DevOps Automation

DevOps Automation

Join this online group to communicate across IBM product users and experts by sharing advice and best practices with peers and staying up to date regarding product enhancements.


#DevOps
 View Only

Introducing DevOps Deploy 8.2.0 which includes a new AI Model Context Protocol (MCP) Server, a new Mainframe Linked Version feature, enhancements to the Deployment Genie, support for RHEL 10, and much more!

By Randall Langehennig posted an hour ago

  

Overview

We are very pleased to announce the release of IBM DevOps Deploy 8.2.0 (2025.12).

IBM DevOps Deploy is an application-release solution that infuses automation into the continuous delivery and continuous deployment (CI/CD) process and provides robust visibility, traceability and auditing capabilities.

Our latest release includes our first AI based Model Context Protocol (MCP) Server with tools that can be accessed through chat prompts in your IDE which acts as an AI agent.   The AI agent interprets your chat prompts to intelligently select and run the appropriate MCP tool to fulfill the request.

The following MCP tools are available from your MCP client:  

Tools

Description

Input Requirement

list_applications

Lists all applications.

None

list_application_processes

Lists all processes defined for an application.

application ID or application name

list_application_environments

Lists all environments defined in an application.

application ID or application name

list_application_snapshots

Lists all snapshots defined in an application.

application ID or application name

get_deployment_status

Returns the application deployment status.

deployment_request_id (optional)

deploy_snapshot_to_environment

Deploys an application snapshot to an environment.

application_id, snapshot_id, process_id, environment_id

Your IDE can show you the tools available for the MCP server as shown from this VS Code example:

A screenshot of a computer

AI-generated content may be incorrect.

The MCP Agent in your IDE can then query the tools above and return information from the DevOps Deploy Server.    Here are a couple of examples from my VS Code IDE running at the 1.106.3 release level with GitHub Copilot Chat version 0.33.1 extension installed:

A screenshot of a computer

AI-generated content may be incorrect.

Here is one more command where I query the target environments for one of our applications in DevOps Deploy called AceCapital-Banking:

A screenshot of a computer

AI-generated content may be incorrect.

The idea is to bring automation that significantly reduces manual work, simplifies integration across different systems, and enable powerful automation for application deployments.   For example, I could say “Deploy snapshot Dec25-01 to the QA environment for the application Ace-Capital-Banking” and the MCP tools will do all the work bringing power automation and resulting in quicker and easier deployment cycles.

We have also introduced other exciting, new enhancements in DevOps Deploy 8.2.0.  In the next section, we review each new feature available to our customers.

Enhancements in DevOps Deploy 8.2.0

Version 8.2.0 includes the following enhancements:

Feature Title

Description

Added Ollama and WatsonX support for the AI-based Deployment Genie

You can now perform an AI analysis on failed deployment steps using the Deployment Genie button with integrations to Ollama or WatsonX as your LLM of choice.   When you click the button, the deployment log is sent to your configured AI service for analysis.

With the help of the AI service, you can receive instructions to help identify the cause of the failure as well as guidance on how to resolve the issue.

See Configuring the AI settings for more information.

New Deploy Model Context Protocol (MCP) Server

The MCP Server allows AI agents to interact with the Deploy server and perform various deployment actions using chat prompts.  This feature significantly reduces manual work, simplifies integration across different systems, and enables powerful automation for application deployments resulting in quicker and easier deployment cycles.

You will configure a mcp.json file in your git repository under .vscode that tells your AI Agent how to start your MCP Server.   The URL is "https://your-deploy-server-url:port/mcp".   Once started, you should see output that looks like this where it is in a Running state and it discovers the tools available for the MCP Server.

See Deploy MCP server for more information.

Added authentication support for proxy-enabled, webhook integrations with MS Teams

Added authentication support for the proxy-enabled webhook integration for MS Teams notifications.

See Adding webhooks to process for more details on how to configure the proxy settings.

Added support for RedHat Enterprise Linux 10

We have added support for RHEL 10 for the DevOps Deploy Server, Deploy Agent, and Agent Relay in the 8.2.0 release.

See the Software Product Compatibility Reports to learn more.

Enhanced environment inventory status

A new environment inventory status assists you in diagnosing inventory differences caused by a failed component uninstall process.

See Environment inventory and compliance for more details.

Added the Description parameter to the Run Application Process window

An optional Description parameter is added to the Run Application Process window to populate the description of the process while reviewing the deployment process request.

See Deploying applications to learn more.

Added support for Java 25

MariaDB Enterprise 10.5 is no longer a supported database version

This release supports Java 25.  

Note: the minimum Java that can be leveraged is Java 17 for the upgrade or installation of DevOps Deploy 8.2.0.   Support for Java 11 has been removed.   You must upgrade to Java 17 or 21.

See Prerequisite software.

API updates

The following API updates were introduced with DevOps Deploy 8.2.0:

-       The getModels endpoint is changed to getLLMVendorModels.  See getLLMVendorModels for the details of how to get the list of supported models for the given vendor with the udclient command.

Note: the endpoint was previously accessed via a GET request but now requires a POST request with a JSON payload.

-       The getVendors endpoint is changed to getLLMVendors.   See getLLMVendors for more information on how to get the list of supported LLM vendors using the udclient command.

Added security updates

Updated Apache Tomcat to 9.0.111

Container updates for Kubernetes installation of the Deploy Server

For those looking to make the journey to containerization where the DevOps Deploy Server is installed in a OpenShift or Kubernetes cluster, the following updates were made to support our container deployments:

  •             Allowed annotations and customization in Helm charts for Kubernetes deployments.
  •             Added Helm annotation so that the appdata PVC is not removed when running the ‘helm delete …’ command.
  •              Upgraded the base image to the latest RHEL 10.x version.
  •              Updated container images to Java 21.

See Journey to Containerization with DevOps Deploy to learn more about how to move from an on-premise installation of DevOps Deploy to a Kubernetes based installation with containers.

Mainframe Enhancements

With the DevOps Deploy 8.2.0 release, we also provided the following new enhancements to help our zOS customers:

Feature Title

Description

Enhanced z/OS component version feature

You can now link z/OS component versions in IBM DevOps Deploy.   You can deploy an application as a whole, allowing environments to reliably represent a specific version of the application, while maintaining the efficiency of incremental deployments.

This simplifies the deployment of mainframe applications as well as rollback.   We believe this is a game changer for our mainframe clients and we are very excited to share this new feature with our mainframe clients.

See Linked versions in z/OS components for more details.

Added ability to download manifest of artifacts of a Linked z/OS component version

Added a button so you can download a manifest of artifacts of a Linked z/OS component version.

See Downloading component version artifacts for more details.

Want to learn more?

As you can see from all the enhancements above, IBM is investing heavily in our IBM DevOps Deploy solution.   The new features, capabilities, and continued investments in technologies like Model Context Protocol, Mainframe, AI, and more make the solution even more valuable to our many customers across the globe.    Our goal is to ensure this solution remains the most trusted application release automation solution for the enterprise.

If you are interested in learning more about DevOps Deploy, you can request a meeting at:  https://ibm.biz/DevOpsMeeting

We host DevOps Summits and Bootcamps in cities around the globe, offering deeper insights into the future of software delivery.   Join our mailing list to stay informed about upcoming events:  https://ibm.biz/DevOpsEventsList

To see further details of What’s New in DevOps Deploy 8.2.0, please see the online documentation.

0 comments
4 views

Permalink