SQL Server Big Data Clusters CU10 brings important new capabilities

[mood]

SQL Server Big Data Clusters CU10 brings important new capabilities

 

 

With the release of SQL Server 2019, Microsoft brought Big Data Clusters (BDC) for customers as a way to utilize both structured and unstructured data. Running exclusively on Linux containers that are hosted on Kubernetes, these clusters can be deployed both on-premise and on the cloud. Last year, the tech giant released the cumulative update 5 (CU5) for SQL Server 2019, focusing on expanding capabilities offered through BDC.

 

Today, Microsoft has unveiled the SQL Server Big Data Clusters CU10 release. With this update, a few new important capabilities have been introduced. These have been highlighted as follows:

• Upgraded base images from Ubuntu 16.04 to Ubuntu 20.04.
• High availability support for Hadoop KMS components.
• Additional configurability of SQL Server networking and process affinity settings at the resource-scope.
• Resource management for Spark-related containers through cluster-scoped settings.

Notably, when it comes to base image upgrading, the documentation for the CU10 release states that due to the stricter security requirements of Ubuntu 20.04, users may face issues if they attempt to use BDC to connect to SQL Server instances before SQL Server 2017. If such a case arises, they should refer to the following solution.

 

When upgrading existing BDC, the new CU10 images will be utilized once the usual in-place upgrade to the CU has been performed. Additionally, High Availability capabilities for Hadoop's key management service (KMS) components further improves the Encryption at Rest feature set. After the upgrade, all namenode pods will have a KMS, not only offering increased high availability, but also providing a performance boost for encryption operations on encryption zones.

 

 

Source: SQL Server Big Data Clusters CU10 brings important new capabilities