Figure 6: Creating a workbench within the sample project.
Add a workbench by clicking the Create workbench button (Figure 6.)
On the Create workbench page, complete the following information. Note: Not all fields are required. (Figure 7)
- Name
- Description
- Notebook image
- Deployment size (container size)
- Environment variables
- Cluster storage name
- Cluster storage description
- Persistent storage size
Once you have entered the information for your workbench, click Create.
Figure 7: Complete the fields to configure your data science project’s workbench.
After creating the workbench, you will return to the anomaly-detection project page. (Figure 8)
Figure 8: The anomaly-detection project now includes a workbench and cluster storage.
Notice that a red exclamation appears under the status indicator in Figure 9. If we hover over that icon, we can see that an “insufficient memory” error has occurred.
Figure 9: Encountering an error in OpenShift Data Science.
Looking closely at the error message, it appears that we don’t have enough resources; we chose a larger storage size than was available in our container. To resolve this error, we need to edit the cluster storage so that it does not exceed our container size.
Once we have our workbench and cluster storage set up, we can add data connections. Click the Add data connection button to open the data connection configuration window. (Figure 10)
Figure 10: Click the Add data connection button to add a new connection.
Within this window, as shown in Figure 11, you can add the following items:
- Name
- AWS_ACCESS_KEY_ID
- AWS_SECRET_ACCESS_KEY
- AWS_S3_ENDPOINT
- AWS_DEFAULT_REGION
- AWS_S3_BUCKET
Figure 11: Configuring a new data connection in OpenShift Data Science.
After completing the required fields, click Add data connection. You should now see the data connection displayed in the main project window. (Figure 12)
Figure 12: The main data science project window shows a newly created data connection.
After creating the data connection, you can configure your model server. Select Configure server. (Figure 13)
Figure 13: Configuring a model server in OpenShift Data Science.
In the pop-up window that appears, as in Figure 14, you can specify the following details:
- How many model server replicas to deploy
- Model server size (compute resources per replica)
- Model route
- Token authorization
Figure 14: Options for configuring your model server.
After adding and selecting options within the Configure model server pop-up window, click Configure to configure the model server.
Your data science project overview shows that your model server has been configured with the default OpenVINO Model Server (ovms), as shown in Figure 15.
Figure 15: The model server in this example was configured with ovms, the OpenVINO Model Server.
Once you have set up your model server, you can deploy your model. From the main Red Hat OpenShift Data Science dashboard, choose Model Serving. (Figure 16)
Figure 16: To deploy your model, select the Model Serving option in the OpenShift Data Science menu.
To add a model to be served, click the Serve model button. Doing so will initiate the Deploy model pop-up window. (Figure 17)
Figure 17: Configure properties for deploying your model.
You can now use your existing data science project configurations to help you deploy your model. Notice that you can select the anomaly-detection project that we created previously. You can also choose the anomaly-detection-connection connection that you created earlier in your data science project!
When you are ready to deploy your model, select the Deploy button. When you return to the Deployed models page, you will see your newly deployed model. (Figure 18)
Figure 18: Viewing the newly deployed anomaly-detection-model for the anomaly-detection project.
This concludes our learning path on creating a data science project.