Skip to main content
The Dashboard is the landing page for an organisation. It rolls up every cluster you manage into a single view: a geographic world map, a health summary of the issues that need attention, and a cost overview across your providers.
The Dashboard is organisation-scoped. It only shows clusters in the organisation you currently have selected - switch organisations from the top-left selector (or with the CLI --org flag) to see a different fleet.

Fleet world map

The world map plots each cluster at its real geographic location, so you can see where your workloads run across regions and providers at a glance. Ankra locates a cluster from one of two sources, in priority order:

Cloud region

For clusters on a known provider, the region is read from the node labels (topology.kubernetes.io/region) and mapped to coordinates. This is the most precise source and is used whenever available.

GeoIP

When no region label is present, Ankra falls back to the cluster’s public egress IP - reported by the agent - and resolves it to an approximate city/country. Locations from this source are marked as approximate.

Clusters that don’t appear on the map

A cluster is listed as unlocated (shown in a separate list rather than on the map) when neither source resolves:
ReasonMeaningHow to fix
no_region_labelThe cluster’s nodes carry no region label and no GeoIP location is availableEnable public IP reporting on the agent (see below), or wait for node sync to populate region labels
unknown_regionThe node has a region label, but Ankra has no coordinates mapped for itThe cluster still works normally; only the map pin is missing
To place imported clusters that run in private networks (no recognisable cloud region) on the map, enable public IP reporting on the agent by setting PUBLIC_IP_REPORTING_ENABLED=true. See Cluster Agent for the full configuration.
Draft clusters are never shown on the map - only fully imported, non-deleted clusters appear.

Health summary

The Dashboard surfaces the most pressing issues across the fleet so you can triage from one place: offline clusters, failed reconciles, and unhealthy add-ons bubble to the top. Open a cluster from the summary to jump straight into its detail view, or ask the AI Assistant to investigate. For a deeper look at how Ankra classifies cluster status, see Cluster States.

Cost overview

The Dashboard includes a fleet-wide cost rollup - estimated monthly spend, month-to-date, and a projected month-end total, broken down by provider and by top-spending cluster. This is the same data covered in detail on the Cloud Cost page.

API

The Dashboard is backed by read-only organisation endpoints:
MethodPathPurpose
GET/org/dashboard/fleet-locationsLocated + unlocated clusters for the world map
GET/org/dashboard/healthHealth summary (use ?max_issues= to cap the list, 1-50)
GET/org/cloud-cost/summaryFleet cost rollup
See the API Reference for request and response schemas.