Quality and stability improvements
R2025b delivers quality and stability improvements, building on the new features introduced in R2025a.
MATLAB Production Server Python Client: Available as PyPI Package
The MATLAB® Production Server™ client library is available as a PyPI package at matlab-prodserver-client · PyPI. For more information on installing the client, see Install the MATLAB Production Server Python Client (R2025a).
RESTful API: Securely upload and manage deployed archives on MATLAB Production Server
The RESTful API for Secure Management of Deployable Archives allows you to upload
deployable archives to MATLAB
Production Server as an alternative to copying an archive into the
auto_deploy folder. The API also supports deleting an
archive, updating an existing archive, and listing all archives deployed to the
server. The API supports access control, allowing you to restrict user access to
applications or archives deployed to a server. For more information, see RESTful API for Secure Management of Deployable Archives (R2025a).
MATLAB Production Server Reference Architecture on AWS: Support for R2025a release (April 2025)
You can now deploy MATLAB Production Server R2025a on Amazon® Web Services (AWS®) using the reference architecture on GitHub®.
To access the reference architecture, see MATLAB Production Server on Amazon Web Services on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Architecture on AWS (R2025a) in the MATLAB Production Server documentation.
MATLAB Production Server Reference Architecture on Azure: Support for R2025a release (April 2025)
You can now deploy MATLAB Production Server R2025a on Microsoft® Azure® using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Microsoft Azure on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on Azure (R2025a) in the MATLAB Production Server documentation.
Functionality being removed or changed
Persistent Storage: Redis Will No Longer Ship with MATLAB Production Server
Warns
Starting in a future release, Redis™ will no longer ship with MATLAB
Production Server and MATLAB
Compiler SDK™. Redis supports the data persistence functionality in these products.
Provide your own Redis persistence service instead. Affected functions include start (R2025a), stop (R2025a), and restart (R2025a).
MATLAB Production Server Reference Architecture on AWS: Support for R2024b release (September 2024)
You can now deploy MATLAB Production Server R2024b on Amazon Web Services (AWS) using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Amazon Web Services on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on AWS (R2024b) in the MATLAB Production Server documentation.
MATLAB Production Server Reference Architecture on Azure: Support for R2024b release (September 2024)
You can now deploy MATLAB Production Server R2024b on Microsoft Azure using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Microsoft Azure on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on Azure (R2024b) in the MATLAB Production Server documentation.
Functionality Being Removed or Changed
Azure Marketplace: MATLAB Production Server for Microsoft Azure Marketplace will no longer accept new users (April 2024)
Behavior change in future release
MATLAB Production Server has been removed from Azure marketplace and is no longer accepting new users. Current users are still able to use the marketplace offering.
For continued support, use the MATLAB Production Server reference architecture on Azure, which is hosted on GitHub. See MATLAB Production Server Reference Architecture on Azure (R2024b) for more information.
Archive-Level Routes: Specify archive-level routes using
compiler.build.productionServerArchive
Specify archive-specific routes with compiler.build.productionServerArchive (R2024a) (MATLAB Compiler SDK).To do so, set the
RoutesFile property when you create a deployable archive
using compiler.build.productionServerArchive.
Previously, archive-level routes were supported only for deployable archives
created using the mcc (R2024a) (MATLAB Compiler) command. For more information on
defining archive-level routes using either method, see Handle Custom Routes and Payloads in HTTP Requests (R2024a) (MATLAB Compiler SDK).
Configuration Properties: Give deployable archive its own pool of workers
You can give deployable archives their own pools of workers using the
component-isolation property. These deployable archives are
selected using a regular expression in ECMAScript that matches the stem of the
corresponding filename in the auto_deploy directory. For these
isolated components, no other component, whether isolated or non-isolated, will ever
use this pool of workers.
Assigning isolated pools of works can be useful for the following:
To prohibit the execution of any other component in one worker process. With this prohibition, no other component can interfere in any way with the configured component within a single process.
To allow different components to use available workers in a more controlled way by assigning to them a portion of the available workers.
For details, see component-isolation (R2024a).
