Show HN: CloudyWithAChanceOfLatency: testing the water for a net monitoring app

Hello HN!

I'm building an umbrella-app for many network-related tools I've accumulated.

The architecture: A fleet of programmable probes run tests and forward all traffic sent/received to a service that analyses it.

So far so good.

I'm unsure of the UI though.

I like (isometric) maps more than the usual dashboards to display data, therefore I'm trying to design the app around that.

For the layout, I throw all possible info into a custom optimizer.

As the number of edges in the graph grew (currently about half a million), ideas that seemed to work fine initially started to fall apart and the layout and representation of this graph has become a major source of frustration and disappointment.

I've been running this UI experiment with a single traceroute-like measurement between several datacenters. Don't look too closely at the numbers as most of the actual computations are not implemented properly yet. I'm not worried about that, I'm worried about the UI.

I have other ideas of representation but web front-end takes an awful lot of time and quite frankly that's not my specialty.

If you could have a look at that demo and picture what I'm trying to achieve despite the rough edges, and tell me what you think of this approach, I'd really appreciate it. (but please take turns as it's all running on a single old server :))

Interesting? Boring? Hopeless? Déjà vu?