Network Insights

Network Insights reads your discovered network topology and surfaces structural risk you’d otherwise have to spot by eye: which devices, if they failed, would cut off part of your network; which devices have no second path in; and which devices would take the most other things down with them.

Where to find it

Open Explore → Insights. The page analyzes your current topology automatically — there’s nothing to configure or run.

If GridNMS hasn’t discovered any topology yet, the page shows nothing. Once device links start coming in (see Graph Explorer for how those links are learned), findings appear on their own.

The three findings

Card	What it flags
Single points of failure	A device whose failure would split part of your network into disconnected pieces — a “choke point” with no alternate path around it.
Single-homed devices	A device reachable over only one link. One cable, port, or upstream failure isolates it completely.
Highest blast radius	The devices with the most other devices sitting downstream of them — good candidates to prioritize for monitoring and redundancy, since losing one affects the most.

Each card shows a count and a short list of the devices behind it. Click a row to expand it and see exactly which devices are affected — the ones that would lose connectivity, the one device it depends on, or its downstream dependents, depending on the card — each one a link straight to that device’s detail page. A row with nothing to expand links directly to the device instead.

When nothing is flagged, the page shows an all-clear message instead of empty cards.

How these are computed

Click How these insights are computed at the bottom of the page for the full explanation. In short: everything here comes from the network topology GridNMS has actually discovered — device-to-device links learned via LLDP or CDP, plus any you’ve added manually. No extra probing runs to produce these findings; it’s a structural read of the map you already have.

Because of that, the findings are only as complete as your discovered topology. A device flagged as a single point of failure might have a real backup path that GridNMS simply can’t see yet — for example if neighbor discovery is turned off somewhere along that path, or the path runs through an unmonitored device. Treat the counts as a well-informed estimate, and confirm against your own knowledge of the network before acting on one.

Where to go next

• See the same topology-and-resilience analysis surfaced as a recommendation feed, alongside cost, coverage, and reliability findings, in Grid Advisor.
• Explore the underlying map yourself, including exactly how devices are connected, in Graph Explorer.
• Chart metrics for a flagged device directly with Metric Explorer.