Shared Runtime Health
System Status
One place for runtime, services, compute readiness, and operator follow-up. Use the summary tab first, then drill down only into the failing lane.
Main action
Refresh metrics, confirm the summary state, then use runtime, services, compute, or JSON only as supporting drill-down panels.
System Status
Shared health view for runtime, services, compute readiness, and operator follow-up.
CPU: --%
RAM: --%
GPU: --
MPR: --
FPS: --
Cloud: --
Updated: --:--:--
|
Use these controls when the page looks stale or the operator needs a fresh snapshot.
Operator checklist
- Sustained high CPU or RAM can reduce live FPS and operator responsiveness.
- Video health shows whether the stream is active, degraded, or stalled.
- Check Healthz and Ready to confirm that the required services are available.
Current runtime state
--
--
--
Source: --
Golden path dashboard
Latest run proof from camera to reset. Use the first failed hop before opening deeper debug pages.
Latest run: --
--
runtime: --
fps gate: --
First failed hop
--
Next action
--
Report proof
--
Source
--
Run ledger manifest
Read-only evidence packet for the latest snapshot. It is not auto-persisted by the runtime.
Status: --
Run: --
Runtime: --
FPS: --
Proof gaps: --
CPU
--%
Temp: -- C
Memory
--%
-- / -- GB
GPU
--
--
Live runtime FPS
--
Observed during runtime
Capacity guidance
--
--
--
Target FPS: --
System CPU: --
Process CPU: --
RAM: --
GPU usage: --
GPU memory: --
This guidance is based on recent metrics. Use sustained load and p95 evidence before making hardware or deployment decisions.
Readiness checks
--
Service Ready: -- | Live Ready: --
Live Ready reason: --
Phase 2: --
Service URL: --
Ingest: --
Payload: --
DB: --
Missing blueprints:
GPU details
Name--
Usage--
Memory--
Temperature--
NVML: --
--
Backend process
CPU%--
Memory (RSS)-- GB
Disk
Usage--%
Read--
Write--
Network
Receive--
Send--
Video health
--
Status--
Reason--
Since--
--
Recent events
CPU by component
Combined view from /api/system, /api/payload_status, /api/diagnostics, and /api/mpr/status.
| Component | CPU | Memory | Status |
|---|---|---|---|
| Waiting for system metrics... | |||
CPU by component (latest)
CPU history (last 10)
Environment
Host: --
Operating system: --
Python: --
PID: --
Uptime: --
NVML/CUDA: --
App version: --
Payload version: --
Git commit: --
Cloud mode: --
Core service status
Healthz
--
reason: --
payload/gpu/video --
warn: --
Ready
--
templates/static/monitor/streamer --
Session / Video
--
--
Diagnostics
--
--
ObjDet/MediaPipe: --
MPR compute: --
Compute mode
--
Locked by environment settings. Changes are not allowed.
Choose the GPU or CPU policy for the current runtime environment.
Status (requested vs resolved)
Requested: --
Resolved: --
Reason: --
Strictness: --
GPU count: --
GPU name: --
GPU build: --
Source: --
ENV override: --
Updated: --
Updated by: --
Stage device summary
GPU and pipeline summary
--
Source: /api/mpr/status (perf_window) + /api/system (GPU)
Observed FPS: --
Budget FPS: --
Samples: --
GPU usage: --
GPU memory: --
Infer (last): --
Stage avg/max (ms)
--
Raw system snapshot (JSON)
Includes system, health, diagnostics, and payload status for deeper follow-up.
{}