fix(agentx): rely on upstream VLLM_PORT handling

Remove the redundant local srt-slurm patch now that main unsets VLLM_PORT for multi-node vLLM endpoints.

中文：依赖上游的 VLLM_PORT 处理。当前 srt-slurm main 已会为跨节点 vLLM 端点取消设置 VLLM_PORT，因此移除冗余的本地补丁。

Author: ivanium

perf-changelog.yaml

@@ -5034,6 +5034,6 @@
     - dsv4-fp4-gb200-dynamo-vllm-agentic-3p1d-dep8-dep16
   description:
     - "Add GB200 Dynamo-vLLM AgentX aggregate TP8 at conc [16] and disaggregated 3P/1D DEP8/DEP16 at conc [1536]."
-    - "Use NVIDIA/srt-slurm main with its upstream per-node DP, Dynamo health-count, and Mooncake master fixes; retain the unmerged multi-node TP VLLM_PORT fix. TP8 uses FlashInfer all-reduce and disables vLLM custom all-reduce."
+    - "Use NVIDIA/srt-slurm main with its upstream per-node DP, multi-node TP VLLM_PORT, Dynamo health-count, and Mooncake master fixes. TP8 uses FlashInfer all-reduce and disables vLLM custom all-reduce."
     - "Image: vllm/vllm-openai:nightly-dev-arm64-cu13.0.1-c188b96"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2300

runners/launch_gb200-nv.sh

@@ -298,18 +298,11 @@ fi
 # TODO(CJQ): make first class upon srt-slurm upstream refactor
 if [[ "$IS_AGENTIC" == "1" ]]; then
     # Agentic recipes use NVIDIA/srt-slurm:main. The per-node DP launcher,
-    # matching Dynamo health counts, and Mooncake master compatibility are all
-    # upstream; keep only the still-unmerged multi-node TP port fix below.
+    # matching Dynamo health counts, multi-node TP VLLM_PORT handling, and
+    # Mooncake master compatibility are all upstream.
     git clone --branch main --single-branch https://github.com/NVIDIA/srt-slurm.git "$SRT_REPO_DIR" || exit 1
     cd "$SRT_REPO_DIR" || exit 1
 
-    # NVIDIA/srt-slurm#243 gives each srt process a private VLLM_PORT range, but
-    # a multi-node TP process still passes that one range to all of its local
-    # vLLM children. Unset it only for multi-node non-DP workers, matching the
-    # still-unmerged NVIDIA/srt-slurm commit de1a4f0257dae5bf871881dc4696e35389c37483.
-    # Keep the upstream allocation for per-node DP and co-located endpoints.
-    git apply "$GITHUB_WORKSPACE/runners/patches/srt-slurm-vllm-multinode-tp-port.patch" || exit 1
-
     mkdir -p recipes/vllm/deepseek-v4/agentic || exit 1
     cp -rT "$GITHUB_WORKSPACE/benchmarks/multi_node/srt-slurm-recipes/vllm/deepseek-v4/agentic" \
         recipes/vllm/deepseek-v4/agentic || exit 1