server · Issues· 895 open

k8s-onprem autoscaling metric avg_time_queue_us under-reports queue time by roughly the number of loaded models

python_backend: error-string shm freed before parent reads it in is_ready() readiness check -> allocator corruption -> permanent hang

TensorRT backend leaks GPU resources after model initialization failure; unload does not recover memory

Random latency regression on L40S nodes after new pod creation — node becomes persistently "bad"

[Server] Support for numpy2.x

Intermittent server crash (`boost::interprocess::lock_exception` / `std::bad_alloc` / SIGSEGV) in Python backend under concurrent BLS load — 25.02 (2.55.0)

Triton TRT-LLM Multimodal guide outdated

Inaccuracies in PyTorch backend documentation

[Question] Reloading models and latency spikes

InferenceRequest.async_exec(decoupled=True) intermittently raises `RuntimeError: Invalid argument` under concurrent in-flight BLS to a decoupled model

vLLM backend passes deprecated raw prompts to InputProcessor

Python stub aborted with glibc heap corruption, leading to 0 RPS on gRPC serving thread

test: L0_lifecycle uses blind time.sleep(5) for model load/unload, causing flaky CI failures

[vLLM backend][Feature Request] Support vLLM-Omni for native audio/video input

[Question] Different performance of vLLM server and Triton Inference Server