Skip to main content

Read the Overview dashboard

Documentation for Tutela v3.9.0.

Use Overview to decide what needs attention, then open the underlying activity or finding before taking action. Its cards measure different things and should not be added together.

Prerequisites​

  • Permission to view Overview. Access to Findings, Cases and approvals depends on your role.

Start with the time and source​

  1. Open Overview and check the selected activity period.
  2. Check the latest record and assessment times. If a source is unavailable or partial, treat the affected card as incomplete information.
  3. Review Highest-priority open findings and open a finding to inspect its concern, evidence limitations and supporting session. Use Cases when it needs an owner, notes or follow-up.
  4. Review Pending approvals separately. These are decisions awaiting review, not resolved findings or completed actions.

What the cards count​

CardWhat it measures
Governed requestsDistinct recorded request IDs within the selected period. Input and output records for one request count once. Records without an ID are reported separately.
Open FindingsThe current unresolved policy-owned Findings backlog across all time. Historical concerns remain in the inventory without joining the current queue. Changing the activity period does not hide older open Findings.
Known-system evidenceCurrent registrations and retained period observations, excluding catalog-only entries. Counts distinguish systems with recorded governance evidence from those without sufficient evidence. Unavailable sources remain explicit; this is not organization-wide protection.
Pending approvalsThe current decision queue, separate from findings and execution results.
Recorded control actionsPolicy decisions in the selected period. Open the associated activity to understand the result.

Governed-request history uses distinct request IDs. Separately labeled inspection-record activity and active-system counts use their own recorded populations. A request can create more than one inspection record, so these counts may differ legitimately.

Follow an item to evidence​

Direct policy outcomes and observed execution are separate facts. For a policy action, open the related Activity session and read the decision, execution status and response handling separately. An Allow decision does not prove execution. A response withheld after a tool completed does not mean that the tool action was prevented.

For a Finding, check when its source last updated and review any evidence gaps before assigning severity or resolution. Link or create a Case when multiple people need to coordinate the investigation.

Use AI Spend for detailed costs, Evidence for evidence reports, and Settings > Support & diagnostics for service health.

Human review queue​

In v3.9, Human review · All time shows Needs review and In investigation totals for the complete authorized queue, not just the rows loaded on the current page. Open either total to review that queue. These human-review states are separate from the open machine backlog and the selected activity period. A reviewed finding can still remain open.

Review totals unavailable means the queue could not be verified; it does not mean zero. Retry by refreshing Overview, then check Findings or ask your administrator if the source remains unavailable. Users without Finding access do not receive these totals.

Recorded history​

Compact summaries keep governed requests, the open backlog and pending approvals separate. The severity distribution retains the exact count and investigation link for each category. The control-action distribution uses the selected period; it does not change the current backlog. A recorded zero stays zero and an unavailable category stays unavailable.

In v3.9, expand Historical activity and availability first. Open View Governed request history data or View Open backlog history data for exact observations behind each chart. Blank intervals are gaps, not an all-clear. Finding rows identify the recorded subject and action where available, with the supporting session available from the finding. Incomplete assessment inputs do not establish which evidence is missing for an individual action.

Review a finding​

Audience: analysts and investigators. In v3.9, select the finding title to open its detail page. Follow the supporting session from there when you need the action history. Earlier releases may also show separate finding and session links on Overview.

Last occurrence is the latest occurrence recorded for that finding. Last assessment is when the assessment was recorded, not when the action ran. Both display the timezone. Time unavailable means no usable timestamp was supplied; it is not replaced by the assessment or page-refresh time.

Ask an investigator to review the finding if your role cannot open Findings. If the destination is unavailable, refresh once and contact your administrator with the finding identifier and displayed time. Do not treat an unavailable page as evidence that the concern was resolved.

Expected result: you can identify the affected item, inspect its recorded outcome and uncertainty, and decide whether it needs a Case.

The dashboard API exposes request history separately from inspection-record trends. Each request counts once, in its earliest recorded bucket inside the selected period. Changing that period can move the request to a later bucket because earlier records are outside the selection. Records with no request ID remain a separate count.

Backlog history is a series of recorded observations, not a total of new findings. It starts when background collection is enabled and retains 90 days. A daily point is the last observation for that day. Missing hours, outages, and dates before collection began remain gaps. Each observation retains its capture time, assessment time and source limitations. Investigation permissions also apply to this history. An unavailable observation does not establish that the backlog was empty.

Known-system evidence​

The dashboard API also returns a separately scoped known-system summary. It combines current system registrations with retained observations during the selected period. Recorded governance evidence requires a matching first-party record for that exact system. A registration, an active credential, or one recorded request does not mean the entire system is protected. This summary does not replace organization-wide control coverage.

When a source cannot be read, combined totals remain unavailable. Available source counts and their recorded times remain separate; source counts may overlap and must not be added. Records without a usable system association are not counted as additional systems. Administrators can review the affected source's connection and identity reporting; other users should send the source and time to their administrator.

Validate​

  • Your interpretation uses the correct time period and count definition.
  • You opened supporting evidence for items requiring action.
  • Missing data and unconfirmed execution remain explicit.

Next steps​

The four aligned summaries are Governed requests, Open Findings, Pending approvals and Known-system evidence. Human review totals stay within Findings. Activity charts show inspection records; gaps mean no recorded observation. System bars compare counts with the most active listed system, not an organization-wide denominator.

Expand Organization-wide control coverage or Metric definitions for methodology and supporting scope. Known-system sources retains source details. Configure DSPM enrichment in Settings > Data enrichment; consequential source failures stay visible in the affected summaries and investigations.