Debugging SDK Processor Remotely

Preparing Traffic manager container

Note: You can debug remotely only in the quick start mode.
  1. Set up the cluster and connect to the tml-tm container. 
     docker ps | grep -i tml-tm
    docker exec -it {tml-tm-container_id} bash
  2. Enable Javaproxy with debugger agent by executing the following command: 
    Edit the file /opt/javaproxy/proxy/proxy.ini and add below line at the end
-agentlib:jdwp=transport=dt_socket,address=8001,server=y,suspend=n
    Save the changes.
  3. Restart the proxy as shown below. 
    /etc/init.d/javaproxy restart

Preparing the IDE

Eclipse

  1. Open IDE and configure the debugger to connect to a remote port.
  2. Click Debug Configurations.



  3. The Debug Configurations window opens, here click Remote Java Application.

  4. Click New button,

    , to create a new configuration. The right side of the screen displays options to create new configuration.
  5. In the Connect tab, select the Project by clicking Browse. In the Project selection window, select the SDK Processor's project.



  6. In the Connection Properties field, for the Host field, input the node's IP and for Port, input 8001. Click Apply and then click Debug.

IntelliJ Idea

  1. From the main menu, click Run option and select Edit Configuration.
  2. Click button and select Remote from the configuration list.
  3. In the tree structure, click SDK Debug. In the Debugger mode field, select Attach to remote JVM. For the Host field, input the node's IP and for Port, input 8001.



On successful completion, you can see the toolbar in the IntelliJ Idea window. Click the green debug icon,when you wish to debug.

Apache NetBeans
  1. In the NetBeans window, click Debug menu, and then click Attach Debugger.



  2. The Attach dialog box opens, here provide the input for Debugger, connector and Transport.



  3. For the Host field, input the node's IP and for Port, input 8001. Click OK.
The debug view opens. Set the required breakpoints in the processor code and call an end point.

Fix, Build and Deploy

Fix

Fix the issues found on debugging by fixing the code.

Build
  1. Run the built script. 
    /home/user/MasheryLocalSDK$ ./build-adapter.sh
    Note: The build script reports the compilation errors. In case of any reported errors rerun the build script.

  2. The script generates an archive file as tml-mashery-customer-extension.zip under MasheryLocalSDK. Upload this file to the Mashery Local installer.

Deploy
  1. Find the container ID of the tml-tm container using the following command: 
    docker ps | grep -i "tml-installer"
  2. Copy the upload-deploy-sdk-processor.sh script from the installer to the local path. 
    docker{tml-installer-contianer-id}:/var/jenkins_home/docker-deploy/onprem/quick_start/system/upload-deploy-sdk-processor.sh {path_to_save}
  3. Use the upload-deploy-sdk-processor.sh script to upload and deploy the already built SDK processors in the tml-tm container. 
    ./upload-deploy-sdk-processor.sh --help
Usage: ./upload_deploy_sdk_processor.sh [--adapter-zip <ADAPTER_FULL_PATH>] [--tml-tm-container <TM_CONTAINER>]
    ( --adapter-zip | -a ) -  path/filename of the adapter zip to be uploaded to container
    ( --tml-tm-container| -c ) - TML-TM Container ID where adapter to be uploaded
Example: ./upload-deploy-sdk-processor.sh --adapter-zip

/home/user/MasheryLocalSDK/dist/tml-mashery-customer-extension.zip --tml-tm-container 9c9e09ea1f95
  4. Log in the tml-tm container, 
    docker exec -it {tml-tm-container-id} bash
    and restart thejavaproxy service using the following command: 
    /etc/init.d/javaproxy restart