One year ago, we took a big step toward making NetApp AFF A-Series systems the most cloud-connected all-flash solutions in the industry by introducing FabricPool. First available in NetApp ONTAP® 9.2, FabricPool is a NetApp Data Fabric technology that enables automated tiering of inactive (cold) data to low-cost object storage tiers. FabricPool can be used with either off-premises public clouds, such as Amazon S3 and Microsoft Azure Blob Storage, or on-premises private clouds such as NetApp StorageGRID® Webscale. The automated cloud tiering is transparent to applications and allows enterprises to take advantage of cloud economics without sacrificing performance or having to rearchitect storage infrastructure.

 

Now, with the launch of ONTAP 9.4, we’ve added several new key capabilities to FabricPool.

Auto Volume Tiering Policy

In ONTAP 9.2 and 9.3, FabricPool was able to tier NetApp Snapshot copies and data in data protection volumes to an external capacity tier by using the Snapshot-Only and Backup volume tiering policies. It was a great start, but the number one feature request from our customers was to add an ability to tier inactive (cold) data in the active file system to the cloud.

 

In ONTAP 9.4, a new volume tiering policy is now available: Auto. When a volume tiering policy is set to Auto, all cold blocks in the volume are moved to the capacity tier. By default, it takes 31 days for inactive blocks to become cold. But the Auto policy’s cooling period is adjustable, so you can customize it, per volume, to the time periods that work best in your environment.

 

Volume modify -vserver <svm_name> -volume <volume_name> 
-tiering-policy auto 
-tiering-minimum-cooling-days <2-63>

Note: To set tiering-minimum-cooling-days, advanced privilege level is required.

 

The following image shows the tiering of inactive data through the auto-tiering policy in FabricPool.

Support for Microsoft Azure Blob Storage

Starting in ONTAP 9.4, FabricPool also supports tiering to Microsoft Azure. It’s done by attaching containers inside Microsoft Azure Blob Storage to aggregates that act as an external capacity tier for your on-premises all-flash system.   Both hot and cool Azure storage tiers are supported, so you can customize pricing schemes to your specific needs. For example, you could use the Azure Blob hot storage tier to for data that is more likely to be accessed and the cool storage tier for data that is less likely to be needed.

 

Providing data mobility across multiple clouds—both public and private— is as easy as moving a volume from one aggregate to another. FabricPool now supports tiering data to and from:

 

  • Amazon S3 (Standard and Standard–Infrequent Access)
    • Includes AWS GovCloud
  • Microsoft Azure Blob Storage (Hot and Cool)
  • StorageGRID Webscale

Support for ONTAP Select

ONTAP 9.4 also adds support for ONTAP Select. Now you can automatically tier data to low-cost clouds anywhere you can use ONTAP. For example:

  • ONTAP in traditional data centers. Create FabricPool aggregates on supported AFF and FAS systems by using solid-state drive (SSD) aggregates.
  • ONTAP Cloud. Tier to Amazon S3 and Azure Blob to reduce storage costs.
  • ONTAP Select. Reduce storage costs from anywhere you use ONTAP Select—even NetApp HCI.

Inactive Data Reporting

Most storage administrators know that they have a ton of inactive (cold) data taking up valuable capacity in their all-flash environments. What they don’t know is exactly how much. It’s an important data point, especially for FabricPool: How much money can I save if I tier this data to somewhere less expensive?

 

First available in ONTAP 9.4, inactive data reporting (IDR) is an excellent tool for determining the amount of inactive (cold) data that can be tiered from an aggregate. IDR is enabled by default on FabricPool aggregates. More important, you can enable it on non-FabricPool aggregates by using the ONTAP CLI, to get visibility on how much you can gain by deploying FabricPool.

 

storage aggregate modify -aggregate <name>

-is-inactive-data-reporting-enabled true


Note:
The amount of cold data tiered depends on the tiering policies set on the aggregate volumes. This amount may differ from the amount of cold data detected by IDR, which uses a 31-day cooling period to determine what data is considered inactive.

 

IDR is displayed on the Storage Tiers page in NetApp OnCommand® System Manager.

To learn more about FabricPool, read TR-4598: FabricPool Best Practices and check out this short lightboard video.

 

mm

John Lantz

John Lantz is a Senior Technical Marketing Engineer at NetApp, who’s currently focused on FabricPool and object storage. His 15+ years of experience are built on a diverse background that includes technical writing, content marketing, and developer advocacy—with an emphasis on content-driven solutions. He's passionate about ONTAP and the development of smart and relevant collateral for technical audiences.