Merge branch 'develop' into wangjiawei04-patch-2

8391764f · MRXLT · GitHub · 9eff225e · 9c6d556b · 8391764f
20 changed file
--- a/README.md
+++ b/README.md
@@ -48,7 +48,7 @@ nvidia-docker exec -it test bash
 pip install paddle-serving-client==0.3.2 
 pip install paddle-serving-server==0.3.2 # CPU
 pip install paddle-serving-server-gpu==0.3.2.post9 # GPU with CUDA9.0
-pip install paddle-serving-server-gpu==0.3.2.post10 # GPU with CUDA9.0
+pip install paddle-serving-server-gpu==0.3.2.post10 # GPU with CUDA10.0
 ```

 You may need to use a domestic mirror source (in China, you can use the Tsinghua mirror source, add `-i https://pypi.tuna.tsinghua.edu.cn/simple` to pip command) to speed up the download.

--- a/core/cube/CMakeLists.txt
+++ b/core/cube/CMakeLists.txt
@@ -12,8 +12,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License

+#execute_process(COMMAND go env -w GO111MODULE=off)
 add_subdirectory(cube-server)
 add_subdirectory(cube-api)
 add_subdirectory(cube-builder)
-add_subdirectory(cube-transfer)
-add_subdirectory(cube-agent)
+#add_subdirectory(cube-transfer)
+#add_subdirectory(cube-agent)
--- a/core/general-server/op/general_response_op.cpp
+++ b/core/general-server/op/general_response_op.cpp
@@ -155,9 +155,11 @@ int GeneralResponseOp::inference() {
      }

      if (model_config->_is_lod_fetch[idx]) {
-        for (int j = 0; j < in->at(idx).lod[0].size(); ++j) {
-          fetch_p->mutable_tensor_array(var_idx)->add_lod(
-              in->at(idx).lod[0][j]);
+        if (in->at(idx).lod.size() > 0) {
+          for (int j = 0; j < in->at(idx).lod[0].size(); ++j) {
+            fetch_p->mutable_tensor_array(var_idx)->add_lod(
+                in->at(idx).lod[0][j]);
+          }
        }
      }


--- a/doc/COMPILE.md
+++ b/doc/COMPILE.md
@@ -75,6 +75,8 @@ export PATH=$PATH:$GOPATH/bin
 ## Get go packages

 ```shell
+go env -w GO111MODULE=on
+go env -w GOPROXY=https://goproxy.cn,direct
 go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-grpc-gateway
 go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-swagger
 go get -u github.com/golang/protobuf/protoc-gen-go