MATLAB Production Server Reference Architecture on AWS: Support for R2024a release (April 2024)
You can now deploy MATLAB Production Server R2024a on Amazon Web Services (AWS) using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Amazon Web Services on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on AWS (R2024a) in the MATLAB Production Server documentation.
MATLAB Production Server Reference Architecture on Azure: Support for R2024a release (March 2024)
You can now deploy MATLAB Production Server R2024a on Microsoft Azure using the reference architecture on GitHub.
The following resources have new names:
servermachineis nowadmindashboardservermachine-nicis nowadmindashboard-nicservermachine-public-ipis nowadmindashboard-public-ip
To access the reference architecture, see MATLAB Production Server on Microsoft Azure on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on Azure (R2024a) in the MATLAB Production Server documentation.
Functionality being removed or changed
Configuration Properties: Support for SSLv2 and SSLv3 discontinued for
ssl-protocols property
Behavior change
MATLAB
Production Server no longer supports SSLv2 and SSLv3 for the
ssl-protocols property in the
main_config file. Additionally, TLSv1 and TLSv1.1 are no
longer default values for ssl-protocols.
For details, see ssl-protocols (R2024a).
Configuration Properties: Default values for ssl-ciphers
configuration property changed
Behavior change
The default value for the ssl-ciphers property in
main_config is now a specific list of cipher suites,
instead of ALL:
ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES128-GCM-SHA256Cipher suites using ECDHE (Elliptic Curve Diffie-Hellman Ephemeral) for key exchange, ECDSA or RSA for authentication, AES128 in GCM (Galois/Counter Mode) for encryption, and SHA256 for message authentication.
ECDHE-ECDSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-GCM-SHA384Cipher suites using AES256 for encryption and SHA384 for message authentication.
ECDHE-ECDSA-CHACHA20-POLY1305:ECDHE-RSA-CHACHA20-POLY1305Cipher suites using ECDHE for key exchange, ECDSA or RSA for authentication, and the ChaCha20-Poly1305 algorithm for combined encryption and message authentication.
DHE-RSA-AES128-GCM-SHA256:DHE-RSA-AES256-GCM-SHA384:DHE-RSA-CHACHA20-POLY1305Cipher suites using DHE (Diffie-Hellman Ephemeral) for key exchange, RSA for authentication, and either AES in GCM mode or ChaCha20-Poly1305 for encryption and message authentication.
For details, see ssl-ciphers (R2024a).
MATLAB Production Server No Longer Supports PAYG on Azure
As of R2024a, MATLAB Production Server's pay-as-you-go offering is no longer available for new users. For alternatives, see MATLAB Production Server Reference Architecture on Azure (R2024a).
Archive-Level Routes: Organize URL route mappings by deployable archive name
When defining URL routes that map client requests to MATLAB functions deployed as web request handlers, you can now define URL
routes in separate JSON files organized by deployable archive name. Previously, you
defined URL routes for all archives deployed on the server in a single JSON file,
which is located on the server instance at the path specified by the routes-file (R2023b) configuration property. Now, for each archive deployed
on the server, you can define a JSON file containing archive-level routes and
compile this file into the archive. Using archive-level routes, you can better
organize your routes and avoid potential route matching conflicts that can occur
when specifying all routes in the server-level routes JSON file.
Archive-level routes are supported only for deployable archives created using the
mcc (R2023b) (MATLAB Compiler) command. These routes require
both MATLAB
Production Server R2023b and MATLAB Runtime R2023b. For more details on defining archive-level routes, see Handle Custom Routes and Payloads in HTTP Requests (R2023b).
MATLAB Production Server Reference Architecture on AWS: Support for R2023b release (October 2023)
You can now deploy MATLAB Production Server R2023b on Amazon Web Services (AWS) using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Amazon Web Services on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on AWS (R2023b) in the MATLAB Production Server documentation.
MATLAB Production Server Reference Architecture on Azure: Support for R2023b release (October 2023)
You can now deploy MATLAB Production Server R2023b on Microsoft Azure using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Microsoft Azure on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on Azure (R2023b) in the MATLAB Production Server documentation.
