Find out why your pod is broken - in seconds
CrashLoopBackOff, OOMKilled, Pending, a flapping node? Kubius pulls the logs, events, metrics and node diagnostics into one window - and traces OOM-kills to the exact container that blew its limit. No agents to install, nothing to stand up.
Works with Google GKE, Amazon EKS, Azure AKS, DigitalOcean, self-managed & more · macOS Sonoma (14) or later
The 2am reality
From “something's wrong” to the root cause
- Five terminal tabs: get pods, describe, logs, top, events
- kubectl logs one pod at a time, no live filter
- OOMKilled - but which container, and why?
- Stand up Prometheus just to see a memory graph
- The node dies and takes the evidence with it
- Logs, events, metrics and node health in one window
- Stream every pod of a workload, filtered live by regex
- OOM-kills traced to the exact container that blew its limit
- Time-series CPU and memory with the limit line, built in
- A node-diagnostics snapshot that survives the node dying
Everything you need to find the cause
Live logs, every pod at once
Live-follow a pod with regex and level filters - or stream every pod of a workload in one interleaved, colour-coded view.
Events that explain it
The event timeline for any resource, right beside it - the reason a pod won't schedule or keeps restarting, in plain sight.
Node diagnostics & OOM
A rule-based node health report: OOM-kills traced to the container responsible, kubelet flapping and memory pressure, ranked worst-first - and it survives the node dying.
See what's connected
An interactive dependency graph - controllers, ReplicaSets, pods, Services, Ingresses - to follow a failure across the chain.
Metrics with the limit line
Time-series CPU and memory with request, limit and capacity reference lines - spot the throttle or the OOM headroom at a glance.
The moment it happens
Native notifications for crashes, OOM-kills, failed jobs and node pressure - so you're looking before the pager finishes buzzing.
Why it's faster
Kubius vs piecing it together with kubectl
| Kubius | kubectl + tabs | |
|---|---|---|
| Logs, events and metrics in one view | ||
| Stream every pod of a workload | one at a time | |
| Live regex & level log filters | ||
| OOM-kills traced to the container | guesswork | |
| Node health report (flapping, pressure) | ||
| Diagnostics survive node death | ||
| Metrics charts with request/limit lines | needs Prometheus | |
| Works with nothing pre-installed |
Less flailing
Everything you'd piece together from five kubectl commands
In one window, on your Mac - no agents, no dashboards to stand up first.
- Logs, events and metrics side by side
- OOM-kills traced to the exact container
- PodDisruptionBudget-aware node drains
- Dependency graph to follow the blast radius
- Last-applied vs live YAML diff
- Nothing to install in-cluster
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
Does it work mid-incident, with nothing pre-installed?
Yes. Connect your kubeconfig and you're looking at live logs, events and diagnostics immediately - no agents, dashboards or Prometheus to stand up first.
How does the OOM tracing work?
Kubius reads the node's cAdvisor metrics and container statuses to attribute OOM-kills to the specific container, and keeps a snapshot so the report survives even if the node dies.
Can I see logs across every pod of a deployment?
Yes - Combined Logs streams every pod behind a workload in one interleaved view, each line prefixed with its pod name in a unique colour.
Which clusters does it support?
Any context in your kubeconfig - Google GKE, Amazon EKS, Azure AKS, DigitalOcean, self-managed and more - on macOS Sonoma (14) or later.