Kubernetes monitoring, without Prometheus
Not every cluster needs a full Prometheus and Grafana stack just to answer "is this pod near its limit?". Kubius reads metrics-server and charts CPU and memory - with request, limit and capacity lines - live, on your Mac, with nothing to deploy.
Works with Google GKE, Amazon EKS, Azure AKS, DigitalOcean, self-managed & more · macOS Sonoma (14) or later
The monitoring most people actually need
CPU & memory charts
Time-series usage for pods and nodes with request, limit, allocatable and capacity reference lines - so a number has context, not just a value.
Node health & pressure
See node conditions, memory and disk pressure and capacity across every namespace, live - the cluster-wide health a dashboard usually gives you.
OOM diagnostics
Out-of-memory kills are surfaced and attributed to the victim pod and container - the thing raw graphs rarely make obvious.
Live, not scraped
Metrics update as the cluster reports them - no scrape interval to tune, no retention to manage.
Nothing to deploy
It uses metrics-server (already on most clusters). No Prometheus, no Grafana, no exporters, no PVCs for time-series storage.
Turns into action
From a chart, jump to right-sizing recommendations - so monitoring leads somewhere, not just a pretty graph.
Why skip Prometheus
Answers, without the stack
For teams who want usage and health, not a monitoring project.
- Live CPU/memory charts with request, limit & capacity lines
- Cluster-wide node health and pressure
- OOM kills attributed to the victim pod/container
- Reads metrics-server - nothing to deploy in-cluster
- Right-sizing recommendations from the same data
- 14-day free trial, no card required
Simple, one-time pricing
Pay once. No subscription.
One-time licence from €49 - no subscription. 14-day free trial, no card required.
Free 14-day trial · macOS Sonoma (14) or later · No agents, nothing to install in-cluster
Questions, answered
Can I really monitor Kubernetes without Prometheus?
For the everyday questions - is this pod near its limit, which workloads are heavy, are nodes under pressure, what got OOM-killed - yes. Kubius reads metrics-server and charts it live with request/limit/capacity context. It complements a full Prometheus stack rather than replacing long-term alerting and retention.
What do I need installed?
Just metrics-server, which most clusters (including GKE, EKS and AKS) already run. There's no Prometheus, Grafana or exporters to deploy.
Does it keep history?
Kubius charts a live rolling window per pod and node while you're connected. For long-term retention and alerting you'd still use Prometheus - this is for fast, on-demand insight.
Is this a Pro feature?
Cluster dashboards and cost/right-sizing are part of Kubius Pro; the 14-day trial includes everything so you can try it on your own clusters.