MATLAB Production Server No Longer Supports PAYG on AWS
As of R2023b, MATLAB Production Server's pay-as-you-go offering is no longer available for new users. For alternatives, see MATLAB Production Server Reference Architecture on AWS (R2023b).
Security: Addressed known SSL security vulnerabilities, which includes disabling SSL renegotiation by default
MATLAB Production Server now addresses known SSL security vulnerabilities, which includes disabling SSL renegotiation by default. When server authentication requirements change, SSL renegotiation lets clients maintain their connection to the server by enabling them to re-enter their authentication details. However, SSL renegotiation exposes security vulnerabilities and must be used with caution.
To enable SSL renegotiation, uncomment the enable-ssl-renegotiation (R2023a) property in the
config/main_config file of your server instance.
MATLAB Production Server in Kubernetes Using Containers: Support for R2023a release (March 2023)
You can now deploy MATLAB Production Server R2023a on a Kubernetes® cluster using Helm® charts and Docker® containers.
To access the Helm charts and for instructions on loading the Docker container images used in the deployment, see MATLAB Production Server on Kubernetes (GitHub).
For additional information on deploying MATLAB Production Server in Kubernetes, see Run MATLAB Production Server on Kubernetes Using Containers (R2023a).
MATLAB Production Server Reference Architecture on AWS: Support for R2023a release (April 2023)
You can now deploy MATLAB Production Server R2023a on Amazon Web Services (AWS) using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Amazon Web Services on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on AWS (R2023a) in the MATLAB Production Server documentation.
MATLAB Production Server Reference Architecture on Azure: Support for R2023a release (April 2023)
You can now deploy MATLAB Production Server R2023a on Microsoft Azure using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Microsoft Azure on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on Azure (R2023a) in the MATLAB Production Server documentation.
Functionality being removed or changed
MATLAB Production Server Reference Architecture on AWS: Support for R2020a discontinued (April 2023)
Errors
The reference architecture template for deploying MATLAB Production Server on Amazon Web Services (AWS) no longer supports the R2020a release of MATLAB Production Server.
For continued support of your applications, use a supported release of MATLAB Production Server on AWS. For a list of currently supported releases, see MATLAB Production Server on Amazon Web Services on GitHub.
Web Request Handlers: Test MATLAB function and URL routes using the testing interface in the Production Server Compiler app
You can use the testing interface in the Production Server Compiler app to test web request handlers. Configure access to the routes file, then start the test server in the app to test MATLAB functions for deployment to MATLAB Production Server. For more information, see Test Web Request Handlers (R2022b) (MATLAB Compiler SDK).
Kerberos: Use Kerberos to authenticate user access to Linux server instances
You can configure Kerberos to authenticate user access to MATLAB Production Server instances running on Linux®, in addition to existing support for Windows®. For more information, see Use Kerberos Authentication and Kerberos Delegation (R2022b).
Data Caching: Using Redis on Windows requires WSL installation
MATLAB Production Server ships a newer version of Redis. To use Redis on server instances that run on Windows machines, an installation of Windows Subsystem for Linux (WSL) is required. For details, see Install WSL for Server Instances Running on Windows (R2022b).
Previous versions of MATLAB Production Server do not require a WSL installation to use Redis.
Python Client: Support for Python 3.10
The Python® client library now supports Python 3.10, in addition to existing support for versions 3.8 and 3.9. For more information on client system requirements, see Product Requirements and Platform Availability for MATLAB Production Server.
Streaming Data Framework for MATLAB Production Server: Interact with Kafka event streams and develop streaming analytic functions
Use Streaming Data Framework for MATLAB Production Server to read from and write to event streaming platforms, such as Kafka®. Using this framework, you can:
Develop a streaming analytic function in MATLAB that filters, transforms, records or processes event stream data.
Connect to a streaming source and test how the analytic function reads from and writes to event streams by using Streaming Data Framework for MATLAB Production Server functions.
Simulate the production environment for testing your streaming analytic algorithms (requires MATLAB Compiler SDK).
Package the analytic function and deploy it to your production environment by using MATLAB Production Server.
For more information on this framework, see Event Stream Processing (R2022b).
