What's new

Review this information for a high-level summary of the new features and changes in 2.10.x release.

Version 2.10.0

Support for Dell servers

Dell manufactured servers are now supported, offering an alternative hardware vendor choice for compute and storage servers. This enhancement broadens hardware flexibility without altering the foundational architecture. You can collect logs, monitor events, and use Call Home support for Dell nodes, which facilitates failure detection and debugging. For more information about the hardware configuration, see Dell hardware overview.

Regional-DR

Regional-DR provides volume-persistent data and metadata replication across geographically dispersed sites. In a public cloud environment, this functionality mitigates risks associated with regional failures, ensuring business continuity while accepting a predictable degree of data loss. Regional DR configuration for provider mode Fusion Data Foundation storage is available so that you can have a centralized control of multiple clusters, cross cluster management, improved resource optimization, and ensure business continuity. For more information and how to configure, see Regional-DR configuration for Fusion Data Foundation.

AI

IBM Fusion Content-Aware Storage (CAS) service

This release introduces a new IBM Fusion service: IBM Fusion Content-Aware Storage (CAS) service 1.0.2.

It uses natural language processing (NLP) and AI technologies to extract semantic meaning from unstructured text data. It allows for more efficient data retrieval and integration with AI workflows, particularly in retrieval-augmented generation (RAG) processes. For more information about the CAS service, see Content-Aware Storage (CAS).

Additional GPU Card support using AMD MI210
This release provides more choices to configure your systems to support AI applications. It provides a powerful GPU with 64 GB of memory to support the processing of large LLMs.

Multi-cluster management

Dedicated drives for the Logical Volume Manager (LVM) file system

LVM is used in hosted control planes to provide persistent storage for etcd data. LVM allows for flexible storage allocation and management, ensuring that the etcd data is stored on fast, dedicated storage devices. For more information about this feature, see Dedicated drives for LVM filesystems.

Remote scale mount on bare metal hosted control plane
You can now configure remote mount on Hosted Control Plane similar to other label-supported service installations.

For more information about this feature, see Remote scale mount on bare metal hosted control plane.

Manage multi-cluster with Machine Engine (MCE)
It is now possible to manage multi-cluster IBM Fusion through MCE. It eliminates the dependency on Red Hat® Advanced Cluster Management for Kubernetes (ACM) so you can optimize your hybrid cloud infrastructure with a more cost-effective and scalable approach. For more information about how to use MCE, see Multi Cluster Engine (MCE).

Backup & Restore enhancements

Improved scalability
As the number of Spoke clusters connect to the Hub cluster changes, additional Kafka Bridge instances are dynamically added or removed as needed based on demand. Prior to this release, the Kafka Bridge on the hub had to scale up to handle more than a certain number of spoke clusters. Now, the Hub cluster has automatic Kafka Bridge scaling based on the number of spoke clusters connected to it.
Install and upgrade Enhancements
  • Custom configuration changes are now seamlessly preserved across upgrades, ensuring continuity and reducing manual adjustments. This enhancement supports Backup & Restore activities for large configurations and environments, providing a more reliable and efficient upgrade experience.
  • If other applications on the cluster use OADP or Streams for Apache Kafka with a version that differs from the one required by the Backup & Restore service, then the auto-upgrade process of Backup & Restore is prevented to maintain compatibility. These products do not support multiple versions of the same product within a single cluster due to conflicts arising from cluster-scoped CRD versions.
  • GPU node anti-affinity for Lenovo systems:

    The Backup & Restore service attempts to schedule deployment to non-GPU nodes. GPU nodes are used only when there are no other non-GPU nodes. For more information about GPU deployment, see AI workloads on IBM Fusion.

More flexible and efficient backup and restore
  • The IBM Fusion HCI 2.10 recipe introduces three additional hooks: JobHook, LabelHook and AnnotationHook, allowing for more controlled orchestration of application backup and recovery sequences. For more information, see Additional hooks.
  • Dynamic recipe provides a capability to handle the backup and restore of applications with multiple sub-components dynamically. With a dynamic recipe, you can define a parent recipe that automatically discovers child recipes that provide extra workflow information, making your recipes more dynamic. For more information, see Dynamic recipe.
  • The Backup & Restore service now has multiple Availability Zone (AZ) support. Note that the OADP and Streams for Apache Kafka operators currently do not have multiple AZ support and may be scheduled to nodes in any AZ.
  • Change Block Tracking (CBT) support of Filesystem mode PVCs on Ceph RBD storage classes. CBT greatly reduces backup times by eliminating the need to read the entire PVC to identify changes from the prior backup.
IBM Fusion HCI user interface enhancements
Enhancements to Backup & Restore user interface:

Several improvements have been introduced to the user interface to enhance usability and operational efficiency.

