Monitoring dashboards

What it is

Observatory gives you two dashboard levels. The Overview tab shows the health of your whole fleet on widgets you can rearrange, resize and hide. Each device you click in the sidebar opens its own tab with live charts. All charts are interactive (Apache ECharts) and refresh automatically.

The Overview dashboard

The Overview tab (default on open, header Monitoring Overview) is a grid of widgets:

• Metrics — Targets Up, Targets Down, Avg Latency, Packet Loss.
• Network Health Score — a 0-100 ring combining uptime (50%), latency (25%) and packet loss (25%).
• Alerts Dashboard — active alert events with severity badges. Click one to jump to the device.
• Quick Notes — free-text notes, auto-saved to the server as you type. If the notes fail to load when the Overview opens, the box shows “Failed to load” and stays read-only instead of accepting typing — this avoids a half-loaded box silently overwriting your team’s saved notes. Reload the page to try again.
• Top 5 Rankings — worst latency and worst packet loss.
• AI Insights — anomalies analyzed by CNS; see [[crearack—monitoring—cns-sentinel]].
• Agent Fleet — your Local Agents; see [[crearack—monitoring—fleet-manager]].
• 24h Availability Matrix — per-target availability heatmap for the last 24 hours.

Header buttons:

Overview data refreshes every 60 seconds; the heatmap every 5 minutes.

Customizing the widgets

The dashboard is locked by default, so nothing moves by accident. Click Customize to enter edit mode:

• Move a widget by dragging its title; resize it from the bottom-right corner or from its right and bottom edges (the handles are visible while you edit).
• Widgets opens a list with a checkbox per widget: untick one to hide it, tick it again to bring it back. A hidden widget stops requesting data until you show it again.
• Set as organization default (administrators only) saves the arrangement you are editing as the default for everyone in the organization, after a confirmation. It does not touch the arrangements colleagues already saved for themselves: it is what new users see, and what Restore default brings back.
• Restore default brings back the organization default if an administrator saved one, or the factory layout if not, and shows the hidden ones.
• Done saves your arrangement; Cancel discards every change made since you clicked Customize.

Your arrangement is personal: it is stored on the server under your account and follows you to any browser. Other users of the organization keep their own; until someone saves theirs, they see the organization default set by an administrator, or the factory layout if there is none. On narrow screens (a phone, a tablet held upright, a half-width window) the widgets stack in a single column, and you can customize that order too: press Customize, drag widgets up or down by their title and hide the ones you do not need, then Done. That mobile arrangement is saved separately from the desktop one, so each keeps its own layout. Until you save one, the phone shows the widgets in the same order as your desktop layout. Administrators can also set an organization default for the mobile arrangement with Set as organization default while on a narrow screen.

Only recent measurements count

The Metrics cards, the Network Health Score and the Top 5 Rankings only use devices with a recent measurement (within the last few polling cycles). A device whose Agent stopped reporting keeps its last latency and packet loss on file, but those frozen numbers are not averaged or ranked as if they were current — a device that is down and was measured two minutes ago still counts; one that has not been measured since yesterday does not.

If no device has a recent measurement — typically when the machine running the Local Agent is switched off — the score and the average cards show -- and the rankings say there is no data yet, instead of a score built from yesterday’s numbers. Bring the Agent back online and the figures return on the next refresh.

Device detail tabs

Click a device in the sidebar to open its tab. The header shows inline metrics (Status, Latency, Loss %, BW In, BW Out, Errors) and the chart controls. Below, collapsible sections — collapse state is remembered:

• Heartbeat (Ping) — latency and packet loss. Toolbar: chart style, AL (alert lines), Console (live ping panel), Alerts.
• Bandwidth (SNMP) — inbound/outbound Mbps. Toolbar: Agg (aggregate line), Sync Y (same scale on both axes), chart style, interface selector (which interface the Bandwidth and Interface Errors charts show — the Agent keeps one series per selected interface, so switching is instant; the primary interface is marked (primary) and is set in Settings), AL, Settings, Alerts.
• Interface Errors — errors and discards per minute (SNMP).
• Resources (CPU / Memory) and Environment (Temperature) — appear only on devices where Deep Discovery found the vendor OIDs.

A device tab refreshes at the device’s Polling cadence (15 s to 5 min, set in its card editor — see [[crearack—network—device-page]]), so the charts move at the same pace the device is polled.

Chart controls

Controls in the tab header apply to all charts of that device:

• Range — six buttons: 01m (1→5→10→20 min), 1h (1→2→4→8 h), 6h (6→12→24→48 h), 24h (24→48→72 h), 7d and 30d; click a cycling button again to extend it through its steps. 7d and 30d are single-value buttons, matching the same ranges already available in UPS, Wireless and Digital Signage — the server pre-summarizes them to one point every 15 minutes (7d) or 30 minutes (30d), so unlike the shorter ranges they do not keep refreshing live; re-click the button to pull the latest data. On the two long ranges, the chart’s time axis marks midnight with the date, not just the hour.
• Show Values — labels the most notable data points for a few seconds.
• Ani — toggles chart animation.
• Reset Zoom — back to 100% on all charts.
• Export — saves the charts as PNG.
• Chart Guide — opens the chart interpretation guide.
• Report — historical mode: pick From and To dates, click Apply, return with Live.
• New Tab — opens the device standalone in a new browser tab.

Each chart also has a style dropdown with 13 looks: Line (smooth, sharp, thick, dots), Area (solid, stacked), Gradient Stack, Step, Step (area), Bar, Bar (thin), Scatter, Scatter (large). Zoom with the slider under each chart or the mouse wheel; zoom state is remembered per chart.

The Chart Defaults panel in the sidebar sets your account-wide defaults (style, range — including 7d and 30d, animation, alert lines, smoothing) and Apply to All pushes them to every open tab.

Ping Console

In the Heartbeat (Ping) section, click Console to open a terminal-style panel with a live ping every 2 seconds. Green means reply, red means timeout. It keeps the last 50 lines and can be dragged around the chart. For a device on a private LAN address the cloud cannot ping it, so each line shows the last result reported by the Local Agent instead (Private IP — last known from Agent: up, 3.2ms, checked 18:40:11) — it is not a live ping and nothing is recorded in the metrics store.

What gets remembered

Widget layout (personal, saved with Done; administrators can also save one as the organization default), Quick Notes and Chart Defaults are stored on the server (they follow your account). Open tabs, time ranges, chart styles, zoom, collapsed sections and sidebar width are stored in your browser. Refresh pace is not a per-tab setting anymore — it follows each device’s polling cadence.

Related

• [[crearack—monitoring—que-es-observatory]] — module overview
• [[crearack—monitoring—metricas]] — the metrics behind the charts
• [[crearack—monitoring—configurar-snmp]] — SNMP setup for the Bandwidth chart
• [[crearack—monitoring—alertas]] — alert lines and the Alerts buttons
• [[crearack—monitoring—ping-http]] — ping and HTTP checks
• [[crearack—monitoring—cns-sentinel]] — AI Insights widget
• [[crearack—monitoring—fleet-manager]] — Agent Fleet widget
• [[crearack—network—device-page]] — setting the polling cadence per device

Véase también

• [[crearack—monitoring—que-es-observatory]]
• [[crearack—monitoring—metricas]]
• [[crearack—monitoring—configurar-snmp]]
• [[crearack—monitoring—alertas]]
• [[crearack—monitoring—ping-http]]
• [[crearack—monitoring—cns-sentinel]]
• [[crearack—monitoring—fleet-manager]]