• more performance, more objects
• S3 RDMA for high performance AI workloads
• S3 Caching Gateway based on the SGF6112 AllFlash Appliance
• StorageGRID Certification

More performance, more objects

To further scale StorageGRID, a few tweaks are introduced to scale better. The Cassandra Database now optimizes and compresses the table spaces to reduce disk footprint, the new maximum of objects is increased to 600 billion within a single cluster.

S3 RDMA

To increase IO bandwith, scalability and performance, but reducing CPU overhead at the same time, the new S3 RDMA protocol is introduced. This helps to reach bandwidths needed for modern AI use cases.

S3 Caching Gateway

To further accelerate IO, a dedicated SGF6112 Appliance can be used as a caching loadbalancer. Think about a SG110 system, just add a few more SSDs as cache. By doing this, frequently accessed/read data can be sent from the local SSDs and does not have to be fetched from the storage nodes in the cluster.
Fun fact: As the objects are placed in the cache, they do not have to be reassembled, no re-assembly from EC chunks to whole objects is needed… further reducing latency.

StorageGRID Certification:

You can finallly get certified in StorageGRID! Did you know, that there is a real certification, not only an accredetation? Log into your certmetrics profile and search for:
StorageGRID 11.9.0 Professional Level Certification Now Live: NetApp StorageGRID Administration Professional Exam (NS0-076)

You’re still hungry for more info? Read more exciting details in Vishnus article.