  • Recipe visibility:

    The recipe visibility is increased across various Backup & Restore pages, allowing users to more easily track and reference the recipe associated with a backup. It provides greater clarity and context during both backup and restore operations.

  • Policy Time Window Configuration:

    A new Policy Time Window setting is added. It is especially useful when multiple backups share the same policy or schedule, which can overload the agent with simultaneous jobs. The Policy Time Window allows backup jobs to remain queued for a user-defined period before getting canceled. This ensures more efficient utilization of a single agent replica throughout the day, reducing the need to spawn additional replicas or manually distribute backup schedules.

  • Jobs Page Enhancements:

    The Jobs page now includes the option to download a CSV file containing a complete list of resources included in a backup. This enables users to review the backup contents without the need to perform a test restore. Additionally, when a backup is configured using a recipe, the recipe name is now displayed in the Application, Policy Assignment, and Backup & Restore job panels, offering improved traceability.

  • Enhanced Namespace Mapping for Multi-Namespace:

    The user interface now supports custom mapping of namespace names during restore operations, making the process more flexible and intuitive.

  • Additional Usability Enhancements:

    You can now download a list of resources associated with any given backup directly from the interface, further improving transparency and control.

IBM Data Cataloging

Scalability Enhancements
  • Ingest up to 1 billion files or objects per day:

    With our optimized architecture, your organization can seamlessly ingest up to 1 billion files or objects daily. This ensures high-volume data operations are handled efficiently, delivering real-time insights and enabling faster, smarter decision-making. Whether you are feeding AI models or driving business intelligence, your datasets will always be accessible, structured, and ready for action.

  • Manage up to 30 billion objects:

    Data Cataloging empowers you to manage and maintain visibility across up to 30 billion data objects or files, delivering enterprise-scale oversight, data governance, and complete control over your digital assets—without compromising on performance or reliability.

Fast data retrieval: Execute complex queries in milliseconds
Run advanced searches and execute data queries in just milliseconds at scale. Our high-performance architecture ensures instant access to mission-critical data, enabling teams to act quickly, whether you are powering dashboards, driving automation through APIs, supporting analysts with self-service data discovery or enabling your ML/AI applications through the AI-ready datasets.
Edge intelligence: Rapid data ingestion and enrichment at the edge
Capture and enrich metadata directly at the data source, even in decentralized or regulated environments. Our edge ingestion capabilities allow for real-time processing and contextual enrichment, ensuring your data is compliant, clean, and analytics-ready before it even hits central storage.
Intuitive data tagging and labeling: improved streamlined tagging and labeling UX
Accelerate data classification and exploration with a simplified, intuitive tagging and labeling experience. Designed with ease-of-use in mind, our UX reduces manual effort while maintaining the accuracy and structure needed for data discovery, governance, and AI readiness.

Platform management

Support static IP addresses
From this release on wards, IBM Fusion HCI offers you the static IP option as an additional choice for IP address assignment. It eliminates the need to create Mac address in DHCP server of your data center. For more information about the benefits of this method, see Static or DHCP IP address assignment methods.
Centralized notification of Fusion events
In this release, a new feature is included to help IT Operations teams manage IBM Fusion HCI incidents and failures. IBM Fusion HCI now supports integrated alert notifications for all CRITICAL and WARNING Fusion events. With this enhancement, you can configure receivers like Slack, email, and webhooks to receive real-time notifications. Additionally, support for multi-cluster configuration enables centralized monitoring. It allows alerts from multiple IBM Fusion HCI clusters to be consolidated into a single, unified notification channel. It ensures that you are promptly notified of the important events across your clusters without the need to manually check the IBM Fusion HCI user interface for updates. For more information, see IBM Fusion event alerts.
Fusion License service enablement
Integration with the IBM License Service allows you to monitor and report the basic license usage of virtual processor cores by Fusion.

Fusion Data Foundation support

IBM Fusion HCI now includes support for Fusion Data Foundation 4.18. For more information about what is new in Fusion Data Foundation 4.18, see New features in 4.18. For enhancements in 4.18, see Enhancements in 4.18.

OpenShift Container Platform support

IBM Fusion HCI 2.10 includes support for OpenShift Container Platform 4.18.

From IBM Fusion HCI 2.10, support for OpenShift Container Platform 4.12 is deprecated.

New currency support

For a complete list of currency support, see Support matrix.