Procedure

1. Create a credentials secret named kai-api-keys in the openshift-mta project.

   1. For Amazon Bedrock as the provider, type:

      oc create secret generic aws-credentials \
       --from-literal=AWS_ACCESS_KEY_ID=<YOUR_AWS_ACCESS_KEY_ID> \
       --from-literal=AWS_SECRET_ACCESS_KEY=<YOUR_AWS_SECRET_ACCESS_KEY>

   2. For Azure OpenAI as the provider, type:

      oc create secret generic kai-api-keys -n openshift-mta \
       --from-literal=AZURE_OPENAI_API_KEY='<YOUR_AZURE_OPENAI_API_KEY>'

   3. For Google as the provider, type:

      oc create secret generic kai-api-keys -n openshift-mta \
       --from-literal=GEMINI_API_KEY='<YOUR_GOOGLE_API_KEY>'

   4. For the OpenAI-compatible providers, type:

      oc create secret generic kai-api-keys -n openshift-mta \
       --from-literal=OPENAI_API_BASE='https://example.openai.com/v1' \
       --from-literal=OPENAI_API_KEY='<YOUR_OPENAI_KEY>'

      Note

      You can also set the base URL as the kai_llm_baseurl variable in the Tackle custom resource.

2. (Optional) Force a reconcile so that the MTA operator picks up the secret immediately

   kubectl patch tackle tackle -n openshift-mta --type=merge -p \
   '{"metadata":{"annotations":{"konveyor.io/force-reconcile":"'"$(date +%s)"'"}}}'

Procedure

1. Log in to the Red Hat OpenShift cluster and switch to the openshift-mta project.

2. Create the Tackle CR in the tackle_hub.yaml file:

   vi tackle

3. Enable kai_llm_proxy_enabled in the Tackle CR:

   ---
   kind: Tackle
   apiVersion: tackle.konveyor.io/v1alpha1
   metadata:
     name: mta
     namespace: openshift-mta
   spec:
     kai_llm_proxy_enabled: true
     kai_solution_server_enabled: true
     kai_llm_provider: <provider-name> #For example, OpenAI.
     # optional, pick a suitable model for your provider
     kai_llm_model: <model-name>
   ---

4. Save the Tackle CR configuration.

5. Apply the Tackle CR in the openshift-mta project:

    $ oc apply -f tackle_hub.yaml

Procedure

1. Open the my-Java project in Visual Studio Code.
2. Download the Red Hat Developer Lightspeed for migration toolkit for applications extension from the Visual Studio Code marketplace.
3. Type Ctrl+Shift+P (Windows and Linux systems) or Cmd+Shift+P (Mac systems).
4. Type Preferences: Open Settings (UI) in the Command Palette to open the Visual Studio Code settings and select Extensions > MTA.
5. Select Gen AI:Agent Mode.
6. In the Red Hat Developer Lightspeed for MTA extension, click Open Analysis View.

7. Type MTA: Manage Analysis Profile in the Command Palette and configure the following fields:

   • Profile Name: Type a profile name
   • Target Technologies: quarkus
   • Custom Rules: Select custom rules if you want to include them while running the analysis. By default, Red Hat Developer Lightspeed for MTA enables Use Default Rules for quarkus.

8. Type MTA: Open the Gen AI model provider configuration file in the Command Palette to configure the following in the provider-settings file:

   models:
     openshift-example-model: &active
       environment:
         OPENAI_API_KEY: "<Server's OPENAI_API_KEY>"
         CA_BUNDLE: "<Servers CA Bundle path>"
       provider: "ChatOpenAI"
       args:
         model: "my-model"
         configuration:
           baseURL: "https://<serving-name>-<data-science-project-name>.apps.konveyor-ai.example.com/v1"

   Note

   You must change the provider-setting configuration if you plan to use a different LLM provider.

9. On the MTA: Open Analysis View page, click Start to start the MTA Remote Procedure Call (RPC) server.

10. Select the profile you configured and click Run Analysis to scan the Java application.

    MTA identifies the issues in the code.

