Fixed Issues

DSE-28066: Port conflict on Single Node

CML UI shows timeout error during upgrade on a single node cluster because of an underlying port conflict error: 1 node(s) didn't have free ports for the requested pod ports.

This issue has been fixed.

DSE-27283: Pod evaluator should not use hostNetworking on Azure and Private Cloud

CML workspace Pod Evaluator used host networking which could cause HA issues and deployment failures owing to port conflicts.

This issue has been fixed.

DSE-28005: Environment variables hidden on Application>Settings page

Previously, environment variables were not visible or modifiable on Application's settings page.

This issue has been fixed.

DSE-32672: Timezone issue for start, stop and schedule session activities

Earlier, the saved date value for sessions was not set to the UTC timezone. This issue caused the sessions to terminate for Private Cloud Data Services if the web pod was running in a timezone different from UTC.

This issue has been fixed.

DSE-34100: Migration readiness check fails on BuildKit-enabled clusters

Earlier, the migration readiness checks failed on clusters having BuildKit instead of Docker.

This issue has been fixed.

DSE-33099: CML team members’ type value could not be configured

Earlier, the team members' type value could not be edited, and team members with the value none could not contribute to synchronized team activities. All team members need to have write permissions so that the roles can be set from the UI.

This issue has been fixed.

DSE-31914: The model metrics ingress configuration does not support TLS

The model metrics ingress configuration did not have a template for specifying the TLS section.

This issue has been fixed.

DSE-32250: The spark executor resource usage is not reported correctly

Earlier, when you exported the Spark executor usage data from the Site Admin > Usage tab, the exported data in the report was the same as the session data the executor was launched in. This included incorrect CPU, memory, create time, and stop time details of the Spark executors.

This issue has been fixed.

DSE-33098: Unable to synchronize teams in CML without having an MLAdmin user in that team

Earlier, CML required that each team had one user with the MLAdmin role, by default, to manage the team, even if the role was not necessary for the team later. It is now possible to manage the team with the MLUser role without having the MLAdmin role in the team.

This issue has been fixed.

DSE-33765: CML workspace upgrade fails due to a wrong logic for getting helm values

Earlier it was wrongly assumed that the PersistentVolumeClaim (PVC) requests are returned from K8 in the same order. This wrong logic could lead to the CML upgrade being broken. To avoid this, the returned list shall rather be searched for the particular volume instead.

This issue has been fixed.

DSE-34088: Unable to view Spark UI executor logs in CML

The names for Spark executor pods changed in the dashboard pods DB table.

This issue has been fixed.

DSE-34112: Migration readiness timeout logic is required to avoid the migration being stuck

A timeout value is implemented to exit the migration readiness workflow.

This issue has been fixed.