Merge branch 'main' into test-cpu-32-runner

Author: JavaPythonAIForBAT

.github/workflows/_selected_tests.yaml

@@ -177,11 +177,11 @@ jobs:
             libcurl4
           ldconfig
 
-          MOONCAKE_WHEEL="mooncake_transfer_engine_ascend-0.3.9-cp312-cp312-manylinux_2_35_aarch64.whl"
           pip install --no-cache-dir --no-deps \
-            "https://vllm-ascend.obs.cn-north-4.myhuaweicloud.com/vllm-ascend/${MOONCAKE_WHEEL}"
+            -i https://mirrors.aliyun.com/pypi/simple/ \
+            mooncake-transfer-engine-npu==0.3.11.post1
 
-          pip show mooncake-transfer-engine-ascend || true
+          pip show mooncake-transfer-engine-npu || true
 
       - name: Install vllm-project/vllm-ascend with device
         if: ${{ matrix.group.npu_type != 'cpu' }}

Dockerfile

@@ -18,7 +18,7 @@
 FROM quay.io/ascend/cann:9.0.0-910b-ubuntu22.04-py3.12
 
 ARG PIP_INDEX_URL="https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple"
-ARG MOONCAKE_TAG="v0.3.9"
+ARG MOONCAKE_TAG="v0.3.11.post1"
 
 WORKDIR /workspace
 

Dockerfile.a3

@@ -18,7 +18,7 @@
 FROM quay.io/ascend/cann:9.0.0-a3-ubuntu22.04-py3.12
 
 ARG PIP_INDEX_URL="https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple"
-ARG MOONCAKE_TAG=v0.3.9
+ARG MOONCAKE_TAG=v0.3.11.post1
 
 COPY ./tools/mooncake_installer.sh /vllm-workspace/
 

Dockerfile.a3.openEuler

@@ -18,7 +18,7 @@
 FROM quay.io/ascend/cann:9.0.0-a3-openeuler24.03-py3.12
 
 ARG PIP_INDEX_URL="https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple"
-ARG MOONCAKE_TAG="v0.3.9"
+ARG MOONCAKE_TAG="v0.3.11.post1"
 
 WORKDIR /workspace
 

Dockerfile.a5

@@ -18,7 +18,7 @@
 FROM quay.io/ascend/cann:9.0.0-950-ubuntu22.04-py3.12
 
 ARG PIP_INDEX_URL="https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple"
-ARG MOONCAKE_TAG=v0.3.9
+ARG MOONCAKE_TAG=v0.3.11.post1
 
 COPY ./tools/mooncake_installer.sh /vllm-workspace/
 

Dockerfile.a5.openEuler

@@ -18,7 +18,7 @@
 FROM quay.io/ascend/cann:9.0.0-950-openeuler24.03-py3.12
 
 ARG PIP_INDEX_URL="https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple"
-ARG MOONCAKE_TAG="v0.3.9"
+ARG MOONCAKE_TAG="v0.3.11.post1"
 
 WORKDIR /workspace
 

Dockerfile.openEuler

@@ -18,7 +18,7 @@
 FROM quay.io/ascend/cann:9.0.0-910b-openeuler24.03-py3.12
 
 ARG PIP_INDEX_URL="https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple"
-ARG MOONCAKE_TAG="v0.3.9"
+ARG MOONCAKE_TAG="v0.3.11.post1"
 
 WORKDIR /workspace
 

docs/source/locale/zh_CN/LC_MESSAGES/user_guide/feature_guide/kv_pool.po

@@ -76,8 +76,8 @@ msgid "vLLM-Ascend：main branch"
 msgstr "vLLM-Ascend：main 分支"
 
 #: ../../source/user_guide/feature_guide/kv_pool.md:17
-msgid "mooncake：>= 0.3.9"
-msgstr "mooncake：>= 0.3.9"
+msgid "mooncake：>= 0.3.11.post1"
+msgstr "mooncake：>= 0.3.11.post1"
 
 #: ../../source/user_guide/feature_guide/kv_pool.md:19
 msgid "KV Pool Parameter Description"
@@ -294,9 +294,9 @@ msgstr "800 I/T A3 系列"
 
 #: ../../source/user_guide/feature_guide/kv_pool.md
 msgid ""
-"HDK >= 26.0<br>or HDK >= 25.5 with mooncake >= v0.3.11<br>CANN >= "
+"HDK >= 26.0<br>or HDK >= 25.5 with mooncake >= v0.3.11.post1<br>CANN >= "
 "9.0.0<br>LingQu Computing Network >= 1.5"
-msgstr "HDK >= 26.0<br>或 HDK >= 25.5 且 mooncake >= v0.3.11<br>CANN >= 9.0.0<br>LingQu 计算网络 >= 1.5"
+msgstr "HDK >= 26.0<br>或 HDK >= 25.5 且 mooncake >= v0.3.11.post1<br>CANN >= 9.0.0<br>LingQu 计算网络 >= 1.5"
 
 #: ../../source/user_guide/feature_guide/kv_pool.md
 msgid "`export ASCEND_ENABLE_USE_FABRIC_MEM=1`"
@@ -534,8 +534,8 @@ msgid "Enable MooncakeStore SSD Offload with Embedded Real Client Mode"
 msgstr "使用嵌入式真实客户端模式启用 MooncakeStore SSD 卸载"
 
 #: ../../source/user_guide/feature_guide/kv_pool.md:430
-msgid "Requires mooncake >= v0.3.11."
-msgstr "此功能需要 mooncake >= v0.3.11。"
+msgid "Requires mooncake >= v0.3.11.post1."
+msgstr "此功能需要 mooncake >= v0.3.11.post1。"
 
 #: ../../source/user_guide/feature_guide/kv_pool.md:432
 msgid "Start the master"
@@ -1272,8 +1272,8 @@ msgstr ""
 #~ msgid "Embedded Real Client Mode（Mooncake ssd-offload.md Step 3A）"
 #~ msgstr "嵌入式真实客户端模式（Mooncake ssd-offload.md 步骤 3A）"
 
-#~ msgid "mooncake >= v0.3.11"
-#~ msgstr "mooncake >= v0.3.11"
+#~ msgid "mooncake >= v0.3.11.post1"
+#~ msgstr "mooncake >= v0.3.11.post1"
 
 #~ msgid "Set `true` to enable SSD offload."
 #~ msgstr "设置为 `true` 以启用 SSD 卸载。"

docs/source/tutorials/features/pd_colocated_mooncake_multi_instance.md

@@ -121,7 +121,7 @@ Moonshot AI. Installation and compilation guide:
 First, obtain the Mooncake project using the following command:
 
 ```bash
-git clone -b v0.3.9 --depth 1 https://github.com/kvcache-ai/Mooncake.git
+git clone -b v0.3.11.post1 --depth 1 https://github.com/kvcache-ai/Mooncake.git
 cd Mooncake
 git submodule update --init --recursive
 ```

docs/source/tutorials/features/pd_disaggregation_mooncake_multi_node.md

@@ -177,7 +177,7 @@ Mooncake is the serving platform for Kimi, a leading LLM service provided by Moo
 First, we need to obtain the Mooncake project. Refer to the following command:
 
 ```shell
-git clone -b v0.3.9 --depth 1 https://github.com/kvcache-ai/Mooncake.git
+git clone -b v0.3.11.post1 --depth 1 https://github.com/kvcache-ai/Mooncake.git
 ```
 
 (Optional) Replace go install url if the network is poor