--- a/doc/COMPILE_CN.md
+++ b/doc/COMPILE_CN.md
@@ -72,6 +72,8 @@ export PATH=$PATH:$GOPATH/bin
 ## 获取 Go packages

 ```shell
+go env -w GO111MODULE=on
+go env -w GOPROXY=https://goproxy.cn,direct
 go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-grpc-gateway
 go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-swagger
 go get -u github.com/golang/protobuf/protoc-gen-go

--- a/doc/FAQ.md
+++ b/doc/FAQ.md
@@ -37,3 +37,25 @@
    client端的日志直接打印到标准输出。

    通过在部署服务之前 'export  GLOG_v=3'可以输出更为详细的日志信息。
+    
+- Q: GPU环境运行Serving报错，GPU count is: 0。
+  ```
+  terminate called after throwing an instance of 'paddle::platform::EnforceNotMet'
+  what():
+  --------------------------------------------
+  C++ Call Stacks (More useful to developers):
+  --------------------------------------------
+  0   std::string paddle::platform::GetTraceBackString<std::string const&>(std::string const&, char const*, int)
+  1   paddle::platform::SetDeviceId(int)
+  2   paddle::AnalysisConfig::fraction_of_gpu_memory_for_pool() const
+  3   std::unique_ptr<paddle::PaddlePredictor, std::default_delete<paddle::PaddlePredictor> > paddle::CreatePaddlePredictor<paddle::AnalysisConfig, (paddle::PaddleEngineKind)2>(paddle::AnalysisConfig const&)
+  4   std::unique_ptr<paddle::PaddlePredictor, std::default_delete<paddle::PaddlePredictor> > paddle::CreatePaddlePredictor<paddle::AnalysisConfig>(paddle::AnalysisConfig const&)
+  ----------------------
+  Error Message Summary:
+  ----------------------
+  InvalidArgumentError: Device id must be less than GPU count, but received id is: 0. GPU count is: 0.
+  [Hint: Expected id < GetCUDADeviceCount(), but received id:0 >= GetCUDADeviceCount():0.] at (/home/scmbuild/workspaces_cluster.dev/baidu.lib.paddlepaddle/baidu/lib/paddlepaddle/Paddle/paddle/fluid/platform/gpu_info.cc:211)
+  ```
+  A: libcuda.so没有链接成功。首先在机器上找到libcuda.so，ldd检查libnvidia版本与nvidia-smi中版本一致（libnvidia-fatbinaryloader.so.418.39，与NVIDIA-SMI 418.39 Driver Version: 418.39）,然后用export导出libcuda.so的路径即可（例如libcuda.so在/usr/lib64/，export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/lib64/）
+  
+ 
--- a/doc/LATEST_PACKAGES.md
+++ b/doc/LATEST_PACKAGES.md
@@ -36,6 +36,10 @@ https://paddle-serving.bj.bcebos.com/whl/paddle_serving_client-0.0.0-cp37-none-a
 ```
 https://paddle-serving.bj.bcebos.com/whl/paddle_serving_client-0.0.0-cp36-none-any.whl
 ```
+### Python 3.5
+```
+https://paddle-serving.bj.bcebos.com/whl/paddle_serving_client-0.0.0-cp35-none-any.whl
+```
 ### Python 2.7
 ```
 https://paddle-serving.bj.bcebos.com/whl/paddle_serving_client-0.0.0-cp27-none-any.whl

--- a/python/paddle_serving_client/__init__.py
+++ b/python/paddle_serving_client/__init__.py
@@ -354,8 +354,9 @@ class Client(object):
                                name))
                    result_map[name].shape = shape
                    if name in self.lod_tensor_set:
-                        result_map["{}.lod".format(
-                            name)] = result_batch_handle.get_lod(mi, name)
+                        tmp_lod = result_batch_handle.get_lod(mi, name)
+                        if np.size(tmp_lod) > 0:
+                            result_map["{}.lod".format(name)] = tmp_lod
                elif self.fetch_names_to_type_[name] == float32_type:
                    result_map[name] = result_batch_handle.get_float_by_name(
                        mi, name)
@@ -367,9 +368,9 @@ class Client(object):
                    shape = result_batch_handle.get_shape(mi, name)
                    result_map[name].shape = shape
                    if name in self.lod_tensor_set:
-                        result_map["{}.lod".format(
-                            name)] = result_batch_handle.get_lod(mi, name)
-
+                        tmp_lod = result_batch_handle.get_lod(mi, name)
+                        if np.size(tmp_lod) > 0:
+                            result_map["{}.lod".format(name)] = tmp_lod
                elif self.fetch_names_to_type_[name] == int32_type:
                    # result_map[name] will be py::array(numpy array)
                    result_map[name] = result_batch_handle.get_int32_by_name(
@@ -382,8 +383,9 @@ class Client(object):
                    shape = result_batch_handle.get_shape(mi, name)
                    result_map[name].shape = shape
                    if name in self.lod_tensor_set:
-                        result_map["{}.lod".format(
-                            name)] = result_batch_handle.get_lod(mi, name)
+                        tmp_lod = result_batch_handle.get_lod(mi, name)
+                        if np.size(tmp_lod) > 0:
+                            result_map["{}.lod".format(name)] = tmp_lod
            multi_result_map.append(result_map)
        ret = None
        if len(model_engine_names) == 1:

--- a/python/paddle_serving_client/io/__init__.py
+++ b/python/paddle_serving_client/io/__init__.py
@@ -74,7 +74,8 @@ def save_model(server_model_folder,
        fetch_var = model_conf.FetchVar()
        fetch_var.alias_name = key
        fetch_var.name = fetch_var_dict[key].name
-        fetch_var.is_lod_tensor = fetch_var_dict[key].lod_level >= 1
+        #fetch_var.is_lod_tensor = fetch_var_dict[key].lod_level >= 1
+        fetch_var.is_lod_tensor = 1
        if fetch_var_dict[key].dtype == core.VarDesc.VarType.INT64:
            fetch_var.fetch_type = 0
        if fetch_var_dict[key].dtype == core.VarDesc.VarType.FP32:

--- a/python/pipeline/gateway/proxy_server.go
+++ b/python/pipeline/gateway/proxy_server.go
@@ -25,7 +25,7 @@ import (
  "github.com/grpc-ecosystem/grpc-gateway/runtime"
  "google.golang.org/grpc"

-  gw "./proto"
+  gw "serving-gateway/proto"
 )

 //export run_proxy_server

--- a/python/setup.py.client.in
+++ b/python/setup.py.client.in
@@ -28,17 +28,11 @@ import util
 py_version = sys.version_info
        
 def copy_lib():
-    if py_version[0] == 2:
-        lib_list = ['libpython2.7.so.1.0', 'libssl.so.10', 'libcrypto.so.10'] 
-    elif py_version[1] == 6:
-        lib_list = ['libpython3.6m.so.1.0', 'libssl.so.10', 'libcrypto.so.10']
-    elif py_version[1] == 7:
-        lib_list = ['libpython3.7m.so.1.0', 'libssl.so.10', 'libcrypto.so.10']
    os.popen('mkdir -p paddle_serving_client/lib')
+    lib_list = ['${OPENSSL_CRYPTO_LIBRARY}', '${OPENSSL_SSL_LIBRARY}', 
+                '${PYTHON_LIBRARY}']
    for lib in lib_list:
-        r = os.popen('whereis {}'.format(lib))
-        text = r.read()
-        os.popen('cp {} ./paddle_serving_client/lib'.format(text.strip().split(' ')[1]))
+        os.popen('cp {} ./paddle_serving_client/lib'.format(lib))

 max_version, mid_version, min_version = util.python_version()

@@ -53,9 +47,6 @@ REQUIRED_PACKAGES = [
    'grpcio-tools >= 1.28.1'
 ]

-if not util.find_package("paddlepaddle") and not util.find_package("paddlepaddle-gpu"):
-    REQUIRED_PACKAGES.append("paddlepaddle")
-

 packages=['paddle_serving_client',
          'paddle_serving_client.proto',

--- a/python/setup.py.server.in
+++ b/python/setup.py.server.in
@@ -29,7 +29,7 @@ util.gen_pipeline_code("paddle_serving_server")

 REQUIRED_PACKAGES = [
    'six >= 1.10.0', 'protobuf >= 3.11.0', 'grpcio >= 1.28.1', 'grpcio-tools >= 1.28.1',
-    'paddle_serving_client', 'flask >= 1.1.1', 'paddle_serving_app'
+    'paddle_serving_client', 'flask >= 1.1.1', 'paddle_serving_app', 'func_timeout', 'pyyaml'
 ]

 packages=['paddle_serving_server',

--- a/python/setup.py.server_gpu.in
+++ b/python/setup.py.server_gpu.in
@@ -29,7 +29,7 @@ util.gen_pipeline_code("paddle_serving_server_gpu")

