Full visibility into
every cluster.
Node health, pod restarts, resource quotas, deployment rollouts — all correlated with your traces and logs. No extra agents.
Pods — production + data
Live · 2s refresh| Pod | Namespace | Status | CPU | Memory | Restarts |
|---|---|---|---|---|---|
| api-server-7d9f | production | Running | 120m | 256Mi | 0 |
| checkout-6b8c | production | Running | 88m | 180Mi | 0 |
| worker-5f7b | production | CrashLoop | — | — | 14 |
| postgres-0 | data | Running | 210m | 512Mi | 0 |
| redis-master-0 | data | Running | 45m | 128Mi | 0 |
| ingress-nginx-4k2p | kube-system | Running | 30m | 64Mi | 0 |
< 15s
Metric resolution
400+
K8s metrics tracked
0
Agents to install
100%
OpenTelemetry native
Real-time metrics
Cluster health at a glance.
CPU Utilisation (%)
Memory Usage (%)
Pod Restarts
Features
Everything Kubernetes, nothing you don't need.
Node & Pod Metrics
CPU, memory, network, and disk — per node, per pod, per namespace. Drill from cluster down to container in one click.
OOMKill & CrashLoop Alerts
Instant notification when a pod is killed or enters a crash loop. See the restart history and logs in the same view.
Namespace Dashboards
Pre-built dashboards per namespace with request rates, error budgets, and resource quotas — zero configuration needed.
Deployment Tracking
See exactly when a rollout happened and correlate it with spikes in latency or errors. Roll back with confidence.
Node Pool Visibility
Compare node pools side-by-side. Spot underutilized nodes and over-provisioned workloads to cut cloud spend.
Persistent Volume Health
Monitor PVC capacity, IOPS, and latency. Get alerted before a volume fills up and takes down a stateful workload.
OpenTelemetry native
No proprietary agents. Ever.
obseria.io ingests metrics via the OpenTelemetry Collector and the Kubernetes Metrics Server. Deploy our pre-built Helm chart and you're fully instrumented in under 5 minutes.
- Helm chart with zero-config defaults
- Auto-discovery of new namespaces & services
- PromQL-compatible metric queries
- Unified with your existing traces & logs
# Install via Helm
helm repo add dataleak
https://charts.obseria.io
helm install dataleak-k8s \
dataleak/kubernetes-monitoring \
--set apiKey=$DATALEAK_API_KEY \
--set cluster.name=production
# Done. Metrics flowing in ~60s.