Observability
From live Prometheus-backed charts to audit trails and secure operations — manage fleet health, capacity, and compliance without leaving Cloud Admin.
Observability
You should not need Grafana, three browser tabs, and a spreadsheet to answer how the fleet is doing. In Cloud Admin, open Monitoring & Ops → Observability for Prometheus-backed charts, inventory, and drill-downs from a quick summary to node- and pod-level detail.
Screens from the live product, with a short note on when you would open each view.
Real-time Performance Analytics area chart for CPU, memory, GPU, and workloads over 24 hours.
Click the screenshot to open full size, zoom, and pan.
Historical trends reveal whether usage is seasonal, spiking, or creeping, real-time performance analytics line chart with metric summary cards. Pair with capacity planning before users feel pain.
Click the screenshot to open full size, zoom, and pan.
In Memory Pool, the Overview view answers one operational question at a time, memory pool analytics overview with total, used, free memory and utilization. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In Memory Pool, the Usage view answers one operational question at a time, memory pool analytics usage tab with working set and cached memory. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In Memory Pool, the Breakdown view answers one operational question at a time, memory breakdown by namespace and top memory-consuming pods. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In Memory Pool, the Trends view answers one operational question at a time, 24-hour memory usage trend with min, max, and average. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In Memory Pool, the Issues view answers one operational question at a time, memory health issues tab with pressure and oom checks. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In CPU Performance, the Overview view answers one operational question at a time, cpu performance analytics overview with usage, cores, and limits. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In CPU Performance, the Usage view answers one operational question at a time, cpu usage tab with used, idle cpu and overcommit ratio. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In CPU Performance, the Breakdown view answers one operational question at a time, cpu breakdown by namespace and top cpu pods. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In CPU Performance, the Trends view answers one operational question at a time, 24-hour cpu usage trend chart and statistics. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In CPU Performance, the Issues view answers one operational question at a time, cpu health checks for usage, overcommit, and throttling. Part of Monitoring workspace; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
Fleet KPIs, saturation charts, flow telemetry, and Prometheus-backed alerts share one investigation cadence: confirm scope, isolate the layer that moved, validate mitigation.
Slice traffic by cluster total, namespace, node NIC, or workload; correlate spikes with pod-level tables instead of tcpdump theater.
Allocation donuts, 24h trends, namespace and node breakdowns, plus ranked pods, ideal when HPA metrics disagree with user-visible latency.
Alert summaries pair severity counts with the charts that explain them. Fewer circular conversations about “whether Prometheus fired.”
Nodes, namespaces, controllers, ConfigMaps, Secrets, NetworkPolicies, and Services appear beside charts, no separate CMDB hunt mid-incident.
Monitoring & Ops
In Cloud Admin, Monitoring & Ops → Capacity Planning turns Prometheus signals and Kubernetes inventory into snapshots, forecasts, and fleet-wide roll-ups, so finance and platform teams agree before spend or outages force the conversation.
Screens from the live product, with a short note on when you would open each view.
Node fleet capacity for CPU, memory, storage, and pods.
Click the screenshot to open full size, zoom, and pan.
Capacity planning with headroom and predictive risk overlay.
Click the screenshot to open full size, zoom, and pan.
When capacity tiles, forecasts, and tenant overlays reference identical Prometheus series and Kubernetes objects, debates shift from “whose export is right” to “which scenario do we fund.”
Forecasts & actions stay grounded because collectors, schemas, and Prometheus health are visible, not hidden jobs in another portal.
Fleet and scenario tabs compare clusters without stripping virtualization, network, or LLM-specific overlays.
Dedicated tenant lenses translate raw cluster metrics into chargeback-ready narratives.
Monitoring & Ops
Open Monitoring & Ops → Events to watch Kubernetes events as they happen: cluster-scoped context, type/reason distributions, rich filters, and a searchable table, ideal when logs are noisy but the API surface tells the truth.
Screens from the live product, with a short note on when you would open each view.
Cluster events charts and searchable event table.
Click the screenshot to open full size, zoom, and pan.
Events become actionable when everyone sees the same Normal/Warning profile, reasons cluster visibly, and filters isolate the blast radius, without copying kubectl output into chat.
Donut charts give executives the headline mix while engineers stay in the detailed table.
Warning badges and reasons isolate kubelet and scheduler narratives before you grep terabytes of stdout.
Switch clusters without breaking context, Monitoring & Ops routes stay consistent.
Monitoring & Ops
Under Monitoring & Ops → Resources (and the FN K8s Dashboard entry point), Cloud Admin shows one cluster up front: version and node count in the header, a KPI strip for CPU, memory, pods, namespaces, and readiness, then capacity vs usage charts and per-node cards with pod slots, CPU/memory/storage bars, and paths to details or console.
Screens from the live product, with a short note on when you would open each view.
Your starting point for Kubernetes cluster operations: KPI cards and charts summarize fleet health, including cluster overview with nodes, pods, gpus, and utilization gauges. Spot drift early, then drill into the tab that explains the root cause.
Click the screenshot to open full size, zoom, and pan.
Cluster nodes view with CPU and memory usage per node.
Click the screenshot to open full size, zoom, and pan.
