Yukab Panel Stress Diagnostic
Controlled workload generator that measures Yukab CPU, memory, persistent storage, bandwidth, logging, long-request, health, and cgroup behavior.
STRESS_TOKEN to a strong random secret. Stress endpoints stay disabled when the token is absent.Full diagnostic report limit-aware
The diagnostic now reads the real cgroup RAM/CPU limits and current /data capacity before choosing workload sizes. It will not accept impossible 25–50 GiB values on a small server.
CPU saturation
Runs worker threads continuously for a bounded duration.
Memory pressure
Commits a real Buffer. Requests are server-side clamped to the detected safe cgroup headroom.
Persistent disk write
Writes real bytes to /data/yukab-stress-test.bin. Requests are clamped to available persistent capacity.
Bandwidth stream
Streams an exact binary response through the reverse proxy.
Log pressure
Writes structured lines to stdout to exercise Docker log limits and rotation.
Held connection
Leaves an HTTP response open for a bounded period to exercise proxy timeout behavior.
Live status manual
Enter STRESS_TOKEN and click Detect limits, or run the full diagnostic.
No diagnostic report yet.