Glossary

Exit classification

In one sentence

Exit classification is the act of deciding, when a dental application process closes, whether that closure was a real crash, a user-initiated quit, an OS-level shutdown, or an expected restart — and only treating real crashes as remediation events. CyberCore’s implementation uses 10 distinct exit signals so it does not react to non-events.

Last updated

Generic monitoring tools tend to alert on any unexpected process termination. In a dental practice, that produces a high false-positive rate: staff close Dentrix at the end of a shift, an OS update reboots a workstation, an intraoral scanner driver restarts after a firmware update. Treating every closure as a crash trains the practice owner to ignore alerts — the opposite of what monitoring is supposed to do.

A dental-trained exit classifier instead reads the closure across multiple signals: the OS-reported exit code, the process's last-logged state, the parent application's session lifecycle, whether the closure correlates with a graceful shutdown event, and whether the application was actively in use immediately before. Only the combinations that map to a real fault produce a remediation event.

The owner-visible artifact is a per-incident reason: not just "Dentrix exited at 14:32" but "Dentrix exited at 14:32 — classified as Lock-Timeout (signals 3, 7), remediation: clear database lock, status: completed in 4.2s."

Related

← All glossary terms

Ask Core AI