The iServer Data Science Service provides two deployment methods. For details, see the Data Science Service Pack Installation Guide . The address of the initialization wizard accessed after startup is different for different deployment methods:

  • Deploy the data science service in Docker, and configure the initialization wizard by accessing HTTP://{ IP }: 8000 after startup
  • Deploy the Data Science Service in the Kubernetes cluster, and access HTTP://< K8S Cluster Node IP >:31000 after startup. Perform the initialization wizard configuration

Initialize the wizard configuration

Starting iServer Data Science Services for the first time brings you to the Initialization Wizard page (enter the address of the deployment in the browser). You need to configuration the data user system and administrator for the science service. Two user systems are currently supported: using the KeyCloak user or operating system user; also configure the Notebook running mode, which currently supports running in local process, in local Docker container and in the Docker container in K8S.

Configure the process

1. Select the authorization type

User system and administrator for configuration the data science services.

Currently, the following two user systems are supported. Click "Next" after configuration.

  • Use the KeyCloak user

Before using KeyCloak users, the KeyCloak service must be available. See KeyCloak Installation and Configuration for details.

    •  Administrator account: can use users in KeyCloak service

    • Callback URL: the IP and port of the current data science service. The format is http://ip:port/hub/oauth_callback.

    • KeyCloak address: KeyCloaks ervice address in the format of http://ip:port.

    • KeyCloak Domain: The domain name set in KeyCloak. For example, jupyter hub ".

    • Client ID defined by KeyCloak: The client ID set in KeyCloak. For example, "jupyterhub".

    • Client secret key defined by KeyCloak: client secret key set in KeyCloak.

  • Use the operating system user

    •  Administrator account: Users with non-root privileges existing in the current operating system can be used.

Note: In the current Docker image, two non-root users, "iserver" and "guest", have been created, and the user name and password are the same.

2. Set the Notebook operating mode

The following three Notebook operating modes are currently available:

  • Run in local process: Run on the local operating system.
  • Docker container running locally:
    • Runs in a local Docker container.
    • Data science service mirror name: The mirror name of the launched data science service. Default: registry.cn-beijing.aliyuncs.com/supermap/iserver-datascience:latest. Note: The filled image must exist in the local Docker environment.
  •  running on K8S Docker container in: running in a K8S cluster.

After setting, click "Next"

3.Confirm the initialization information.

 After the configuration of the above steps is completed, click" Initialize "; to modify the configuration, click" Back "

4.Initialization completed.

After initialization, if the Data Science Service is running on Docker in a non-K8S cluster In the container, a manual restart is required with the following command:

docker Start Docker Container ID

5. Log in to Data Science Services.

Log in by using the account password of the user in the user system according to the user system selected in the first step.