MATLAB Production Server Reference Architecture on Azure: Support for R2022b release (November 2022)
You can now deploy MATLAB Production Server R2022b on Microsoft Azure using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Microsoft Azure on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on Azure (R2022b) in the MATLAB Production Server documentation.
MATLAB Production Server Reference Architecture on AWS: Support for R2022b release (December 2022)
You can now deploy MATLAB Production Server R2022b on Amazon Web Services (AWS) using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Amazon Web Services on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on AWS (R2022b) in the MATLAB Production Server documentation.
Functionality being removed or changed
Python Client: Support for Python 2.7 discontinued
Errors
The Python client library no longer supports Python 2.7. For more information on client system requirements, see Product Requirements and Platform Availability for MATLAB Production Server.
To ensure continued support for your applications, use a supported version of Python, version 3.8, 3.9, or 3.10.
MATLAB Production Server (PAYG) on Azure: Support for R2020b, R2021a, and R2021b discontinued (September 2022)
Errors
Azure Marketplace no longer supports the R2020b, R2021a, and R2021b releases of MATLAB Production Server.
Web Request Handlers: Support for custom URL routes and request payloads
MATLAB Production Server now supports flexible mapping of request URLs to deployed MATLAB functions and can serve static content. Clients can send RESTful requests to the server that contain custom HTTP headers and custom payloads.
To enable the server to access the request headers and payload, and to return custom headers, status codes, and payload in the response, you configure the server and write the deployed MATLAB function to handle them specifically. For details, see Handle Custom Routes and Payloads in HTTP Requests (R2022a).
RESTful API for Server Metrics: Provide custom metrics
Create custom Prometheus® metrics related to your algorithm or request processing by using the
functions prodserver.metrics.setGauge (R2022a) (MATLAB Compiler SDK) and prodserver.metrics.incrementCounter (R2022a) (MATLAB Compiler SDK) in the deployed MATLAB function. These functions create Prometheus gauge and counter metrics, respectively. The MATLAB
Production Server metrics service returns the custom metrics in its output. For details,
see Create Custom Prometheus Metrics (R2022a).
Java Client: mps_client.jar client library hosted in
Maven™ repository
The mps_client.jar
Java® client library is hosted in a Maven™ repository at https://mvnrepository.com/artifact/com.mathworks.prodserver/mps_java_client. To use the JAR file 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>R2022a</version>
</dependency>
Python Client: Support for HTTPS
Using the MATLAB Production Server Python client library, you can now write applications that use HTTPS to execute MATLAB functions deployed to server instances. For details, see Use HTTPS for Client-Server Communication (R2022a).
HTTPS communication using the Python client API is supported only on Windows and Linux platforms. macOS is not supported.
Python Client: Improved performance with large multidimensional arrays
The Python multidimensional array component used by the MATLAB Production Server Python client library shows improved performance in both of the following cases:
Converting data from Python sequences to the data types defined by the
matlabmoduleDeserializing data from MATLAB into Python
For both cases, the improvement is noticeable when operating on arrays with at least 10 elements. For example, in the following Python code, the first line in bold is about 11 times faster than in the previous release, and the second line in bold is about 12 times faster:
import random
import time
from matlab.production_server._internal_mps.matlab_array_builder import create_matlab_array
from matlab.production_server._internal_mps.matlab_response_builder import create_matlab_result
import matlab
exp = 8
rand_array = [random.random() for i in range(10**exp)]
s0 = time.perf_counter()
array_md = matlab.double(rand_array, size=(1, 10**exp))
s1 = time.perf_counter() - s0
print('conversion to matlab.double: {:< 10f} seconds'.format(s1))
matlab_array = create_matlab_array(array_md)
s0 = time.perf_counter()
matlab_result = create_matlab_result(matlab_array.__get_ml_ref__())
s1 = time.perf_counter() - s0
print('execution of create_matlab_result(): {:< 10f} seconds'.format(s1))The approximate execution times for the first line in bold are:
R2021b: 40 s
R2022a: 3.7 s
The approximate execution times for the second line in bold are:
R2021b: 43 s
R2022a: 3.5 s
The code was timed on a Windows 10, Intel®
Xeon® CPU E5-1650 v4 @ 3.60 GHz test system using Python
perf_counter() statements.