11. Click the solutions icon ( ) in an issue to request suggestions to resolve the issue.

    Red Hat Developer Lightspeed for MTA streams the issue description, a preview of the code changes in files that resolve the issue.

    You can review the code changes in the editor and accept or reject the changes. If you accept the changes, Red Hat Developer Lightspeed for MTA creates a new file with the accepted code changes.

12. Click Continue to allow Red Hat Developer Lightspeed for MTA to run a follow-up analysis.

    This round of analysis detects lint issues, compilation issues, or diagnostic issues that can occur when you accepted the suggested code change.

    Repeat the review and accept or reject the resolutions. Red Hat Developer Lightspeed for MTA continues to run repeated iterations of scan if you allow until it resolves all issues.

Procedure

1. Go to the Podman AI Lab extension and click Catalog under Models.
2. Download one or more models.
3. Go to Services and click New Model Service.
4. Select a model that you downloaded in the Model drop down menu and click Create Service.
5. Click the deployed model service to open the Service Details page.
6. Note the server URL and the model name. You must configure these specifications in the Red Hat Developer Lightspeed for MTA extension.

7. Export the inference server URL as follows:

   export OPENAI_API_BASE=<server-url>

8. In the Red Hat Developer Lightspeed for MTA extension, type Open the GenAI model provider configuration file in the Command Palette to open the provider-settings.yaml file.

9. Enter the model details from Podman Desktop. For example, use the following configuration for a Mistral model.

   podman_mistral: &active
       provider: "ChatOpenAI"
       environment:
         OPENAI_API_KEY: "unused value"
       args:
        model: "ibm-granite/granite-3.3-8b-instruct-GGUF"
        configuration:
          baseURL: "http://localhost:56885/v1"

   Note

   The Podman Desktop service endpoint does not need a password but the OpenAI library expects the OPENAI_API_KEY to be set. In this case, the value of the OPENAI_API_KEY variable does not matter.

Procedure

1. Go to the Red Hat Developer Lightspeed for MTA settings in one of the following ways:

   1. Click Extensions > Migration toolkit for applications Extension > Settings
   2. Type Ctrl+Shift+P or Cmd+Shift+P on the search bar to open the Command Palette and enter Preferences: Open Settings (UI). Go to Extensions > MTA to open the settings page.

2. Configure the settings described in the following table:

   Table 4.1. Red Hat Developer Lightspeed for MTA extension settings

   Settings	Description
   Log level	Set the log level for the MTA binary. The default log level is debug. The log level increases or decreases the verbosity of logs.
   Analyzer path	Specify an MTA custom binary path. If you do not give a path, Red Hat Developer Lightspeed for MTA uses the default path to the binary.
   Auto Accept on Save	Red Hat Developer Lightspeed for MTA enables this option by default. When you accept the changes that the LLM suggests, Red Hat Developer Lightspeed for MTA automatically saves the updated code in a new file. Disable this option if you want to manually save the new file after accepting the code changes.
   Gen AI:Enabled	Red Hat Developer Lightspeed for MTA enables this option by default. You can get code fixes by using Red Hat Developer Lightspeed for MTA with a large language model.
   Gen AI: Agent mode	Enable the experimental Agentic AI flow for analysis. Red Hat Developer Lightspeed for MTA runs an automated analysis of a file to identify issues and suggest resolutions. After you accept the solutions, Red Hat Developer Lightspeed for MTA makes the changes in the code and re-analyzes the file.
   Gen AI: Excluded diagnostic sources	Add diagnostic sources in the settings.json file. Red Hat Developer Lightspeed for MTA excludes the issues that such diagnostic sources generate from the automated Agentic AI analysis.
   Cache directory	Specify the path to a directory in your filesystem to store cached responses from the LLM.
   Trace directory	Configure the absolute path to the directory that has the saved LLM interaction.
   Trace enabled	Enable to trace MTA communication with the LLM model. Red Hat Developer Lightspeed for MTA stores traces in the trace directory that you configured.
   Demo mode	Enable to run Red Hat Developer Lightspeed for MTA in demo mode that uses the LLM responses from the cache directory for analysis.
   Debug:Webview	Enable debug level logging for Webview message handling in Visual Studio Code.

