These subsystems are retired and replaced by better alternatives (FleetNow / FleetOps SPAs via dashboard_api; in-process pooling; reporting.v_ingest_health). Remove them so the repo reflects the live stack only. Nothing running depends on the deleted artifacts. Deleted (dead artifacts): - n8n-workflows/ (retired webhook exports), grafana/ (provisioning for the removed service), dwh/ (migrations for the decommissioned external warehouse) - runbooks: DWH_PIPELINE.md, DWH_Execution_Manual.md, grafanaDeployment.md, grafanaOperationalManual.md Code/config: - run_migrations.py: drop sync_role_passwords() (its only entries were the now -dead grafana_ro + pgbouncer syncs; the guard already made it inert) - .env: remove the two unused GRAFANA_* vars - ingest_movement_rev.py / db_audit / deploy_dashboard_api_staging.sh: reword stale Grafana/grafana_ro comments Docs: scrub n8n/Grafana/DWH from CLAUDE.md, CONNECTIONS, DATA_FLOW, OPERATIONS_MANUAL, docker_commands, KPI_FRAMEWORK, PLATFORM_OVERVIEW, STAGING_FLEETOPS, and deprecation-banner the two large SQL libraries (dwh_gold was already dropped 2026-06-05). Kept deliberately: the grafana_ro DB role (now an unused read-only login), applied migration history, dated docs/reports/*, and docs/superpowers/* specs. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
13 lines
503 B
SQL
13 lines
503 B
SQL
-- Idempotent: safe to run on every audit start
|
|
CREATE TABLE IF NOT EXISTS tracksolid.health_checks (
|
|
id BIGSERIAL PRIMARY KEY,
|
|
checked_at TIMESTAMPTZ NOT NULL DEFAULT NOW(),
|
|
check_name TEXT NOT NULL,
|
|
status TEXT NOT NULL CHECK (status IN ('ok', 'warning', 'critical')),
|
|
detail JSONB,
|
|
row_count INT
|
|
);
|
|
|
|
-- Index for time-range queries
|
|
CREATE INDEX IF NOT EXISTS health_checks_checked_at_idx
|
|
ON tracksolid.health_checks (checked_at DESC);
|