Health score

The health score is a number from 0 to 100 that sums up how an environment is doing right now. Curade Mini works it out itself from the problems open in the environment, so it is not a figure your monitoring platform publishes.

How the score is worked out

1. The app reads the problems open right now, whatever time window the page shows. A problem that started before the window and is still open counts, and one that closed inside the window does not.
2. It splits them in two: application and service problems on one side, infrastructure and environment problems on the other.
3. In each group, every open problem takes a penalty by its severity, and the penalty for each severity is capped, so one noisy kind of problem cannot swamp the rest. The group's score is 100 minus its capped penalties, and never falls below zero.
4. The overall score takes 70% of the service score and 30% of the infrastructure score, since what users experience counts for more than what a host is doing.

Reading the card

The health card on the dashboard has three tabs: Overall, Service and Infrastructure. Its badge names the band the score is in:

Tap the card to open the breakdown: each group's score, its problem count and its penalty, the active problems, the services and hosts they affect, and for each severity how many problems are open, the penalty each carries and the cap. How is this score calculated? shows the sums with your own numbers.

A clean environment

When the tab in view has no open problems, a green All good, no open problems strip sits above the score, naming what is clear. The strip stays away while monitoring data is unavailable, since a score built on missing data has no clean state to claim.

A score of zero

A zero means the capped penalties add up to 100 or more: enough weighted problems are open to use up the whole score. It does not mean the environment is offline or that monitoring has stopped.

When the service score is perfect and the overall score is still low, the whole loss comes from infrastructure problems: users may not feel anything yet while hosts or processes are struggling. The card says so when it happens.

Why it differs from your monitoring platform

Your monitoring platform does not publish this number. The score is Curade Mini's own, with its own weighting, and it always reads the whole environment's problems open right now, whatever the time window and the dashboard filters say. Comparing it with a figure from another tool compares two different calculations.