All issues

Search results

Model load/unload over HTTP blocks inference requests — fix already open in #8737

Status: Open.

#8958 In triton-inference-server/server;

· tnsahr2580 opened on Sep 4, 2026

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

performanceA possible performance tune-upA possible performance tune-up

Status: Open.

Bug

#8947 In triton-inference-server/server;

· dengroc opened on Aug 31, 2026

Triton TRT-LLM Multimodal guide outdated

Status: Open.

#8945 In triton-inference-server/server;

· faradawn opened on Aug 31, 2026

vLLM backend passes deprecated raw prompts to InputProcessor

Status: Open.

#8924 In triton-inference-server/server;

· DimensionSTP opened on Aug 10, 2026

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

Status: Open.

#8922 In triton-inference-server/server;

· protonicage opened on Aug 7, 2026

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

Status: Open.

#8919 In triton-inference-server/server;

· Moyummmm opened on Aug 6, 2026

[vllm_backend] Switch to current engine to support the new multi modal models in vllm

Status: Open.

#8912 In triton-inference-server/server;

· protonicage opened on Jul 30, 2026

How to build a lite windows version?

Status: Open.

#8911 In triton-inference-server/server;

· yanrunming opened on Jul 30, 2026

ONNXRuntime won't build

Status: Open.

#8906 In triton-inference-server/server;

· kpedro88 opened on Jul 28, 2026

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

Status: Open.

#8894 In triton-inference-server/server;

· SlackerProMax opened on Jul 21, 2026

Cuda SHM Module Cuda driver entry point API mismatch

Status: Open.

#8884 In triton-inference-server/server;

· RohanAsokan-Perceptyne opened on Jul 15, 2026

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

Status: Open.

#8882 In triton-inference-server/server;

· CarlosNacher opened on Jul 14, 2026