Metrics & Monitoring
EmberCORE keeps the numbers that matter for what you're doing right now in front of you at all times, through the Metrics block, a row of live metric cards that sits above Tools and changes with the category you're working in.
This is operational, at-a-glance monitoring: "are my nodes ready, is anything alerting, how full is storage." For deep historical analysis and custom charting, EmberNet ships Grafana in the App Store. See Custom dashboards below.
The Metrics block
The strip shows a set of live cards, refreshed continuously. Which cards appear follows the category you've selected in the carousel above Tools:
| Category | What the strip shows |
|---|---|
| All | Whatever you have chosen. This is the only category with no fixed set |
| Zero Trust Network | Network up/down, mesh latency average and worst, reconciled services, identities and routers online, network alerts |
| Industrial Automation | PLC drivers up, broker tag count and throughput, OT devices, Ignition gateway status, industrial alerts |
| App Deployment | Fleet bundles, clusters and GitRepos ready, reconciled services, App Store catalog size, running services |
| Edge Infrastructure | Nodes ready, running pods, scheduling pressure, storage used, Cinder volumes, healthy deployments |
| Device Management | Total and offline devices, device alerts, inventory devices and online count, endpoints online |
| Monitoring & Diagnostics | Alert timeline over 24h, critical and warning alerts, alerts in the last 24h, probes, mesh latency |
| Sites | Sites online, alerting and total, site devices, buildings, and bindings |
| Compliance & Reporting | Audit events and permission denials over 24h, tenant audit events, total audit entries, scheduled reports |
| Platform | Platform uptime, AI service status, broker connections and status, active tenants, namespaces |
Selecting a category swaps the whole strip to that category's metrics. Everything is scoped to your current tenant. For platform staff, that's whatever the Client Switcher is set to.
Customizing the strip
Each category's strip is yours to tailor.
Choosing metrics
Open the gear on the Metrics block to choose which metrics appear for the current category. Pick at least 3, up to the whole catalog of 60 metrics; the layout expands to fit, and the column count follows how many cards actually fit rather than how many you picked, so cards stay legible. Your selection is saved per category and remembered between visits, so each category can show exactly what you care about. Use Reset to defaults to return a category to its standard set.
The list scrolls while the header and the Save button stay put, so your selection is always one click away no matter how far down the catalog you are.
Choosing how each metric is drawn
Every metric card can be drawn several ways, so pick the one that reads best for that value from the card's visualization picker:
| Visualization | Best for |
|---|---|
| Stat | A single current value with a trend arrow |
| Sparkline | A value plus its recent shape over time |
| Gauge | A bounded value against warning and critical thresholds, such as storage percent |
| Donut | A part-of-whole reading, with a legend naming each slice |
| Status pill | An up/down or online/offline state |
| Heat strip | Severity or intensity at a glance |
| Bar | A small set of counts side by side |
| Big trend | One headline reading over a full-width plot, for the value you watch most |
| Histogram | How something was distributed across the last 24 hours |
| Ranked | A leaderboard, such as the noisiest services |
| Timeline | When events happened, rather than how many |
Not every metric offers every one. A card's picker only shows the modes that make sense for what it measures, so a plain count is not offered a gauge.
Your choice per card is remembered, and colors follow the value, so a metric with thresholds turns amber as it approaches warning and red at critical.
Metrics you can't see
Some metrics are platform-wide and only meaningful to platform staff. If your role can't see a given metric, its card shows a dash rather than another tenant's number. The strip never leaks data across tenants.
What's measured
The catalog spans the whole platform: edge infrastructure (nodes, pods, scheduling, deployments), storage (volumes, capacity, health), the zero-trust mesh (peers, services, identities, routers), industrial systems (PLC drivers, broker tags and throughput, Ignition), devices and endpoints, sites, alerts, and compliance (audit events, permission denials). Every value is read live from the platform and scoped to your tenant.
Custom dashboards and historical analysis
The Metrics block is for live operational awareness, not ad-hoc charting. When you need to build your own dashboards, whether that's historical trends, correlated signals, OEE, or custom queries, deploy Grafana from the App Store. Point it at the time-series data your apps write to:
- InfluxDB or TimescaleDB for high-resolution time-series
- Prometheus for infrastructure and app metrics
- Telegraf to collect and route metrics into them
This keeps the operations console fast and focused while giving you a full-featured analytics surface when you need one.
Next steps
- Alerts & Notifications: turn a metric crossing a line into a page
- EmberCORE Overview: the console and its tools
- App Store: Grafana: build custom dashboards
- API Reference: read the same metrics programmatically