Microservice Integration: Create a microservice Docker image to access MATLAB code
You can create microservice Docker images on Linux operating systems from the MATLAB command prompt using the compiler.package.microserviceDockerImage (R2022a) (MATLAB Compiler SDK) function. This function,
along with the compiler.package.MicroserviceDockerImageOptions (R2022a) (MATLAB Compiler SDK) function, provides
an interface to specify various options associated with creating Docker images. The microservice image provides an HTTP or HTTPS endpoint to
access MATLAB code.
Python Client: Support for Python 2.7 will be discontinued
The Python client library will not support Python 2.7 in future releases. For more information on client system requirements, see Product Requirements and Platform Availability for MATLAB Production Server.
If you want to use Python 2.7 to develop client applications, you can continue using the R2022a version of the Python client library in future releases.
Python Client: Support for Python 3.7 discontinued
Errors
The Python client library no longer supports Python 3.7. For more information on client system requirements, see Product Requirements and Platform Availability for MATLAB Production Server.
To ensure continued support for your applications, use a supported version of Python, version 2.7, 3.8, or 3.9.
Java Client: Log4j and Logback no longer shipped in the client library
Errors
The Java client library no longer ships the Log4j and Logback logging frameworks.
To support logging in your applications, install an SLF4J-supported logging framework, then update your Java class path to use the framework. For details, see Logging (R2022a).
MATLAB Production Server in Kubernetes: Support for R2022a release (April 2022)
You can now deploy MATLAB Production Server R2022a in Kubernetes using the reference architecture on GitHub. For deployment details, see the MATLAB Production Server in Kubernetes repository on GitHub.
MATLAB Production Server on Azure: Support for R2022a release (June 2022)
You can now deploy MATLAB Production Server R2022a using the reference architecture on GitHub, and the BYOL and PAYG offerings on the Azure marketplace.
MATLAB Production Server Reference Architecture on AWS: Support for R2022a release (June 2022)
You can now deploy MATLAB Production Server R2022a on Amazon Web Services (AWS) using the reference architecture on GitHub.
To access the reference architecture, see MATLAB Production Server on Amazon Web Services on GitHub.
For information on deploying MATLAB Production Server using this reference architecture, see MATLAB Production Server Reference Architecture on AWS (R2022a) in the MATLAB Production Server documentation.
MATLAB Production Server (PAYG) on AWS: Support for R2022a release (October 2022)
You can now deploy MATLAB Production Server R2022a on Amazon Web Services (AWS) using the pay-as-you-go (PAYG) offering on AWS Marketplace.
To access the PAYG offering, see MATLAB Production Server (PAYG) on AWS Marketplace.
For information on deploying MATLAB Production Server using this offering, see MATLAB Production Server (PAYG) on AWS (R2022a) in the MATLAB Production Server documentation.
RESTful API for Server Metrics: Retrieve server metrics
Use the GET Metrics (R2021b) API to retrieve server metrics in Prometheus metrics format. The metrics service returns information about requests that client applications send to the server, and the time and memory that the server utilizes to execute the requests. You can use the metrics to monitor the server when working with Kubernetes and microservices. For more information, see Metrics Service (R2021b).
Japanese Language Localization Support: Parameters, variables, and error messages are available in Japanese
MATLAB Production Server Dashboard is available in Japanese for Japanese localized systems.
Python Client: Support for Python 3.9
The Python client library now supports Python 3.9. For more information on client system requirements, see Product Requirements and Platform Availability for MATLAB Production Server.
MATLAB Production Server (PAYG) on AWS: Deploy MATLAB Production Server on AWS without a license and pay for the MATLAB Production Server workers that you use (August 2021)
Pay as you go for the AWS resources and the MATLAB Production Server workers that your MATLAB Production Server deployment uses. MATLAB Production Server (PAYG) uses a CloudFormation template that enables you to quickly create, configure, and deploy a server environment in AWS. You do not need a MATLAB Production Server license to use this offering. This offering is useful for organizations that want to run proofs-of-concept, require extended paid trials, and experience bursty workloads. To deploy from the AWS marketplace, see MATLAB Production Server (PAYG). For information on how to deploy, see AWS Deployment for MATLAB Production Server (PAYG) (R2021b).
