NetApp boosts StorageGRID performance, capacity

NetApp has upped the capacity and performance of its StorageGRID object storage systems.

StorageGRID is an S3-compatible on-premises object storage system. It can manage and store objects in its own globally distributed infrastructure, and in Amazon S3 or S3-compatible object stores or public clouds. They can function as backup or archival tiers for ONTAP storage arrays, using FabricPool tiering software. NetApp has introduced five new StorageGRID models adding capacity and performance while reducing the total cost of ownership.

There are two kinds of StorageGRID system: disk-based SG5000 and SG6000, flash-based SGF6000 storage nodes, alongside SG100 and SG1000 admin and load-balancing nodes. All of the existing models get an upgrade with a higher-capacity flash storage mode introduced as well.

Vishnu Vardhan, NetApp
Vishnu Vardhan

A blog by NetApp’s Vishnu Vardhan, StorageGRID Director of Product Management, says: “These new appliances support our growth in primary-only object storage use cases where the first and all subsequent copies of data reside on object only. These new StorageGRID appliances will … accelerate the rollout of AI/ML applications and data lakehouses, which are increasingly being deployed on object storage.”

The new systems are the SG5812, SG5860, SG6160 and higher-capacity all-flash SGF6112 storage nodes plus two new load-balancers, SG110 and SG1010. They run v11.8 of the StorageGRID software, which introduced QLC (4bits/cell) SSD support and improved information lifecycle management (ILM) policy features.

NetApp positions them like this:

NetApp StorageGRID

Here’s how the old and new StorageGRIDs compare: 

NetApp StorageGRID specs
Upgrade paths are indicated by arrows. We view the SGF6112 as additive to the SGF6024

The disk-based SG5812 and SG5860 are cost-optimized for secondary storage while the SG6000s are more performance-oriented:

  • SG6024 – entry-level disk
  • SG6160 – mid-range disk, Fabric Pool optimized for transactional small object workloads
  • SG6160 expansion – large-scale (hundreds of PBs) deployments, data lake with >3.2 PB/node
  • SGF6112 – all-flash performance focused on analytics, IoT, and streaming primary workloads

Vardhan says: “The all-flash SGF6112 appliance is a game changer for customers looking to upgrade from older systems. It offers a similar price per GB while delivering up to a 3x performance increase.” 

When tiered to ONTAP all-flash systems, “StorageGRID enables customers to achieve the lowest $/GB all-flash solution in the industry today. FabricPool + StorageGRID frees up to 90 percent of the capacity from an all-flash system, while retaining performance for the hot data.”  

The admin/load-balancing systems have also been upgraded and get faster CPUs:

NetApp StorageGRID load-balancers

The SG1010 is for large deployments needing high performance with small-to-midsize deployment needs met by the SG110. Get a StorageGRID datasheet here.