From ac9a7d130148def4636431cffe337dfea0e78c1b Mon Sep 17 00:00:00 2001 From: Li Wang Date: Wed, 25 Feb 2026 10:14:51 +0800 Subject: [PATCH] [Nightly] Increase VLLM_ENGINE_READY_TIMEOUT_S to avoid nightly failure (#6778) ### What this PR does / why we need it? After some observation, I found some cases failed for timeout, just like https://github.com/vllm-project/vllm-ascend/actions/runs/22280996034/job/64487867977#step:9:921 and https://github.com/vllm-project/vllm-ascend/actions/runs/22315540111/job/64574590762#step:9:1809, this may caused by the excessively long model loading time (currently we are still loading weights from network storage), it is necessary to adjust the timeout seconds 600s -> 1800s ### Does this PR introduce _any_ user-facing change? ### How was this patch tested? - vLLM version: v0.15.0 - vLLM main: https://github.com/vllm-project/vllm/commit/9562912cead1f11e8540fb91306c5cbda66f0007 Signed-off-by: wangli --- tests/e2e/nightly/multi_node/scripts/run.sh | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tests/e2e/nightly/multi_node/scripts/run.sh b/tests/e2e/nightly/multi_node/scripts/run.sh index 6d8b774a..0aced732 100644 --- a/tests/e2e/nightly/multi_node/scripts/run.sh +++ b/tests/e2e/nightly/multi_node/scripts/run.sh @@ -28,6 +28,8 @@ export VLLM_LOGGING_LEVEL="INFO" export GLOG_minloglevel=1 # Set transformers to offline mode to avoid downloading models during tests export HF_HUB_OFFLINE="1" +# Default is 600s +export VLLM_ENGINE_READY_TIMEOUT_S=1800 # Function to print section headers print_section() {