 REQUIRED_PACKAGES = [
    'six >= 1.10.0', 'protobuf >= 3.11.0', 'grpcio >= 1.28.1', 'grpcio-tools >= 1.28.1',
-    'paddle_serving_client', 'flask >= 1.1.1', 'paddle_serving_app'
+    'paddle_serving_client', 'flask >= 1.1.1', 'paddle_serving_app', 'func_timeout', 'pyyaml'
 ]

 packages=['paddle_serving_server_gpu',

--- a/python/util.py
+++ b/python/util.py
@@ -44,8 +44,8 @@ def gen_pipeline_code(package_name):
    ret = os.system(
        "cd {}/pipeline/gateway/proto/ && "
        "../../../../../third_party/install/protobuf/bin/protoc -I. "
-        "-I$GOPATH/src "
-        "-I$GOPATH/src/github.com/grpc-ecosystem/grpc-gateway/third_party/googleapis "
+        "-I$GOPATH/pkg/mod "
+        "-I$GOPATH/pkg/mod/github.com/grpc-ecosystem/grpc-gateway\@v1.15.2/third_party/googleapis "
        "--go_out=plugins=grpc:. "
        "gateway.proto".format(package_name))
    if ret != 0:
@@ -54,14 +54,18 @@ def gen_pipeline_code(package_name):
    ret = os.system(
        "cd {}/pipeline/gateway/proto/ && "
        "../../../../../third_party/install/protobuf/bin/protoc -I. "
-        "-I$GOPATH/src "
-        "-I$GOPATH/src/github.com/grpc-ecosystem/grpc-gateway/third_party/googleapis "
+        "-I$GOPATH/pkg/mod "
+        "-I$GOPATH/pkg/mod/github.com/grpc-ecosystem/grpc-gateway\@v1.15.2/third_party/googleapis "
        "--grpc-gateway_out=logtostderr=true:. "
        "gateway.proto".format(package_name))
    if ret != 0:
        exit(1)

    # pipeline grpc-gateway shared-lib
+    ret = os.system("cd {}/pipeline/gateway/ && go mod init serving-gateway".
+                    format(package_name))
+    ret = os.system("cd {}/pipeline/gateway/ && go mod vendor && go mod tidy".
+                    format(package_name))
    ret = os.system(
        "cd {}/pipeline/gateway && "
        "go build -buildmode=c-shared -o libproxy_server.so proxy_server.go".

--- a/tools/Dockerfile.centos6.cuda9.0-cudnn7.devel
+++ b/tools/Dockerfile.centos6.cuda9.0-cudnn7.devel
@@ -41,6 +41,12 @@ RUN yum -y install wget && \
    echo 'export LD_LIBRARY_PATH=/usr/local/python3.6/lib:$LD_LIBRARY_PATH' >> /root/.bashrc && \
    source /root/.bashrc && \
    cd .. && rm -rf Python-3.6.8* && \
+    wget https://github.com/protocolbuffers/protobuf/releases/download/v3.11.2/protobuf-all-3.11.2.tar.gz && \
+    tar zxf protobuf-all-3.11.2.tar.gz && \
+    cd protobuf-3.11.2 && \
+    ./configure && make -j4 && make install && \
+    make clean && \
+    cd .. && rm -rf protobuf-* &&\
    yum -y install epel-release && yum -y install patchelf libXext libSM libXrender && \
    yum clean all && \
    echo "export LANG=en_US.utf8" >> /root/.bashrc && \

--- a/tools/Dockerfile.centos6.devel
+++ b/tools/Dockerfile.centos6.devel
@@ -41,6 +41,12 @@ RUN yum -y install wget && \
    echo 'export LD_LIBRARY_PATH=/usr/local/python3.6/lib:$LD_LIBRARY_PATH' >> /root/.bashrc && \
    source /root/.bashrc && \
    cd .. && rm -rf Python-3.6.8* && \
+    wget https://github.com/protocolbuffers/protobuf/releases/download/v3.11.2/protobuf-all-3.11.2.tar.gz && \
+    tar zxf protobuf-all-3.11.2.tar.gz && \
+    cd protobuf-3.11.2 && \
+    ./configure && make -j4 && make install && \
+    make clean && \
+    cd .. && rm -rf protobuf-* && \
    yum -y install epel-release && yum -y install patchelf libXext libSM libXrender && \
    yum clean all && \
    localedef -c -i en_US -f UTF-8 en_US.UTF-8 && \

