Module health 

GET /health and GET /health/deep handlers.

/health returns lightweight readiness status from in-memory atomics. /health/deep runs a tiny canary inference (batch=1, seq≈8 tokens) and returns 503 if the actual embedding pipeline is broken.

Constants

DEEP_HEALTH_CANARY 🔒

Fixed canary text for /health/deep. Tokenises to ~8 tokens with the BGE-M3 SentencePiece vocabulary, meeting the “batch=1, seq=8” goal.

DEEP_HEALTH_TIMEOUT 🔒

Timeout for the canary embed call in /health/deep.

Functions

health

Handles GET /health — returns readiness status, worker counts, and tuning diagnostics.

health_deep

Handles GET /health/deep — runs a tiny canary inference and returns 503 if the actual embedding pipeline is broken.