With StorageGRID® Webscale 11.0 announced at NetApp Insight, NetApp has again pushed object storage into the frontier of hybrid cloud use cases. The new features make StorageGRID:

  • The best cloud integrated object storage solution, with new metadata search integration
  • The most S3 compatible platform, with support for self-service, allowing developers and end users to use object storage efficiently
  • One of the most performant solutions, supporting multiple primary storage use cases

And now, a short history lesson. NetApp® StorageGRID was a pioneer in object storage. As described in theregister.com, StorageGRID has many years of enterprise hardening and a rich feature set that has few, if any, rivals.

 

Over the past few years, StorageGRID has continued to innovate and offer many first-to-market features:

  • First to provide dynamically modifiable ILM policies that can tier back and forth between replication and geo-dispersed erasure coding.
  • First to support a simple software consumption licensing model based on used capacity.
  • First to support mixed infrastructure deployments, where a single object storage cluster is deployed across compute and storage platforms from different vendors, freeing users from vendor lock-in.
  • First to support the most granular access control and security possible, at a per user/per API call level, and compatible with AWS IAM.

Now, StorageGRID 11.0 Steps into the Cloud

In addition to the existing 2-year-old feature of Cloud Tiering, StorageGRID now supports two additional cloud features:

  • CloudMirror. With CloudMirror, StorageGRID replicates objects in native format into a target S3 bucket in the cloud or on premises. This means that you can apply cloud resources, such as AWS Elastic Map Reduce, directly on the data that has been replicated to the cloud.
  • Event Notifications. StorageGRID now sends event notifications for events occurring on premises to AWS SNS (or to your on-premises notifications receiver). This means that you can build serverless, event-driven pipelines using on premises and cloud resources. As an example, you could use IBM Watson or AWS Image Rekognition and annotate your on-premises data with metadata generated by cloud AI services.
Figure 1) StorageGRID 11.0 hybrid cloud features.

Event Notifications and CloudMirror provide powerful Lego-like building blocks that can be combined to create workflows that span on premises and public clouds.

 

These features (Event Notifications, CloudMirror, and Cloud Tiering) make StorageGRID the leader in cloud integrated object storage.

Searching Your Metadata

Our customers have  been asking for metadata search, and after considering  all the different approaches to providing search, we concluded that the fastest, most scalable search is achieved by exporting StorageGRID metadata into purpose-build search and analytics clusters.

 

Many object store solutions build these capabilities into the object storage system, and therefore have either scalability or performance challenges.

 

StorageGRID 11.0 exports metadata into elastic search clusters that can be scaled to hundreds of billions of keys and tags, with extremely fast search performance. The export functionality is built in a generic manner, allowing us to support additional endpoints, such as Hadoop clusters, in the future.

 

Figure 2) StorageGRID 11.0 metadata search.

Most S3 Compatible Object Store

StorageGRID uses the standard S3 APIs to deliver on the features of replication and notification.

In addition:

  • StorageGRID now supports S3 metadata tagging
  • Supports end-user-specified S3 regions, providing additional granular control of ILM by end users and making the most powerful policy engine even more powerful.

Between the features of replication, notification, and tagging, we now support 9 additional S3 APIs, with equivalent IAM access control levers, making StorageGRID the most S3 compatible object storage.

Self-Service for Dev-Ops

StorageGRID 11.0 continues to emphasize self-service, allowing developers and end users to use the object store, independent of the storage administrator.

 

In addition to the already extensive AD/LDAP integration, which allows self-service for end users, the StorageGRID self-service capabilities have been extended into all of the new cloud features.

 

The features of Event Notifications, Metadata Search, and CloudMirror are all self-service oriented, allowing individual users and developers to configure and use them, accelerating time to market while reducing the burden on storage administrators.

StorageGRID 11.0: A High-Performance Solution

Performance has always been a key focus for StorageGRID, and 11.0 increases its already fast performance by 2x for multicore platforms. This means that StorageGRID can now be the primary tier for many workloads.

 

To learn more about how you can store and manage unstructured data at scale using StorageGRID 11.0, visit netapp.com. To see it in action, request a live custom demo.

mm

Vishnu Vardhan