--- a/tools/Dockerfile.ci
+++ b/tools/Dockerfile.ci
@@ -34,6 +34,13 @@ RUN wget http://nixos.org/releases/patchelf/patchelf-0.10/patchelf-0.10.tar.bz2
    && cd .. \
    && rm -rf patchelf-0.10*

+RUN wget https://github.com/protocolbuffers/protobuf/releases/download/v3.11.2/protobuf-all-3.11.2.tar.gz && \
+    tar zxf protobuf-all-3.11.2.tar.gz && \
+    cd protobuf-3.11.2 && \
+    ./configure && make -j4 && make install && \
+    make clean && \
+    cd .. && rm -rf protobuf-*
+
 RUN yum install -y python3 python3-devel

 RUN yum -y update >/dev/null \

--- a/tools/Dockerfile.cuda10.0-cudnn7.devel
+++ b/tools/Dockerfile.cuda10.0-cudnn7.devel
@@ -5,7 +5,14 @@ RUN yum -y install wget >/dev/null \
    && yum -y install git openssl-devel curl-devel bzip2-devel python-devel \
    && yum -y install libSM-1.2.2-2.el7.x86_64 --setopt=protected_multilib=false \
    && yum -y install libXrender-0.9.10-1.el7.x86_64 --setopt=protected_multilib=false \
-    && yum -y install libXext-1.3.3-3.el7.x86_64 --setopt=protected_multilib=false
+    && yum -y install libXext-1.3.3-3.el7.x86_64 --setopt=protected_multilib=false 
+
+RUN wget https://github.com/protocolbuffers/protobuf/releases/download/v3.11.2/protobuf-all-3.11.2.tar.gz && \
+    tar zxf protobuf-all-3.11.2.tar.gz && \
+    cd protobuf-3.11.2 && \
+    ./configure && make -j4 && make install && \
+    make clean && \
+    cd .. && rm -rf protobuf-*

 RUN wget https://cmake.org/files/v3.2/cmake-3.2.0-Linux-x86_64.tar.gz >/dev/null \
    && tar xzf cmake-3.2.0-Linux-x86_64.tar.gz \

--- a/tools/Dockerfile.cuda9.0-cudnn7.devel
+++ b/tools/Dockerfile.cuda9.0-cudnn7.devel
@@ -6,6 +6,13 @@ RUN yum -y install wget >/dev/null \
    && yum -y install libXrender-0.9.10-1.el7.x86_64 --setopt=protected_multilib=false \
    && yum -y install libXext-1.3.3-3.el7.x86_64 --setopt=protected_multilib=false

+RUN wget https://github.com/protocolbuffers/protobuf/releases/download/v3.11.2/protobuf-all-3.11.2.tar.gz && \
+    tar zxf protobuf-all-3.11.2.tar.gz && \
+    cd protobuf-3.11.2 && \
+    ./configure && make -j4 && make install && \
+    make clean && \
+    cd .. && rm -rf protobuf-*
+
 RUN wget https://cmake.org/files/v3.2/cmake-3.2.0-Linux-x86_64.tar.gz >/dev/null \
    && tar xzf cmake-3.2.0-Linux-x86_64.tar.gz \
    && mv cmake-3.2.0-Linux-x86_64 /usr/local/cmake3.2.0 \

--- a/tools/serving_build.sh
+++ b/tools/serving_build.sh
@@ -18,14 +18,20 @@ function init() {
    export PYTHONROOT=/usr
    cd Serving
    export SERVING_WORKDIR=$PWD
+
    $PYTHONROOT/bin/python -m pip install -r python/requirements.txt
+    $PYTHONROOT/bin/python -m pip install paddlepaddle
+
    export GOPATH=$HOME/go
    export PATH=$PATH:$GOPATH/bin

