An uptime and incident dashboard should answer three questions in under ten seconds: is anything broken right now, are we meeting our availability targets, and is anything about to break. That means a top layer of status, uptime against SLA, latency percentiles and active incidents, backed by drill-down panels that explain them.
Start With Two Layers: At a Glance vs Drill-Down
Most dashboards fail by giving every panel equal weight. Separate the view an on-call engineer opens during an alert from the one a team lead reviews weekly.
- At a glance: current status, uptime percentage vs target, response time trend, and open incidents. These should fit above the fold.
- Drill-down: error and status code breakdown, per-region availability, MTTD/MTTR history, certificate and domain expiry, and alert volume per monitor. These explain why the top layer looks the way it does.
The At-a-Glance Panels
Current status overview
One row per service showing up, degraded, or down, with failing services sorted to the top. Include the time of the last state change: "down for 40 seconds" and "down for 40 minutes" need very different responses.
Uptime percentage vs SLA target
Show uptime per service over 24 hours, 7 days and 30 days against an SLA reference line. A 99.9% monthly target allows roughly 43 minutes of downtime, so showing remaining error budget in minutes is often more useful than the percentage. The 30-day view tells you whether you can afford a risky deploy.
Response time trend with p50 and p95
Averages hide pain. Plot the median (p50) for typical experience and p95 for your slowest users. A stable p50 with a rising p95 often signals a saturated dependency or one degraded region before a full outage.
Active and recent incidents
List open incidents with start time, affected monitors and running duration, then incidents resolved in the last 72 hours. Duration matters more than count.
The Drill-Down Panels
- MTTD and MTTR: Mean time to detect shows whether check intervals and thresholds are tight enough; mean time to resolve shows response effectiveness. Trend them over 30 to 90 days.
- Error rate and status code breakdown: Split responses into 2xx, 3xx, 4xx and 5xx. A 5xx spike points to the server; a surge in 404s after a release usually means broken routes.
- SSL and domain expiry countdown: Display days remaining for every certificate and domain, with warnings at 30, 14 and 7 days. Expiry outages are entirely preventable.
- Per-region availability: Checks from multiple locations separate a global outage from a regional DNS, CDN or routing issue.
- Alert volume: Count alerts per monitor per week. Monitors that fire often without matching incidents need retuning.
Recommended Panels and Time Windows
| Panel | Question it answers | Recommended time window |
|---|---|---|
| Current status | What is broken right now? | Live |
| Uptime vs SLA | Are we within our availability target? | 24h / 7d / 30d |
| Response time p50/p95 | Is performance degrading for real users? | 24h, compared to 7d |
| Active and recent incidents | What is open and how long has it lasted? | Live + 72h |
| MTTD / MTTR | Are we detecting and fixing faster? | 30–90d |
| Status code breakdown | Is the failure server-side or client-side? | 24h / 7d |
| SSL and domain expiry | What will expire before we notice? | Next 30–90d |
| Per-region availability | Is the problem global or regional? | 24h |
| Alert volume | Which monitors create noise? | 7d / 30d |
For deciding which of these metrics to prioritize first, see our guide on which website uptime metrics SaaS teams should track first.
How UpScanX Covers These Panels
UpScanX gives each monitor a real-time performance view with current status, uptime percentage, response time history and incident counts, across HTTP/HTTPS, API, SSL, domain, ping and port checks run from 15+ global locations. Alerts go to Email, Slack, Telegram, Microsoft Teams or Webhook, and AI reports add anomaly detection, alert correlation and scheduled email summaries.
On the traffic side, the free UpScanX analytics dashboard adds an HTTP status code breakdown (2xx/3xx/4xx/5xx and error pages) from real visits, so you can see whether an incident actually reached users.
Key Takeaways
- Separate at-a-glance panels (status, uptime vs SLA, p50/p95, open incidents) from drill-down panels.
- Show uptime over 24h, 7d and 30d against a visible SLA line and error budget.
- Use percentiles, not averages, for response time.
- Track expiry countdowns and per-region availability to catch preventable and partial outages.
- Review MTTD, MTTR and alert volume monthly.