Tutorial

Data Science Service Creation

Environment Creation

SuperMap iManager supports to ‘one-click’ creating Data Science Service environment, please refer to the following steps:

1. Open Sites Management on the left navigation bar.
2. Clicks on + Create.

3. Fills in the name of the Data Science Service environment, and select Data Science Service in the Site type.

   Advanced Options:

   • Namespace: Specifies the namespace for the created site. If not specified, the system will assign a namespace for it. The specified namespace will be created automatically when it does not exist and will be deleted when the site is deleted. The existing namespace will not be deleted.
   • Data Science service proxy-public external port: Set the port of proxy-public access address, a random port would be used if not set.
   • Spark block manager external port: Set the port for accessing the distributed storage system BlockManager of Spark, a random port would be used if not set.
   • Spark driver external port: Set the port for accessing the Driver of Spark, a random port would be used if not set.
   • Spark executor external port: Set the port for accessing the Executor of Spark, a random port would be used if not set.
   • Spark web ui external port: Set the port for accessing the Web client of Spark, a random port would be used if not set.
   • Use Service Mesh: Click to open service mesh function, the site is available to achieve service trace and service metrics. Modify the value of deploy_istio_enabled to ‘true’ in the ‘values.yaml’ file of iManager deployment package first if you want to use service mesh, please refer to Guides > Deployment Guide > Install iManager.

   • Resources Allocation: Set the limitation of CPU and Memory of the Site. The CPU/Memory in the Site is different to the CPU/Memory in hardware machine, the CPU/Memory in the Site is the CPU/Memory occupy of a service node in Kubernetes, the total CPU/Memory of the Site is equal to the sum of CPU/Memory of service nodes.

     • CPU(Required): Set the maximum available CPU of the Site.
     • Memory(Required): Set the maximum available Memory of the Site.
     • Pods: Set the maximum number of available pods in the Site.
     • Default container CPU limit: Set the maximum available CPU of a single container when creating.
     • Default container Memory limit: Set the maximum available Memory of a single container when creating.

   Notes:
   If the Site enable Service Mesh, each service node would create a istio-proxy, a istio-proxy occupies 2 core CPU and 1 GB Memory, the total CPU/Memory of the Site equal to the sum of CPU/Memory of service nodes plus the occupy of the istio-proxy.

4. Clicks on OK.
5. Check the status on the Data Science Service page, while the status change to ‘Normal’, the Data Science Service created successfully.

User Creation

Data Science Service supports to multiple users concurrent access, each login user will create a pod in the environment, all the pods are working independently. There are two users by default, you can view the users information by checking Account on the service details page. If you need more users, follow the steps below to create:

1. On the home page of Data Science Service, clicks on the name of hub to enter container details page;
2. Open Command Pad;

3. Executes the following command to create a user(‘newuser’ in the command is the username, you can replace to your username):

   adduser newuser

4. Fills in the password;
5. Fills in the user’s information;
6. Confirm the user’s information.

After creation, login and use Data Science Service directly.