Following on from the SolarWinds World Tour dashboard, I wanted to take the same idea in a more operational direction.
The first dashboard was intentionally a bit of fun. It was an experiment in how far the SolarWinds presentation layer could be pushed beyond the usual tables, gauges, and standard widgets.
This one started with a different question:
What if a dashboard did not just show the state of the monitored estate, but also showed how much confidence we should have in the monitoring evidence itself?
That became the Observability Lens.
From monitoring data to monitoring trust
Most dashboards surface individual signals:
- nodes down
- interfaces under pressure
- applications in warning or critical state
- active alerts
- polling completion
- unknown or unmanaged entities
All of that information is useful, but it still leaves the viewer to interpret the wider operational story.
The Observability Lens takes that one step further.
It collects live evidence from SolarWinds, interprets that evidence into a monitoring-trust posture, and presents the result as a single visual state:
Trusted · Degraded · Noisy · Blind Spot · At Risk
The underlying model is simple:
Evidence → State → Action
The goal is not to replace detailed dashboards or engineer drill-down views.
It is to provide a fast, recognisable summary of whether the monitoring evidence is clean, complete, current, and actionable.
The centre lens
At the centre of the dashboard is the current monitoring-trust posture.
The main ring is split into two parts:
- the completed arc represents achieved trust
- the remaining segment represents pressure against that trust
A score of 75/100 therefore reads as:
75% achieved trust25% remaining pressure
The state label provides the operational interpretation.
A blended score can still be overridden by a stronger signal where needed. For example, the estate may still be classified as At Risk if polling completion drops significantly, coverage risk becomes too high, or aged-alert pressure crosses a defined guardrail.
The contributing signals
Around the lens are six supporting signals:
- Evidence Quality — how much of the managed estate is currently healthy
- Polling Completion — whether the polling layer is keeping up
- Alert Confidence — how clean and actionable the alert behaviour looks
- Blind Spot Risk — how much of the estate lacks dependable evidence
- Active Alerts — the current alert burden
- Noise Pressure — how much recent alert activity is driven by recurrence
The dashboard is driven by live Orion evidence across:
- nodes
- interfaces
- SAM applications
- active alerts
- alert-trigger events
- polling engines
Evidence distribution
The lower evidence deck separates the monitored estate into:
- Healthy
- Warning
- Critical
- Coverage
That distinction matters.
A critical entity and an unknown entity are not the same thing.
One is evidence of failure.
The other is a blind spot where the evidence cannot currently be trusted.
Coverage therefore includes unknown, unmanaged, or unclassified objects so the dashboard can distinguish operational degradation from missing or incomplete evidence.
Why This State
The dashboard also includes a compact explanation panel:
Why This State
This surfaces:
- the current posture
- the trust score
- the primary driver
- the next action
For example:
AT RISK · AGED ALERT PRESSURE 96% · INVESTIGATE THE PRIMARY STATE DRIVER
That turns the Lens from a visual showpiece into an operationally useful summary.
Alert pressure and monitoring hygiene
The seven-day trend beneath the Lens shows completed-day alert-trigger pressure.
The purpose is not simply to show alert volume.
It is to expose behaviour that can reduce confidence in the monitoring layer:
- repeated alert signatures
- aged active alerts
- alert backlog pressure
- recurrence-driven noise
These issues can sit beneath the surface of operational dashboards until monitoring becomes noisy, stale, or difficult to trust.
Built inside SolarWinds
The entire visual is rendered natively inside SolarWinds and driven by live Orion data.
The evidence layer determines the posture, and the visual layer makes that state easier to recognise, explain, and act on.
The wider idea
The aim is not simply to make monitoring prettier.
It is to make operational confidence visible.
Most dashboards answer:
What is currently down or under pressure?
This one asks:
How much confidence should we have in the monitoring evidence itself?
Still a bit of fun.
But this one is much closer to the direction I think monitoring dashboards should be heading. ✌️
Update: Part 3 is now available
Following on from the Observability Lens, I have taken one part of the evidence layer and explored it in more detail:
Alert behaviour.
Part 3 looks at how trigger volume, repeated signatures, aged backlog, domain concentration, and signal trust can be interpreted together as an operational posture:
Part 3. Alert Pressure Skyline: Turning Alert Behaviour into Operational Posture - THWACK