Procedure

1. Open the MTA View Analysis page in either of the following ways:

   1. Click the book icon on the MTA: Issues pane of the MTA extension.
   2. Type Ctrl + Shift + P or Cmd + Shift + P on the search bar to open the Command Palette and enter MTA:Open Analysis View.

2. Click the settings button on the MTA View Analysis page to configure a profile for your project.

   The Get Ready to Analyze pane lists the profile configuration options.

   Table 4.2. Red Hat Developer Lightspeed for MTA profile settings

   Profile settings	Description
   Select profile	Create a profile. The profile name is part of the context provided to the LLM for analysis.
   Configure label selector	A label selector filters rules for analysis based on the source or target technology. Specify one or more target or source technologies (for example, cloud-readiness). Red Hat Developer Lightspeed for MTA uses this configuration to determine the rules to apply to a project during analysis. If you mentioned a new target or a source technology in your custom rule, you can type that name to create and add the new item to the list. Note You must configure either target or source technologies for running an analysis.
   Set rules	Enable default rules and select your custom rule that you want MTA to use for an analysis. You can use the custom rules in addition to the default rules.
   Configure generative AI	This option opens the provider-settings.yaml file that has API keys and other parameters for all supported LLMs. By default, Red Hat Developer Lightspeed for MTA uses OpenAI LLM. To change the model, update the anchor &active to the desired block. Modify this file with the required arguments, such as the model and API key, to complete the setup.

1. Open MTA Analysis View.
2. Click Run Analysis to verify that your configuration works by running an analysis.

Procedure

1. Review the issues from the Analysis results space of the MTA view analysis page by the following tabs:

   1. All: lists all incidents identified in your project.
   2. Files: lists all the files in your project for which the analysis identified issues that you must resolve.
   3. Issues: lists all issues across different files in your project.
2. Use the Category drop down to filter issues based on how crucial the fix is for the target migration. You can filter mandatory, potential, and optional issues.
3. Click Has Success Rate to check how many times you accepted the same issue resolution in an earlier analysis.
4. Click the solution tool to trigger automated updates to your code. If you applied any category filter, Red Hat Developer Lightspeed for MTA makes code updates for all incidents, specific files, or specific issues based on the filter. Red Hat Developer Lightspeed for MTA generates new files with the updated code.
5. Review and (optionally) edit the code.
6. Click Apply all in the Resolutions pane to permanently apply the changes to your code.

Procedure

1. Enable the agent mode in one of the following ways:

   1. Open the MTA Analysis View page.

   2. Enable the Agent Mode on.

      OR

   3. Go to Extensions > MTA-core > GenAI: Agent Mode
   4. Select the Agent Mode checkbox.
2. Click the Red Hat Developer Lightspeed for MTA extension and click Open MTA Analysis View.
3. Select a profile for the analysis.
4. Click Start to start the MTA Remote Procedure Call (RPC) server.
5. Click Run Analysis on the MTA Analysis View page. The Resolution Details tab opens, where you can view the automated analysis that makes changes in applicable files.
6. Click the Review Changes option to open the editor that shows the diff view of the modified file.
7. Review the changes and click Apply to update the file with all the changes or Reject to reject all changes. If you applied the changes, then Red Hat Developer Lightspeed for MTA creates the updated file with code changes.
8. Open Source Control to access the updated file.
9. In the Resolution Details view, accept the proposal from Red Hat Developer Lightspeed for MTA to make further changes. The stream of analysis repeats, after which you can review and accept the changes. Red Hat Developer Lightspeed for MTA creates the file with the code changes, and the stream continues until you reject the proposal for further analysis.