| Knowing what to allow |
Drawn from observed traffic, with the owning process on every flow. |
Ask around, read the docs, guess, and find the rest in production. |
| Writing the rules |
Proposed for you and narrowed to the tightest form the evidence supports. |
By hand, per host, per port, forever. |
| Knowing what will break |
A list of every flow that would stop, before you enforce. |
You find out when someone calls. |
| Keeping them current |
Rules describe roles, so a rebuilt or re-addressed host keeps its policy. |
Every change is a manual edit somewhere, and the drift is invisible. |
| Proving it |
Coverage measured per observed flow, and an attributable change record. |
A spreadsheet, and a degree of faith. |
| Cost |
A$20 per server per month, after a 30-day trial. |
Free, plus however many days a year this actually takes you. |
| When it wins |
Once you have more hosts than you can hold in your head. |
A handful of hosts with a stable, well-understood set of dependencies. |