In Namespace Analytics, the Overview view answers one operational question at a time, namespace environment analytics overview and top consumers. Part of Kubernetes cluster operations; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In Namespace, the Resources view answers one operational question at a time, cpu and memory usage by namespace. Part of Kubernetes cluster operations; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In Namespace, the Quotas view answers one operational question at a time, namespace quota limits and pod limit issues. Part of Kubernetes cluster operations; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
In Namespace, the Issues view answers one operational question at a time, namespaces without limits and quota violations. Part of Kubernetes cluster operations; use it when you need evidence before changing limits, scaling, or opening a ticket.
Click the screenshot to open full size, zoom, and pan.
Your starting point for Kubernetes cluster operations: KPI cards and charts summarize fleet health, including kubernetes cluster dashboard with cpu, memory, pod, and node gauges. Spot drift early, then drill into the tab that explains the root cause.
Click the screenshot to open full size, zoom, and pan.
Cluster Resources bridges executive KPIs and SRE forensics: the same refresh cycle feeds the top strip, capacity vs usage bars, and node cards, so debates reference one dataset.
Capacity charts call out when allocated share crosses policy bands, less guesswork than raw gauge colors alone.
Node cards expose console access for operators who need immediate shell validation beside the charts.
Version and build metadata help compare clusters during upgrades without opening vendor PDFs.
Monitoring & Ops
Open Monitoring & Ops → Resources Map for a spatial view of the cluster: namespaces as swimlanes, workload and storage cards inside each lane, and controls to group by namespace, instance, node, or none, plus namespace and resource-type filters, an Issues only toggle, expand/collapse, and zoom to fit.
Screens from the live product, with a short note on when you would open each view.
Cluster resource topology graph by namespace.
Click the screenshot to open full size, zoom, and pan.
Resources Map is for the moment when everyone asks “what shares that namespace?”, grouping, filters, and issue-only mode answer it without exporting a dozen kubectl transcripts.
When list APIs return slices or validation warnings, the UI can surface them, so “empty” never means “healthy” by accident.
Flip to degraded cards only during incidents; expand all when you need the full context for a change window.
The map respects the cluster picker used across Monitoring & Ops, no duplicate “which kubeconfig am I holding?” moments.
Multi-cluster
When Kubernetes estates span teams and regions, “green” is not enough. FusioNative ranks environments, surfaces connectivity or metric gaps, and keeps remediation paths obvious.
Screens from the live product, with a short note on when you would open each view.
Infrastructure environment cards with health scores and utilization.
Click the screenshot to open full size, zoom, and pan.
All Clusters list with health, version, and resource bars.
Click the screenshot to open full size, zoom, and pan.
Cluster management table with connection and status details.
Click the screenshot to open full size, zoom, and pan.
Security & Audit
Open Security & Audit → Security for the Security Center: live counts for roles, standards, Kubernetes NetworkPolicies, Fusionative policies, and secrets; quick actions to create roles, standards, policies, secrets, and ConfigMaps; tabbed drill-downs; analytics tiles for namespaces, bindings, service accounts, and data objects, plus audit logging context and Pod Security admission visibility.
Screens from the live product, with a short note on when you would open each view.
Security center overview and analytics.
Click the screenshot to open full size, zoom, and pan.
Kubernetes RBAC roles list.
Click the screenshot to open full size, zoom, and pan.
Pod security standards enforcement.
Click the screenshot to open full size, zoom, and pan.
Kubernetes network policies tab.
Click the screenshot to open full size, zoom, and pan.
Secrets management in security center.
Click the screenshot to open full size, zoom, and pan.
ConfigMaps in security center.
Click the screenshot to open full size, zoom, and pan.
Security Center ties counts to actions, inventory informs standards; standards inform policies; policies and RBAC guard secrets and ConfigMaps, still inside the same Cloud Admin session finance and ops already trust.
Ribbon metrics answer “how exposed are we?” without exporting kubectl dumps.
Quick creates encode validation. Fewer broken YAML cycles during onboarding sprints.
Unified ingest counters preview how much evidence flows into Audit Compliance for frameworks like ISO/IEC 27001 and GDPR themes.
Security & Audit
Under Security & Audit → Audit compliance, Cloud Admin aggregates API and control-plane signals into one queryable trail, aligned to informational themes for ISO/IEC 27001, PDPL (KSA), ISO/IEC 42001, and GDPR. Monitor ingest health, slice by tenant, cluster, source, trace, HTTP status, and framework filters, then export CSV (with optional tags) for auditors and SIEM pipelines.
Screens from the live product, with a short note on when you would open each view.
Actionable signals instead of raw logs, unified audit trail with framework filters. Each item ties back to audit and compliance so owners know what to fix now versus what can wait.
Click the screenshot to open full size, zoom, and pan.
Actionable signals instead of raw logs, tenant-scoped audit events and export. Each item ties back to audit and compliance so owners know what to fix now versus what can wait.
Click the screenshot to open full size, zoom, and pan.
Audit Compliance closes the loop between “we log everything somewhere” and “here is the CSV with tenant, cluster, trace, and framework filters your auditor requested.”
CSV exports carry structured tags when enabled. Fewer manual joins before Splunk or BigQuery handoff.
Sampling percentage is visible, legal and engineering agree what fraction of reality each report represents.
Inline alerts flag risky default credentials on archive paths before exports leak into the wrong bucket lifecycle.