SIGNAL BOARD / DURABLE TRANSITIONS

Alert incidents

Small, typed rules over retained logs, expected heartbeat events, stable HTTP checks, and Simon's own bounded runtime health. Incidents record state changes—not raw payloads, credentials, or response bodies.

Rules: 0 / 64
Pending: 0
Firing: 0
Open: 0

01 / INCIDENT LEDGER

Transition record

0 open 0 resolved

No incidents recorded. Firing and recovery transitions will be retained here.

02 / INCIDENT DETAIL

No transition selected

When a typed condition fires, its safe evidence and complete transition history appear here.

04 / NEW SIGNAL

Define rule

Operator label Rule kind internal_health / selected above Severity

Evaluation and transition policy Evaluation / seconds 10–300 seconds

Transition debounce / fixed 1 breach / 1 healthy evaluation Canonical runtime states are immediate; initializing, stopping, and degraded storage remain neutral.

Cooldown / seconds 60–86400 seconds

Component health configuration

This rule reads one bounded in-process state snapshot. Quiet logs are not a failure signal, and the rule cannot inspect or control Docker.

Storage pressure critical → firing healthy → recovery; initializing and degraded are neutral

Ingestion pipeline stopped → firing healthy → recovery; initializing and stopping are neutral

Notification delivery failed → firing healthy → recovery; disabled, initializing, and degraded are neutral

Runtime signal Only canonical process state is stored in evidence and incident history.

Enabled Evaluate current runtime health immediately after saving.