What goes in an IT health dashboard in Metabase?

An IT health dashboard gives IT leadership one pane for the estate's overall condition: availability, open incidents, patch compliance, backup success, endpoint health, and warranty/EOL exposure. It is the rolled-up scorecard — the minute-by-minute view lives on an IT monitoring dashboard — and Metabase builds it from daily rollups of your monitoring, patching, backup, and asset systems in the warehouse.

What does an IT health dashboard look like?

Here’s the layout this guide builds. The health score and estate-level KPIs sit at the top so a director gets the verdict first; incidents and recovery come next because they explain most bad weeks; patching, endpoints, and lifecycle exposure sit at the bottom — the slow-moving debt that never pages anyone but decides how bad the next incident is.

Which cards belong on an IT health dashboard?

The eight below cover the question a director actually asks — “are we healthy, and if not, which discipline is slipping?” — without dropping into per-host telemetry.

• Overall estate health score, a weighted composite of the pillars below (gauge)
• System availability by month, with month-over-month comparison (trend)
• Incidents opened by severity, weekly (stacked bar)
• Mean time to resolve by month (line)
• Backup success rate, nightly, against a 99% goal (line)
• Patch compliance by device group (row)
• Endpoint health status mix — healthy, at risk, critical, offline (donut)
• Assets nearing warranty or EOL, next 90 days (table)

What data does the dashboard need?

• uptime_checks — device ID, timestamp, and status from your monitoring stack, for availability rollups.
• incidents — ID, severity, opened and resolved timestamps, and category, from ticketing or PagerDuty.
• patch_status — one row per device per day with missing-patch counts and an explicit is_compliant boolean.
• backup_jobs — job ID, device, start time, and success/failure status from your backup tool.
• assets — CMDB export with device type, business unit, purchase date, warranty_end, and eol_date.

How do you build it?

1. Land nightly exports from monitoring, patch management, backup, and the CMDB into the warehouse, keyed on one shared device ID — the joins are the hard part, so fix ID mismatches here, not in cards.
2. Define the health score once as a Metabase model (for example 40% availability, 35% patching, 25% backups) so the gauge, the KPI row, and any per-unit breakdown share the same formula.
3. Build the availability trend from uptime_checks grouped by month, and the incident cards from your ticketing sync — see the PagerDuty integration guide for getting incident data into the warehouse.
4. Add the lifecycle table filtered to warranty_end or eol_date within 90 days, sorted soonest-first, with already-expired assets flagged.
5. Add filters for business unit, device group, and date range, then set up a weekly dashboard subscription for the leadership review.

Example card SQL

WITH availability AS (
SELECT
  d.business_unit,
  ROUND(100.0 * COUNT(*) FILTER (WHERE c.status = 'up')
    / NULLIF(COUNT(*), 0), 2)                    AS availability_pct
FROM uptime_checks c
JOIN devices d ON d.device_id = c.device_id
WHERE c.checked_at >= CURRENT_DATE - 30
GROUP BY d.business_unit
),
patching AS (
SELECT
  d.business_unit,
  ROUND(100.0 * COUNT(*) FILTER (WHERE p.is_compliant)
    / NULLIF(COUNT(*), 0), 1)                    AS patch_pct
FROM patch_status p
JOIN devices d ON d.device_id = p.device_id
WHERE p.snapshot_date = CURRENT_DATE
GROUP BY d.business_unit
),
backups AS (
SELECT
  d.business_unit,
  ROUND(100.0 * COUNT(*) FILTER (WHERE b.status = 'success')
    / NULLIF(COUNT(*), 0), 1)                    AS backup_pct
FROM backup_jobs b
JOIN devices d ON d.device_id = b.device_id
WHERE b.started_at >= CURRENT_DATE - 7
GROUP BY d.business_unit
)
SELECT
a.business_unit,
a.availability_pct,
p.patch_pct,
b.backup_pct,
ROUND(0.40 * a.availability_pct
    + 0.35 * p.patch_pct
    + 0.25 * b.backup_pct, 1)                    AS health_score
FROM availability a
JOIN patching p USING (business_unit)
JOIN backups  b USING (business_unit)
ORDER BY health_score;

Related

Metrics

Integrations

Dashboards

FAQ