Network critical paths
Flow model and attack simulation run in Erlang. This page is Elixir LiveView.
Demo data
{:not_supported, :azure, :edges} — showing the sample network
Critical path from internet to sql-db
The route with the least resistance, where most traffic flows. This is the path to defend first.
internet
front-door
app-gateway
web-subnet
api-subnet
sql-db
Links
Trunks carry most of the flow; branches carry the rest. Block a link to test a response.
| Link | Traffic | Role | |
|---|---|---|---|
| internet → front-door critical |
22.5%
|
trunk | |
| front-door → app-gateway critical |
21.3%
|
trunk | |
| app-gateway → web-subnet critical |
20.1%
|
trunk | |
| web-subnet → api-subnet critical |
14.7%
|
trunk | |
| api-subnet → sql-db critical |
10.2%
|
trunk | |
| api-subnet → storage |
3.6%
|
branch | |
| api-subnet → ai-endpoint |
2.2%
|
branch | |
| internet → vpn-gateway |
1.7%
|
branch | |
| vpn-gateway → mgmt-subnet |
1.6%
|
branch | |
| web-subnet → storage |
1.0%
|
branch | |
| mgmt-subnet → domain-controller |
0.7%
|
branch | |
| mgmt-subnet → web-subnet |
0.3%
|
branch | |
| domain-controller → sql-db |
0.1%
|
branch | |
| mgmt-subnet → sql-db |
0.1%
|
branch |
Monte Carlo attack simulation
Each simulated attacker starts at internet, prefers busy links, and may be detected at every step.
Demo data. Replace the traffic list in network_live.ex with flows from Azure VNet flow logs.