Make CI happy

Signed-off-by: Yikun Jiang <yikunkero@gmail.com>

Author: Yikun

.github/workflows/vllm_ascend_doctest.yaml

@@ -15,15 +15,21 @@
 # This file is a part of the vllm-ascend project.
 #
 
-name: 'e2e test'
+name: 'Doctest'
 
 on:
+  workflow_dispatch:
   pull_request:
     branches:
       - 'main'
       - '*-dev'
     paths:
+      # If we are changing the doctest we should do a PR test
       - '.github/workflows/vllm_ascend_doctest.yaml'
+      - 'tests/doctest/**'
+  schedule:
+    # Runs every 2 hours
+    - cron:  '0 */2 * * *'
 
 # Bash shells do not use ~/.profile or ~/.bashrc so these shells need to be explicitly
 # declared as "shell: bash -el {0}" on steps that need to be properly activated.
@@ -38,36 +44,58 @@ jobs:
       # Each version should be tested
       fail-fast: false
       matrix:
-        vllm_verison: [main, v0.7.3-dev, v0.8.4rc2, v0.8.4rc1]
+        vllm_verison: [main, v0.7.3-dev, main-openeuler, v0.7.3-dev-openeuler]
     name: vLLM Ascend test
     runs-on: linux-arm64-npu-1
     container:
       image: m.daocloud.io/quay.io/ascend/vllm-ascend:${{ matrix.vllm_verison }}
     steps:
-      - name: Check npu and CANN info
+      - name: Check NPU/CANN and git info
         run: |
-          pwd
+          echo "====> Prnint NPU/CANN info"
           npu-smi info
           cat /usr/local/Ascend/ascend-toolkit/latest/"$(uname -i)"-linux/ascend_toolkit_install.info
 
-      - name: Config mirrors
+          echo "====> Prnint vllm-ascend git info"
+          cd /vllm-workspace/vllm-ascend
+          git --no-pager log -1 || true
+          echo "====> Prnint vllm git info"
+          cd /vllm-workspace/vllm
+          git --no-pager log -1 || true
+
+      - name: Config OS mirrors
+        if: ${{ !endsWith(matrix.vllm_verison, '-openeuler') }}
         run: |
           sed -i 's|ports.ubuntu.com|mirrors.tuna.tsinghua.edu.cn|g' /etc/apt/sources.list
-          pip config set global.index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple
           apt-get update -y
           apt install git curl -y
+
+      - name: Config OS mirrors
+        if: ${{ endsWith(matrix.vllm_verison, '-openeuler') }}
+        run: |
+          yum update -y
+          yum install git curl -y
+
+      - name: Config pip mirrors
+        run: |
+          pip config set global.index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple
           git config --global url."https://gh-proxy.test.osinfra.cn/https://github.com/".insteadOf https://github.com/
 
       - name: Checkout vllm-project/vllm-ascend repo
         uses: actions/checkout@v4
 
-      - name: Run vllm-project/vllm test for V0 Engine
+      - name: Run vllm-ascend/tests/doctest/run_tests.sh
         run: |
-          pwd
-          ls -la
-          if [ ! -d /workspace/vllm-ascend/tests/doctest ]; then
+          # PWD: /__w/vllm-ascend/vllm-ascend
+          # Address old branch like v0.7.3:
+          if [ ! -d /vllm-workspace/vllm-ascend/tests/doctest ]; then
             echo "Warning: the doctest path doesn't exists, copy now"
-            cp -r tests/doctest /workspace/vllm-ascend/tests/
+            cp -r tests/doctest /vllm-workspace/vllm-ascend/tests/
           fi
+
+          # Simulate container to enter directory
           cd /workspace
-          ./vllm-ascend/tests/doctest/run_tests.sh
+
+          # Run real test
+          echo "Test:"
+          /vllm-workspace/vllm-ascend/tests/doctest/run_tests.sh

docs/source/quick_start.md

@@ -146,9 +146,8 @@ vLLM is serving as background process, you can use `kill -2 $VLLM_PID` to stop t
 it's equal to `Ctrl-C` to stop foreground vLLM process:
 
 ```bash
-ps -ef | grep "/.venv/bin/vllm serve" | grep -v grep
-VLLM_PID=`ps -ef | grep "/.venv/bin/vllm serve" | grep -v grep | awk '{print $2}'`
-kill -2 $VLLM_PID
+  VLLM_PID=$(pgrep -f "vllm serve")
+  kill -2 "$VLLM_PID"
 ```
 
 You will see output as below:

tests/doctest/001-quickstart-test.sh

@@ -27,38 +27,26 @@ function wait_url_ready() {
   i=0
   while true; do
     _info "===> Waiting for ${serve_name} to be ready...${i}s"
-    i=$((i + ${CURL_COOLDOWN}))
-
+    i=$((i + CURL_COOLDOWN))
     set +e
-
-    curl \
-      --silent \
-      --max-time "$CURL_TIMEOUT" \
-      ${url} \
-      >/dev/null
-
+    curl --silent --max-time "$CURL_TIMEOUT" "${url}" >/dev/null
     result=$?
-
     set -e
-
     if [ "$result" -eq 0 ]; then
       break
     fi
-
     if [ "$i" -gt "$CURL_MAX_TRIES" ]; then
       _info "===> \$CURL_MAX_TRIES exceeded waiting for ${serve_name} to be ready"
       return 1
     fi
-
     sleep "$CURL_COOLDOWN"
   done
-
   _info "===> ${serve_name} is ready."
 }
 
 function wait_for_exit() {
   local VLLM_PID="$1"
-  while kill -0 $VLLM_PID; do
+  while kill -0 "$VLLM_PID"; do
     _info "===> Wait for ${VLLM_PID} to exit."
     sleep 1
   done
@@ -71,7 +59,7 @@ function simple_test() {
 
 function quickstart_offline_test() {
   export VLLM_USE_MODELSCOPE=true
-  python3 ${SCRIPT_DIR}/../../examples/offline_inference_npu.py
+  python3 "${SCRIPT_DIR}/../../examples/offline_inference_npu.py"
 }
 
 function quickstart_online_test() {
@@ -86,10 +74,10 @@ function quickstart_online_test() {
         "max_tokens": 5,
         "temperature": 0
     }' | python3 -m json.tool
-  ps -ef | grep "vllm serve" | grep -v grep
-  VLLM_PID=$(ps -ef | grep "vllm serve" | grep -v grep | awk '{print $2}')
-  kill -2 $VLLM_PID
-  wait_for_exit $VLLM_PID
+  VLLM_PID=$(pgrep -f "vllm serve")
+  _info "===> Try kill -2 ${VLLM_PID} to exit."
+  kill -2 "$VLLM_PID"
+  wait_for_exit "$VLLM_PID"
 }
 
 _info "====> Start simple_test"