To allow for more granular scaling, you can now install data-only Storage Nodes. Where metadata processing isn't critical, you can optimize your infrastructure cost-effectively. This flexibility helps accommodate varying workloads and growth patterns.

StorageGRID now supports short term credentials using IAM Roles Anywhere in Amazon S3 for Cloud Storage Pools.

Using long-term credentials to access S3 buckets poses security risks if these credentials are compromised. Short-term credentials have a limited lifespan, which reduces the risk of unauthorized access.

You can now configure a Cloud Storage Pool using an Amazon S3 endpoint. S3 Object Lock helps prevent accidental or malicious deletion of objects. If you tier data from StorageGRID to Amazon S3, having object lock enabled on both systems enhances data protection across the data's lifecycle.

By setting limits on S3 buckets, you can prevent tenants from monopolizing capacity. Additionally, uncontrolled growth can result in unexpected costs. By having defined limits, you can better estimate tenant storage expenses.

To enhance scalability, StorageGRID now supports up to 5,000 S3 buckets per tenant. Each grid can have a maximum of 100,000 buckets.

To support 5,000 buckets, each Storage Node in the grid must have a minimum of 64 GB of RAM.

AutoSupport now automatically creates hardware failure cases for legacy appliances.

Node clone usability has been expanded to support larger storage nodes.

ILM ingest time rules with a time period of Days now also remove expired object delete markers. Delete markers are only removed when a time period of Days has passed and the current delete maker has become expired (there are no non-current versions).

Refer to How S3 versioned objects are deleted and Example of bucket lifecycle taking priority over ILM policy.

To provide a smooth and efficient transition to StorageGRID next-generation hardware, node decommissioning has been improved.

Load balancer endpoint access logs contain troubleshooting information, such as HTTP status codes. StorageGRID now supports exporting these logs to an external syslog server. This enhancement allows for more efficient log management and integration with existing monitoring and alerting systems.

Grid administrators can now update and rotate SSH keys. The ability to rotate SSH keys is a security best practice and a proactive defense mechanism.

When an unknown entity signs in to the Grid Manager as root, an alert is triggered. Monitoring root SSH logins is a proactive step toward safeguarding your infrastructure.

The Erasure-coding profiles page is now located at CONFIGURATION > System > Erasure coding. It used to be in the ILM menu.

The search field in the Grid Manager now includes better matching logic, allowing you to find pages by searching for common abbreviations and by the names of certain settings within a page. You can also search for more types of items, like nodes, users, and tenant accounts.