Summary of changes

This topic summarizes changes to the IBM Storage Scale licensed program and the IBM Storage Scale library. Within each topic, these markers ( ) surrounding text or illustrations indicate technical changes or additions that are made to the previous edition of the information.

Summary of changes
for IBM Storage Scale 5.2.0
as updated, April 2024

This release of the IBM Storage Scale licensed program and the IBM Storage Scale library includes the following improvements. All improvements are available after an upgrade, unless otherwise specified.

End of support for the RHEL 7 operating system

Starting with IBM Storage Scale 5.2.0, Red Hat Enterprise Linux (RHEL) 7 is no longer supported. For more information, see What is supported on IBM Storage Scale for AIX, Linux, Power®, and Windows?

TCT is now discontinued

You can address the discontinuation of Transparent Cloud Tiering (TCT) either by migrating to Active File Management (AFM) or by disabling TCT. Our strategic direction for cloud object storage tiering is IBM Storage Scale AFM. If you are currently using TCT, you must migrate to AFM before you upgrade to 5.2.0; otherwise, the existing TCT environment will be lost.

For information about how to migrate from TCT to AFM, see the documented processes for filesets and for file systems in IBM Storage Scale: Administration Guide.
For information about how to disable TCT, see Permanently uninstall Cloud services and clean up the environment.

AFM and AFM DR-related changes

Added the following new parameters for the configuration:
- afmFastLookup
- afmLookupMapSize
- afmObjMUCheckFName
For more information, see mmchconfig command.
Added the afmDRRPOSync parameter for the callback. For more information, see mmaddcallback command.

AFM to Cloud Object Storage

For the multi-site replication, IBM Storage Scale Advanced Edition, IBM Storage Scale Data Management Edition, or IBM Storage Scale Erasure Code Edition are required. For more information about this feature, see Multi-site replication.

Asymmetric data replication

Added the asymmetric data replication feature to enable a file system to concurrently place data in more than one pool at a time. This feature allows to concurrently make productive use of a pool of high-performance storage (called a performance pool) and a pool of IBM Storage Scale Native RAID storage (called a reliable pool). By harnessing these pools, the asymmetric data replication feature provides an excellent balance of performance, reliability, and availability. This feature is initially supported only in specific IBM Storage Scale Systems configurations. For more information about what configurations are currently supported and platform-specific instructions on how to create an asymmetric data replication file system, see the IBM Storage Scale Frequently Asked Questions and Answers.

Big data and analytics changes

For information on changes in IBM Storage Scale Big Data and Analytics support and HDFS protocol, see Big Data and Analytics - summary of changes.

IBM Storage Scale Container Storage Interface driver changes

For information on changes in the IBM Storage Scale Container Storage Interface driver, see IBM Storage Scale Container Storage Interface driver - Summary of changes.

IBM Storage Scale Container Native Storage Access changes

For information on changes in the IBM Storage Scale Container Native Storage Access, see IBM Storage Scale Container Native Storage Access - Summary of changes.

IBM Storage Scale Erasure Code Edition changes

For information on changes in the IBM Storage Scale Erasure Code Edition, see IBM Storage Scale Erasure Code Edition - Summary of changes.

Cloudkit changes

Cloudkit now offers composable deployment architecture.
Enhancements to cloudkit to upgrade clusters created on Amazon Web Services (AWS) or Google Cloud Platform (GCP).
Support for AFM-COS on AWS through cloudkit.
As a technology preview, IBM Storage Scale deployment on Azure by using cloudkit. For more information, see the following IBM Storage Scale support page: https://www.ibm.com/support/pages/node/7149295
As a technology preview, AFM-COS support on GCP through cloudkit. For more information, see the following IBM Storage Scale support page: https://www.ibm.com/support/pages/node/7148057
As a technology preview, creation of multiple file systems on AWS by using cloudkit. For more information, see the following IBM Storage Scale support page: https://www.ibm.com/support/pages/node/7148058
As a technology preview, fleet (also known as rapid expansion) on GCP through cloudkit. For more information, see the following IBM Storage Scale support page: https://www.ibm.com/support/pages/node/7148055

Discontinuation of the CES Swift Object protocol feature

CES Swift Object protocol feature is not supported from IBM Storage Scale 5.1.9 onwards. CES Swift Object will be replaced by CES S3.
IBM Storage Scale 5.1.8 is the last release that has CES Swift Object protocol.
IBM Storage Scale 5.2.0 tolerates the update of a CES node from IBM Storage Scale 5.1.8 or 5.1.9.
- Tolerate means:
  - The CES node will be updated to 5.2.0.
  - Swift Object support will not be updated as part of the 5.2.0 update.
  - You may continue to use the version of Swift Object protocol that was provided in IBM Storage Scale 5.1.8 on the CES 5.2.0 node.
  - IBM will provide usage and known defect support for the version of Swift Object that was provided in IBM Storage Scale 5.1.8 until you migrate to a supported object solution that IBM Storage Scale provides.
Please contact IBM for further details and migration planning.

