Main Content

MATLAB Production Server Java Client Basics

The mps_client.jar Java® client library lets you evaluate MATLAB® functions deployed on remote servers using native Java data.

Obtain mps_client.jar Client Library

There are several options to obtain the mps_client.jar client library based on your project set up:

  • In a MATLAB Production Server™ installation, mps_client.jar is located in $MPS_INSTALL/client/java.

  • The library is available for download at MATLAB Production Server Client Libraries. Select your release to download the folder, then unzip it. mps_client.jar is located in /java.

  • The library is also hosted in a Maven™ repository at https://mvnrepository.com/artifact/com.mathworks.prodserver/mps_java_client. To use the jar in your Maven project, include the following coordinates in the pom.xml file:

    <!-- https://mvnrepository.com/artifact/com.mathworks.prodserver/mps_java_client -->
    <dependency>
        <groupId>com.mathworks.prodserver</groupId>
        <artifactId>mps_java_client</artifactId>
        <version>release_number</version>
    </dependency>
    

Configure your development environment to use mps_client.jar by adding it to your Java class path.

Choose Workflow for Client-Server Communication

The Java client API offers two workflows for client-server communication:

Use MWHttpClient Class

This workflow uses the MWHttpClient class and hides the implementation details of request creation and data serialization when evaluating MATLAB functions deployed on servers. This workflow provides the following two options to evaluate a deployed MATLAB function.

Based on your requirements, decide if the client uses a static proxy or a dynamic proxy.

  • A static proxy uses an object implementing an interface that mirrors the deployed MATLAB functions. You provide the interface for the static proxy. This is a type-safe API that enforces passing the proper data types to the function at compile time.

    See Static Proxy Interface Guidelines.

  • A dynamic proxy creates server requests based on the MATLAB function name provided to the invoke() method. You pass the function name as a parameter to the proxy along with the function arguments. You provide the function name, the number of output arguments, and all of the input arguments required to evaluate the functions. Doing so defers type checking until runtime.

    See Invoke MATLAB Functions Dynamically.

Write Java code to instantiate a proxy to a MATLAB Production Server instance and call the MATLAB functions.

  1. Create an MWClient object for communicating with the service hosted by a MATLAB Production Server instance.

  2. Create MATLAB data structures to hold the data passed between the client and server.

  3. Invoke MATLAB functions.

  4. Free system resources using the close method of the MWClient object.

Use RESTful API and protobuf

This workflow uses the MATLAB Production Server RESTful API for MATLAB Function Execution for MATLAB function execution for request creation and protocol buffers (protobuf) for data serialization. Protocol buffers are a language-neutral and platform-neutral method of serializing structured data.

To use protobuf when making a request to the server, set the HTTP Content-Type header to application/x-google-protobuf in the client code. The Java client library provides helper classes to internally create protobuf messages based on a proto format and returns the corresponding byte array. Use this byte array in the HTTP request body. The Java client library provides methods and classes to deserialize the protobuf responses.

For examples, see Asynchronous RESTful Requests Using Protocol Buffers in the Java Client and Synchronous RESTful Requests Using Protocol Buffers in the Java Client.

Logging

You can record details such as HTTP request statuses, server URLs, and output data in your Java client application using the logging capability available in the MATLAB Production Server Java client library. To offer logging options, the Java client library, mps_client.jar, packages the SLF4J API module (version 1.7.25) as part of the client library. You can use any SLF4J-supported logging framework such as Log4j, Logback, or the java.util.logging package. Without a binding, SLF4J defaults to a no-operation implementation. For details about using a binding, see the Bridging Legacy APIs documentation on the SLF4J website.

When using a logging framework or a different version of SLF4J, you must add it in front of the mps_client.jar on the Java class path.

If you use the java.util.logging package for logging, you must load and use the java.util.logging.Logger class in your Java application code before you load the com.mathworks.mps.client.MWHttpClient class. For more information about java.util.logging, see the Oracle® Package java.util.logging Javadoc.

Related Topics

External Websites