Platform health

Run health scans, triage the issues queue, and resolve or waive findings.

Super Admin → Platform Health → /health

Platform Health scans your data against a rule set and reports what it finds: leases billing nothing, units with no owner, payments unallocated for too long, configuration that will fail the next time a job runs. It catches the class of problem that never throws an error — it just produces quietly wrong numbers.

The pages

PageRoutePurpose
Overview/health/overviewHealth score, open issues, critical issues, last scan, incidents over time.
Issues/health/findingsThe issues queue.
Issue/health/finding/:idOne finding in detail.
Rules/health/rulesThe rule catalogue, and each rule's findings.
Scan history/health/scansEvery scan run and its outcome.

Running a scan

Run Health Scan starts a scan, with options including a full scan and severity selection (for example errors only, or specific rule codes). Progress is reported as the scan runs, and the result appears as a run snapshot with a scan timeline and outcomes.

  1. Run a full scan after any large change

    A migration, a bulk import, a configuration overhaul.

  2. Run a targeted scan when investigating

    Restrict to the rule codes relevant to what you are chasing.

  3. Compare against the last scan

    New findings since the last run are what you act on; a stable backlog is a different conversation.

The issues queue

Each finding carries:

ElementMeaning
Rule / rule codeWhich check produced it.
SeverityHow serious it is — critical, high, and below.
ProblemWhat is wrong.
ImpactWhat it affects if left.
Affected record / entityThe specific record, with entity type and ID.
How this was detectedThe evidence behind the finding.
Recommended actionWhat to do about it.
Suggested next stepThe immediate move.
Detected by / last seenWhich scan found it and when it was last confirmed.
StatusOpen, resolved, snoozed, waived.

Quick filters, sorting and optional columns work as elsewhere; findings can be exported.

Resolving a finding

  1. Read the impact, not just the problem

    Severity sorts the queue; impact decides whether it matters to you this week.

  2. Fix the underlying record

    Health findings point at real data. Fixing the record is the resolution; the finding closing is the consequence.

  3. Record the resolution

    Choose the resolution type and add notes. Some resolutions require a reason.

  4. Re-scan to confirm

    A finding that reappears was not actually fixed.

Snooze and waive

ActionUse it when
Snooze untilThe finding is real but cannot be fixed yet — a supplier is due to confirm something next week.
Waiver (reason required)The finding does not apply to your situation and never will.

Rules

The rules page lists every check with its code and enabled severity, and lets you drill into the findings a rule has produced. Rule execution failures are shown separately — a rule that could not run is not the same as a rule that found nothing, and treating them the same is how a blind spot forms.

Health score

The overview shows a health score and incidents over time. Use the trend, not the absolute number: a score that is falling week on week is a signal even when the absolute value looks acceptable.

Working health into a routine

  1. Weekly: clear new critical and high findings

    They are usually few and usually fast.

  2. Monthly: review snoozed findings

    Snoozes expire; findings that were snoozed twice should be fixed or waived.

  3. Before month end: scan and clear

    Health findings frequently explain reconciliation differences before you go looking for them.

  4. Quarterly: review waivers

    A waiver that has outlived its reason is a hidden problem.