[RFC] Replace shared-memory routed experts with ModelRunnerOutput transfer and HTTP support (#39568)

Signed-off-by: xhx1022 <1737006628@qq.com>
Signed-off-by: arlenxu <arlenxu@tencent.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: arlenxu <arlenxu@tencent.com>
Co-authored-by: Junjie Zhang <junj.jay.zhang@gmail.com>
This commit is contained in:
Hongxin Xu
2026-05-14 14:12:30 +00:00
committed by GitHub
co-authored by Claude Opus 4.6 arlenxu Junjie Zhang
parent 2317682f95
commit c7560af424
12 changed files with 605 additions and 369 deletions
+24
View File
@@ -785,6 +785,30 @@ class VllmConfig:
self.parallel_config.is_moe_model = self.model_config.is_moe
if (
self.model_config is not None
and self.model_config.enable_return_routed_experts
):
if self.parallel_config.pipeline_parallel_size > 1:
raise ValueError(
"--enable-return-routed-experts is incompatible with "
"pipeline parallelism (PP > 1)."
)
# Incompatible with any KV connector — covers both PD disaggregation
# (kv_producer/kv_consumer: routing captured on P can't reach D) and
# single-instance KV offload/sharing (kv_both: slot_mapping semantics
# change when KV blocks live outside local GPU memory, breaking the
# slot-indexed routed_experts buffer).
if (
self.kv_transfer_config is not None
and self.kv_transfer_config.is_kv_transfer_instance
):
raise ValueError(
"--enable-return-routed-experts is incompatible with KV "
"connectors (PD disaggregation, KV cache offload)."
)
if self.lora_config is not None:
self.lora_config.verify_with_model_config(self.model_config)