Per-second metrics platform
Netra collects per-second host, network, workload and application metrics on
Metric alerts
internal/metricalert evaluates threshold and anomaly-rate rules against the
Anomaly detection
Every per-second metric dimension (see metrics.md) gets its own
Application collectors
The agent can scrape the status endpoints of common infrastructure software
Multi-channel alerting
Netra can push anomaly findings to one or more notification channels on an
Netra controller high availability
Netra v0.7 retains the active/passive controller HA introduced in v0.6. Kubernetes coordination.k8s.io/v1 Lease election chooses a single active replica. A candidate is not promoted until it also acquires the shared state file lock.
Multi-cluster fleet (read-only)
GET /api/v1/fleet remains the single-cluster inventory.
Flow history, RED, traces
The live Connections page is still a short sample. This is the longer