Action Required Service Updates

Leverage the Benefits of the Changes to IBM Analytics Engine Clusters

Share this post:

Updates to IBM Analytics Engine

IBM Analytics Engine now includes many updates to its cluster setup and management processes. Not only have several new fixes been added to the cluster software stack, but significant performance improvements have been made to network throughput between IBM Analytics Engine and IBM Cloud Object Storage.

These and many other changes have been implemented to help you maximize on maintaining stateless clusters. Why are stateless clusters more efficient and effective? One enormous benefit is that it gives you the freedom to spin up new clusters and remove old clusters driven by your data analysis use-cases.

Customizing the cluster creation process

Customizing your cluster creation process and separating compute from storage are vital elements of moving toward achieving stateless clusters. If your data resides outside of the cluster in IBM Cloud Object Storage, you can readily change the compute power (the cluster) to include new features, other packages, important fixes, or even to increase the compute power—all without losing your data or job applications. With your data safely stored outside of the cluster, applying your use-case specific cluster customization makes creating a new cluster an easy process for you.

And now that IBM Analytics Engine is available in different regions across the world, you can create clusters in multiple regions to avoid cluster downtime.

Action required: Create new clusters

In view of these value-driven changes to IBM Analytics clusters, we encourage you to create new clusters, especially if you are still using clusters created before August 17, 2018, as these clusters do not include any of the latest updates. Before you delete your old clusters, you should back up your data, metadata, and any changes done on the clusters so that no data is lost.

To help you create your new clusters efficiently and effectively, we have created a Best Practices guide for recommended cluster design and deployment practices that maximize on maintaining stateless clusters.

For any additional help on this topic, please feel free to reach out through the IBM Cloud support center.

Release Manager - IBM Analytics Engine and BigInsights on Cloud

More Action Required Service Updates stories
October 31, 2018

Increased Availability with HA Masters in the Kubernetes Service: Actions You Must Take

IBM Cloud Kubernetes Service will soon increase your cluster availability by adding replicas of the Kubernetes master with fix pack version 1.11.3_1531, released on November 1, 2018. After you update, clusters will have three Kubernetes master replicas that you set up with a highly available (HA) configuration.

Continue reading

October 9, 2018

Updates to Voice Agent with Watson Service

As part of our drive to deliver improvements to the Voice Agent with Watson service, we are making changes in October around how we handle transferring calls and updating the pricing for the Voice Agent service.

Continue reading

September 11, 2018

Migrating from Retrieve and Rank and Document Conversion to Watson Discovery

Watson Retrieve and Rank and Document Conversion services will be permanently sunset on October 3, 2018. Switch now to Watson Discovery!

Continue reading