MATLAB Production Server Reference Architecture on AWS: New dashboard and support for generic identity providers; mandatory SSL certificate requirement during deployment (November 2021)
The dashboard has replaced the cloud console for editing the server configuration and uploading applications. The dashboard also allows role-based access control, where administrators can grant a user the privileges to perform tasks on the dashboard based on their role. For information on using the dashboard, see Manage MATLAB Production Server Using the Dashboard (R2021b). For information on configuring role-based access control, see Dashboard Access Control (R2021b).
The MATLAB Production Server reference architecture now supports integration with OAuth 2.0 providers such as Google® Identity and PingFederate®, in addition to the preexisting integration with Azure Active Directory, to restrict user access to applications deployed to the server and for configuring role-based access control for the dashboard. For details, see Application Access Control (R2021b).
The MATLAB Production Server deployment procedure requires a preexisting SSL certificate in your AWS account to use during deployment. Providing an SSL certificate is no longer optional.
MATLAB Production Server Reference Architecture, BYOL, and PAYG on Azure: Support for R2021b release (November 2021)
You can now deploy MATLAB Production Server R2021b using the reference architecture on GitHub, and the BYOL and PAYG offerings on the Azure marketplace.
MATLAB Production Server in Kubernetes: Deploy and run MATLAB Production Server in Kubernetes using a Helm chart (February 2022)
You can now deploy MATLAB Production Server in a Kubernetes cluster using a Helm® chart provided by MathWorks®. The chart is a collection of YAML files that define the resources you need to deploy MATLAB Production Server in Kubernetes. For deployment details, see the MATLAB Production Server in Kubernetes repository on GitHub.
Application Access Control: Support for generic identity providers
Still runs
MATLAB Production Server now integrates with OAuth 2.0 providers such as Google and PingFederate, in addition to the preexisting integration with Azure Active Directory, to restrict user access to applications deployed to the server. For more information, see Application Access Control (R2021a).
Use OAuth2 instead of AzureAD when
specifying the identity provider for the access-control-provider (R2021a) property. Specifying
AzureAD will not be supported in a future release.
The default name for the access control configuration file is
jwt_idp.json instead of azure_ad.json,
if the access-control-provider property is set to OAuth2. For
more information, see access-control-config (R2021a).
Python Client: Support for Python 3.6 discontinued
Errors
The Python client library no longer supports Python 3.6. For more information on client system requirements, see Product Requirements and Platform Availability for MATLAB Production Server.
To ensure continued support for your applications, use a supported version of Python, version 2.7, 3.7, or 3.8.
MATLAB Client for MATLAB Production Server: Support for HTTPS (November 2020; Version 20.2.1)
You can now write client applications using MATLAB Client for MATLAB Production Server that use HTTPS to execute MATLAB functions deployed to server instances. For more information, see Execute Deployed Functions Using HTTPS (R2021a).
MATLAB Client for MATLAB Production Server: Support for access control (January 2021; Version 20.2.2)
You can now write client applications using MATLAB Client for MATLAB Production Server that support access control. Client applications can send a bearer token to server instances that require user authorization for executing deployed applications. For more information, see Application Access Control (R2021a).
MATLAB Production Server on Azure: Support for generic identity providers (July 2021)
MATLAB Production Server (BYOL), MATLAB Production Server (PAYG), and MATLAB Production Server reference architecture on Azure now support integration with OAuth 2.0 providers such as Google Identity and PingFederate, in addition to the preexisting integration with Azure Active Directory, to restrict user access to applications deployed to the server and for configuring role-based access control for the dashboard.
MATLAB Production Server on Azure: Provide SSL certificate during deployment (July 2021)
The deployment templates for MATLAB Production Server (BYOL), MATLAB Production Server (PAYG), and MATLAB Production Server reference architecture on Azure are no longer preconfigured to use a self-signed SSL certificate. You must provide an SSL certificate during deployment.