoh Router).

On device one:

import meshnetworkx as mx

G = mx.Graph()
G.add_node(1, color="red")
G.add_edge(1, 2, color="pink")

After that on device two:

import meshnetworkx as mx

G = mx.Graph()
print(G.nodes(data=True))
>>> [(1, {'color': 'red'}), (2, {})]

print(G.edges(data=True))
>>> [(1, 2, {'color': 'pink'})]

(back to top)