Install the Data Studio Interpreters

The Data Studio interpreters are part of the Oracle Healthcare Translational Research Notebook solution. This installation assumes that the docker containers for the interpreters are to be running in the same host VM as the Oracle Healthcare Translational Research Middle Tier. The installation and configurations are easier if they are run on the same VM. However, if running on a different host, see the advanced section on "Running on a different Host".

1. Download the Part 1 and Part 2 zip files of the Data Studio interpreters onto the host VM.
   • PgXDataStudio-Docker_3.4_part1.zip
   • PgXDataStudio-Docker_3.4_part2.zip
2. Create a directory such as datastudio.
3. Unzip the zip files onto the directory:
   • unzip -a /scratch/ohtr/stage/PgXDataStudio-Docker_3.4_part1.zip
   • unzip -a /scratch/ohtr/stage/PgXDataStudio-Docker_3.4_part2.zip

   Note:

   The subfolder docker-images contains the tar file of the docker images. The PgXDataStudio contains the configuration and set up files.
4. Copy the docker-load.sh file from the folder PgXDataStudio/docker onto the docker-images folder.
5. Run the docker-load.sh file (sudo sh ./docker_load.sh).

   Note:

   The docker images is loaded from the tar files into the local docker registry. This will take a few minutes and the progress is visible in the console.
6. Create the Oracle wallet for storing the Oracle Healthcare Translational Research database api schema credentials for python interpreter as follows:
   1. Create a directory "shared_volume" inside the PgxDataStudio directory (or any other location).
   2. Change directory into the PgxDataStudio/database/utilities folder. Edit the cx_oracle_conn_config.sh file and provide the values specified in the file.

      Provide the database hostname, port and service name.

      Provide the API (CGA) schema name for API_SCHEMA_NAME

      Provide Weblogic Home directory full path for WEBLOGIC_HOME

      Provide the Java Home directory full path for JAVA_HOME

      Provide the full path to the shared_volume directory created above for HOST_TNS_ADMIN

   3. Run the script create_cx_oracle_conn.sh. The script prompts for a wallet password (give any password) and the API schema password. Upon successful completion, the shared_volume directory contains the sqlnet.ora, tnsnames.ora and the wallet directory.
7. Change directories into the PgxDataStudio/docker folder. Edit the docker-compose.yml file. For the python-interpreter entry (python-interpreter), update the volumes entry to map to the host shared_volume directory location. That is, provide the value for the <full_path> , the directory path of the shared_volume.
8. Start the datastudio docker containers using docker-compose as sudo docker-compose up -d.

   This step starts all the docker containers. The console shows the status. To confirm, verify using sudo docker ps. This should list the 5 data studio docker containers listed in the docker-compose file.

You can now login to the Oracle Healthcare Translational Research Next Gen and click on the datastudio images to navigate to the data studio notebooks.

For more information, see:

• Post-Installation Configuration Steps
  

Post-Installation Configuration Steps

One of the login users should be given the notebook admin role. In the Oracle Healthcare Translational Research Middle Tier WebLogic console, assign the user with the pgxds_admin_group role. This can be done in Users & Groups of Security Realm in the WebLogic console.

The following configuration steps are needed to enable and work with data studio:

• Oracle Healthcare Translational Research Middle Tier WebLogic - Listen Address
  
• Configurations in Notebooks
  

Oracle Healthcare Translational Research Middle Tier WebLogic - Listen Address

The Data Studio apps deployed in weblogic communicates within and with interpreters. By default, "localhost" is used in configurations. The managed server thus should able reachable via this address.

1. In the Oracle Healthcare Translational Research Middle Tier WebLogic, go to the managed server (hosting the apps). In the Configurations tab, if there is a value specified for the Listen Address, clear the field. Note that you will have to first Lock & Edit, then clear the value and click save.
   
2. Restart the Managed Server.

Configurations in Notebooks

1. Login to the Next Gen with the notebook admin user. Navigate to the Data studio by clicking on the left nav. The data studio app is displayed in a separate tab and the user will be logged into the data studio with this user id.
2. Navigate to the Interpreter tab in the notebook.
   
3. Navigate to the pgql tab.
4. Update the base_url to point to the pgx server running in the managed server. Provide in the format of http://hostname:port/pgx.

   Example: http://myhostname:8010/pgx.

5. Click Update.
6. Similarly, do the same for the for the base_url in the pgx-java tab, click Update.
   
7. For the python tab, ensure that the value of zeppelin.python is python3.

That's it. Now you can navigate to the Notebooks tab to access the notebooks.