-    go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-grpc-gateway
-    go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-swagger
-    go get -u github.com/golang/protobuf/protoc-gen-go
-    go get -u google.golang.org/grpc
+    go env -w GO111MODULE=on
+    go env -w GOPROXY=https://goproxy.cn,direct
+
+    go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-grpc-gateway@v1.15.2
+    go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-swagger@v1.15.2
+    go get -u github.com/golang/protobuf/protoc-gen-go@v1.4.3
+    go get -u google.golang.org/grpc@v1.33.0
 }

 function check_cmd() {
@@ -605,7 +611,7 @@ function python_test_grpc_impl() {

            # test load server config and client config in Server side
            cd criteo_ctr_with_cube # pwd: /Serving/python/examples/grpc_impl_example/criteo_ctr_with_cube
-
+<<COMMENT #comment for compile bug, todo fix conflict between grpc-gateway and cube-agent 
            check_cmd "wget https://paddle-serving.bj.bcebos.com/unittest/ctr_cube_unittest.tar.gz > /dev/null"
            check_cmd "tar xf ctr_cube_unittest.tar.gz"
            check_cmd "mv models/ctr_client_conf ./"
@@ -626,9 +632,11 @@ function python_test_grpc_impl() {
                echo "error with criteo_ctr_with_cube inference auc test, auc should > 0.67"
                exit 1
            fi
+COMMENT
+
            echo "grpc impl test success"
            kill_server_process
-            ps -ef | grep "cube" | grep -v grep | awk '{print $2}' | xargs kill
+            #ps -ef | grep "cube" | grep -v grep | awk '{print $2}' | xargs kill

            cd .. # pwd: /Serving/python/examples/grpc_impl_example
            ;;
@@ -665,6 +673,7 @@ function python_test_grpc_impl() {
            cd .. # pwd: /Serving/python/examples/grpc_impl_example

            # test load server config and client config in Server side
+<<COMMENT #comment for compile bug, todo fix conflict between grpc-gateway and cube-agent 
            cd criteo_ctr_with_cube # pwd: /Serving/python/examples/grpc_impl_example/criteo_ctr_with_cube

            check_cmd "wget https://paddle-serving.bj.bcebos.com/unittest/ctr_cube_unittest.tar.gz"
@@ -689,10 +698,11 @@ function python_test_grpc_impl() {
                echo "error with criteo_ctr_with_cube inference auc test, auc should > 0.67"
                exit 1
            fi
+COMMENT
            echo "grpc impl test success"
            kill_server_process
            ps -ef | grep "test_server_gpu" | grep -v serving_build | grep -v grep | awk '{print $2}' | xargs kill
-            ps -ef | grep "cube" | grep -v grep | awk '{print $2}' | xargs kill
+            #ps -ef | grep "cube" | grep -v grep | awk '{print $2}' | xargs kill
            cd .. # pwd: /Serving/python/examples/grpc_impl_example
            ;;
        *)
@@ -829,8 +839,8 @@ EOF
            kill_process_by_port 18080
            
            # test: process servicer & thread op
-            pip uninstall grpcio -y
-            pip install grpcio --no-binary=grpcio
+            #pip uninstall grpcio -y
+            #pip install grpcio --no-binary=grpcio
            cat << EOF > config.yml
 rpc_port: 18080
 worker_num: 4
@@ -944,7 +954,7 @@ function python_run_test() {
    local TYPE=$1 # pwd: /Serving
    cd python/examples # pwd: /Serving/python/examples
    python_test_fit_a_line $TYPE # pwd: /Serving/python/examples
-    python_run_criteo_ctr_with_cube $TYPE # pwd: /Serving/python/examples
+    #python_run_criteo_ctr_with_cube $TYPE # pwd: /Serving/python/examples
    python_test_bert $TYPE # pwd: /Serving/python/examples
    python_test_imdb $TYPE # pwd: /Serving/python/examples
    python_test_lac $TYPE # pwd: /Serving/python/examples