Tiering inactive data to low-cost object storage
Contributors Download PDF of this page
You can reduce storage costs for Cloud Volumes ONTAP by combining an SSD or HDD performance tier for hot data with an object storage capacity tier for inactive data. Data tiering is powered by FabricPool technology. For a high-level overview, see Data tiering overview.
To set up data tiering, you need to do the following:
Choose a supported configuration
Most configurations are supported. If you have a Cloud Volumes ONTAP Standard, Premium, or BYOL system running the most recent version, then you should be good to go. Learn more.
Ensure connectivity between Cloud Volumes ONTAP and object storage
Ensure that you have an aggregate with tiering enabled
Data tiering must be enabled on an aggregate in order to enable data tiering on a volume. You should be aware of the requirements for new volumes and for existing volumes. Learn more.
Choose a tiering policy when creating, modifying, or replicating a volume
Cloud Manager prompts you to choose a tiering policy when you create, modify, or replicate a volume.
What’s not required for data tiering?
Configurations that support data tiering
You can enable data tiering when using specific configurations and features:
Data tiering is supported with Cloud Volumes ONTAP Standard, Premium, and BYOL, starting with the following versions:
Version 9.2 in AWS
Version 9.4 in Azure with single node systems
Version 9.6 in Azure with HA pairs
Version 9.6 in GCP
Data tiering is not supported in Azure with the DS3_v2 virtual machine type.
In AWS, the performance tier can be General Purpose SSDs, Provisioned IOPS SSDs, or Throughput Optimized HDDs.
In Azure, the performance tier can be Premium SSD managed disks, Standard SSD managed disks, or Standard HDD managed disks.
In GCP, the performance tier can be either SSDs or HDDs (standard disks).
Data tiering is supported with encryption technologies.
Thin provisioning must be enabled on volumes.
Depending on your cloud provider, certain connections and permissions must be set up so that Cloud Volumes ONTAP can tier cold data to object storage.
Requirements to tier cold data to AWS S3
Ensure that Cloud Volumes ONTAP has a connection to S3. The best way to provide that connection is by creating a VPC Endpoint to the S3 service. For instructions, see AWS Documentation: Creating a Gateway Endpoint.
When you create the VPC Endpoint, be sure to select the region, VPC, and route table that corresponds to the Cloud Volumes ONTAP instance. You must also modify the security group to add an outbound HTTPS rule that enables traffic to the S3 endpoint. Otherwise, Cloud Volumes ONTAP cannot connect to the S3 service.
If you experience any issues, see AWS Support Knowledge Center: Why can’t I connect to an S3 bucket using a gateway VPC endpoint?.
Requirements to tier cold data to Azure Blob storage
You don’t need to set up a connection between the performance tier and the capacity tier as long as Cloud Manager has the required permissions. Cloud Manager enables a VNet service endpoint for you if the Cloud Manager policy has these permissions:
The permissions are included in the latest Cloud Manager policy.
Requirements to tier cold data to a Google Cloud Storage bucket
The subnet in which Cloud Volumes ONTAP resides must be configured for Private Google Access. For instructions, refer to Google Cloud Documentation: Configuring Private Google Access.
You need a service account that has the predefined Storage Admin role. You’ll need to select this service account when you create a Cloud Volumes ONTAP working environment.
Assign the predefined Storage Admin role to the tiering service account.
Add the Connector service account as a Service Account User to the tiering service account.
You can provide the user role in step 3 of the wizard when you create the tiering service account, or grant the role after the service account was created.
You’ll need to select the tiering service account later when you create a Cloud Volumes ONTAP working environment.
If you don’t enable data tiering and select a service account when you create the Cloud Volumes ONTAP system, then you’ll need to turn off the system and add the service account to Cloud Volumes ONTAP from the GCP console.
Enabling data tiering after implementing the requirements
Cloud Manager creates an object store for cold data when the system is created, as long as there are no connectivity or permissions issues. If you didn’t implement the requirements listed above until after you created the system, then you’ll need to manually enable tiering, which creates the object store.
On the Working Environments page, double-click the name of the Cloud Volumes ONTAP instance.
Click the menu icon and select Enable capacity tiering.
You’ll only see this option if data tiering couldn’t be enabled when Cloud Manager created the system.
Click Enable so Cloud Manager can create the object store that this Cloud Volumes ONTAP system will use for tiered data.
Ensuring that tiering is enabled on aggregates
Data tiering must be enabled on an aggregate in order to enable data tiering on a volume. You should be aware of the requirements for new volumes and for existing volumes.
If you’re enabling data tiering on a new volume, then you don’t need to worry about enabling data tiering on an aggregate. Cloud Manager creates the volume on an existing aggregate that has tiering enabled, or it creates a new aggregate for the volume if a data tiering-enabled aggregate doesn’t already exist.
If you want to enable data tiering on an existing volume, then you’ll need to ensure that data tiering is enabled on the underlying aggregate. If data tiering isn’t enabled on the existing aggregate, then you’ll need to use System Manager to attach an existing aggregate to the object store.
Open the working environment in Cloud Manager.
Click the menu icon, click Advanced, and then click Advanced allocation.
Verify whether tiering is enabled or disabled on the aggregate.
In System Manager, click Storage > Tiers.
Click the action menu for the aggregate and select Attach Cloud Tiers.
Select the cloud tier to attach and click Save.
You can now enable data tiering on new and existing volumes, as explained in the next section.
Tiering data from read-write volumes
Cloud Volumes ONTAP can tier inactive data on read-write volumes to cost-effective object storage, freeing up the performance tier for hot data.
In the working environment, create a new volume or change the tier of an existing volume:
Create a new volume
Click Add New Volume.
Modify an existing volume
Select the volume and click Change Disk Type & Tiering Policy.
Select a tiering policy.
For a description of these policies, see Data tiering overview.
Cloud Manager creates a new aggregate for the volume if a data tiering-enabled aggregate does not already exist.
Tiering data from data protection volumes
Cloud Volumes ONTAP can tier data from a data protection volume to a capacity tier. If you activate the destination volume, the data gradually moves to the performance tier as it is read.
On the Working Environments page, select the working environment that contains the source volume, and then drag it to the working environment to which you want to replicate the volume.
Follow the prompts until you reach the tiering page and enable data tiering to object storage.
For help with replicating data, see Replicating data to and from the cloud.
Changing the storage class for tiered data
After you deploy Cloud Volumes ONTAP, you can reduce your storage costs by changing the storage class for inactive data that hasn’t been accessed for 30 days. The access costs are higher if you do access the data, so you must take that into consideration before you change the storage class.
The storage class for tiered data is system wide—it’s not per volume.
For information about supported storage classes, see Data tiering overview.
From the working environment, click the menu icon and then click Storage Classes or Blob Storage Tiering.
Choose a storage class and then click Save.