chore: upgrade mooncake 0.3.5 (#8341)

Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com>

.github/workflows/pr-test-pd-router.yml

@@ -114,7 +114,7 @@ jobs:
       run: |
         echo "Installing SGLang with all extras..."
         python3 -m pip --no-cache-dir install -e "python[all]" --break-system-packages
-        python3 -m pip --no-cache-dir install mooncake-transfer-engine==0.3.4.post2
+        python3 -m pip --no-cache-dir install mooncake-transfer-engine==0.3.5
 
     - name: Build and install sgl-router
       run: |

docker/Dockerfile

@@ -86,7 +86,7 @@ RUN wget https://developer.download.nvidia.com/compute/redist/nvshmem/3.3.9/sour
 # Python tools
 RUN python3 -m pip install --no-cache-dir \
     datamodel_code_generator \
-    mooncake_transfer_engine==0.3.4.post2 \
+    mooncake-transfer-engine==0.3.5 \
     pre-commit \
     pytest \
     black \

docker/Dockerfile.gb200

@@ -145,14 +145,8 @@ RUN apt update -y \
     && apt update -y \
     && apt install nsight-systems-cli -y
 
-RUN git clone https://github.com/kvcache-ai/Mooncake.git \
-    && cd Mooncake \
-    && bash dependencies.sh -y \
-    && mkdir build \
-    && cd build \
-    && cmake .. -DUSE_MNNVL=ON \
-    && make -j \
-    && make install
+# --- Install Mooncake ---
+RUN pip install mooncake-transfer-engine==0.3.5
 
 # Set up locale
 RUN locale-gen en_US.UTF-8

scripts/ci_install_dependency.sh

@@ -23,7 +23,7 @@ pip install -e "python[dev]"
 pip list
 
 # Install additional dependencies
-pip install mooncake-transfer-engine==0.3.4.post2 nvidia-cuda-nvrtc-cu12
+pip install mooncake-transfer-engine==0.3.5 nvidia-cuda-nvrtc-cu12
 
 # For lmms_evals evaluating MMMU
 git clone --branch v0.3.3 --depth 1 https://github.com/EvolvingLMMs-Lab/lmms-eval.git