Diagnostics and Notifications

The webserver continuously samples system health and records diagnostic history. A diagnostics page surfaces live metrics, historical trends, and saved snapshots, while a toast notification system pushes transient alerts to every connected browser.

Diagnostics

The diagnostics subsystem tracks:

API request latency and error rate
CPU, memory, and disk utilization
ROS diagnostic status and fault reporting

Recent activity is held in an in-memory ring buffer, and periodic snapshots are persisted to disk so that history survives restarts. Snapshots are retained for a configurable number of days and bounded by a maximum disk budget.

Configuration

Diagnostics behavior is configured under the diagnostics key in robot_info.yaml.

Key	Default
`diagnostics.enabled`	`true`
`diagnostics.ring_buffer.max_entries`	`5000`
`diagnostics.storage.data_dir`	`diagnostics`
`diagnostics.storage.snapshot_interval_sec`	`3600`
`diagnostics.storage.retention_days`	`7`
`diagnostics.storage.max_disk_mb`	`500`
`diagnostics.resources.sample_interval_sec`	`5`
`diagnostics.thresholds.cpu_percent`	`85.0`
`diagnostics.thresholds.memory_percent`	`90.0`
`diagnostics.thresholds.disk_percent`	`95.0`
`diagnostics.thresholds.api_latency_ms`	`2000`
`diagnostics.thresholds.api_error_rate_percent`	`10.0`

When a measured value crosses its threshold, the diagnostics subsystem flags the condition and can raise a notification.

Notifications

Toast notifications are short, transient messages delivered to all connected browsers over a server-sent event (SSE) stream. They are used for status changes, threshold breaches, and one-shot informational messages.

The notifications key in robot_info.yaml controls:

queue size
default display duration
severity colors and icons

Integration Endpoints

Endpoint	Purpose
`/api/diagnostics/*`	Live metrics, history, snapshots, and configuration.
`/api/notifications/*`	SSE toast stream and notification pushes.