File system core improvements

Enhanced online file system checking to repair corruptions
The mmfsckx command now supports a [--repair] option, which can repair some file system errors/corruptions while the file system remains mounted. For more information, see mmfsckx command.
For new clusters that are installed with IBM Storage Scale 5.2.0 or higher. The default values for the following config options are changed to provide better out-of-the-box performance for a wide variety of workloads: pagepool, workerThreads, numaMemoryInterleave, and ignorePrefetchLUNCount. Refer to the mmchconfig command for the new default values.
When IBM Storage Scale creates inode alloc regions, the default value of the -n option to mmcrfs is now interpreted as 512; before, this default value was interpreted as 256. When a new file system is created, and no explicit -n option is specified in the mmcrfs command line, IBM Storage Scale now allocates at least 512 inode alloc regions, instead of at least 256 inode alloc regions. This change allows larger clusters (over 100 nodes) that are created in this manner to have better inode allocation performance. For more information, see mmcrfs command.
IBM Storage Scale now supports the inode segment manager, which is a feature useful for workloads of parallel file creation. For more information, see mmcrfs command.
The configuration option prefetchPartitions is introduced to contribute to reducing lock contention for workloads that depend on prefetch or write-behind in an IBM Storage Scale filesystem. For more information, see mmchconfig command.
IBM Storage Scale introduces the graceful emergency recovery procedure for thin-provisioned storage systems, LUNs, and volumes created from FCM drives. This new recovery procedure provides a new mounting mode for space reclaim that allows system administrators and users to delete files to expand the free physical space for the system.

File system protocol changes

CES NFS is upgraded to version 5.7. NFS 4.1 is enabled by default in CES NFS. This is applicable for the installation of a new cluster with IBM Storage Scale 5.2.0.

Installation toolkit changes

As a technology preview feature, toolkit supports CES S3 installation and configuration.
Extended operating system certification and support.
Toolkit code enhancement to work with the latest Ansible library with base operating system repositories.
ECE multi-vdisk support for file system creation.

Management API changes: The following endpoints are modified:
For more information, see IBM Storage Scale management API endpoints.

Python-related changes: IBM Storage Scale 5.2.x supports Python 3.8 or later. It is recommended that Python 3.8 is installed through the OS package manager (For example, yum install python311). If you install Python 3.8 by other means, then unexpected results might occur, such as failure to install gpfs.base for prerequisite checks, and workarounds might be required.

Linux® on Z changes: Added shared memory communications - direct memory access (SMC-D) support. For more information, see Shared memory communications on Linux on Z nodes.

System heath changes

Python 3.8 or later must be installed for system health and call home.
Added monitoring of CES S3.
The mmhealth event list command has a new CLI option resolvable to list all resolvable events.
The mmhealth node show command has a new CLI option --last-check to show last check times instead of the times of the last state change.
The mmhealth command has new monitoring tips for new mmchconfig recommendations in 5.2.0:
Tips:
- pagepool must be >= 1/3 RAM or 4 GB.
- workerThreads must be >= 256.
- ignorePrefetchLUNCount must be yes.
- numaMemoryInterleave on NUMA systems must be yes.
- Health monitoring listens to GNR callbacks to reduce failure detection time.
For more information about the mmhealth command, see mmhealth command.

Call home changes:

From IBM Storage Scale 5.1.9.2, the gpfs.snap command has a new CLI option --upload-snap to automatically upload through call home. For more information, see gpfs.snap commandthe gpfs.snap command in IBM Storage Scale: Command and Programming Reference Guide.
From IBM Storage Scale System 6.1.9.2, a unified call home is used for all IBM Storage Scale System installations.

Performance monitoring changes:

A Prometheus Exporter for perfmon data was implemented.
Expired keys of the performance monitoring can be automatically deleted.
- This feature is enabled by default for IBM Storage Scale Container Native Storage Access (CNSA) installations.
- In IBM Storage Scale, this feature is disabled by default, it can be enabled on demand.
For more information, see Removing non-detectable resource identifiers from the performance monitoring tool database.

Native REST API (technology preview): IBM Storage Scale introduces the native REST API as a technology preview. The feature adds a new control plane component to the IBM Storage Scale stack for administering clusters instead of the mm-command layer. The native REST API also adds a few security enhancements. For more information, see the following IBM Storage Scale support page: https://www.ibm.com/support/pages/node/7145585.

CES S3 (Technology preview): IBM Storage Scale 5.2.0 supports technology preview for the CES S3 (Simple Storage Service) access protocol. CES S3 provides high-performant and scalable S3 object access to data, which is stored in IBM Storage Scale file systems. It is integrated with Cluster Export Services (CES) to provide highly available S3 access service in an IBM Storage Scale cluster. For more information, see the following IBM Storage Scale support page: https://www.ibm.com/support/pages/node/7145681.

Advanced RISC Machine (ARM) architectures: IBM Storage Scale is supported on ARM architectures (aarch64 and arm64). For more information, see Is IBM Storage Scale supported on Advanced RISC Machine (ARM) architectures? in the FAQ page.

Commands, data types, and programming APIs

The following section lists the modifications to the documented commands, structures, and subroutines:

New commands

There are no new commands.

New structures

There are no new structure changes.

New subroutines

There are no new subroutines.

New user exits

There are no new user exits.

Changed commands

cloudkit
mmaddcallback
mmafmcosconfig
mmafmcosctl
mmafmctl
mmapplypolicy
mmbackup
mmces
mmchconfig
mmchfileset
mmchfs
mmcrfs
mmdiag
mmedquota
mmexpelnode
mmfsckx
mmhdfs
mmhealth
mmimportfs
mmkeyserv
mmlsfs
mmobj
mmnfs
mmnetverify
mmremotecluster
mmrestripefile
mmsdrrestore
mmsetquota
spectrumscale
mmxcp
mmaddpdisk

Changed structures

There are no changed structures.

Changed subroutines

gpfs_fcntl

Deleted commands

There are no deleted commands.

Deleted structures

There are no deleted structures.

Deleted subroutines

There are no deleted subroutines.

Messages

The following are the new, changed, and deleted messages:

New messages: 6027-2428, 6027-3418
Changed messages: 6027-3415 and 6027-3416
Deleted messages: There are no deleted messages.