Storage Edit on GitHub
How ONTAP Cloud uses cloud storage
ONTAP Cloud uses AWS and Azure volumes as back-end storage. It sees these volumes as disks and groups them into one or more aggregates. Aggregates provide storage to one or more ONTAP Cloud volumes.
In AWS, an aggregate can contain up to 6 disks, with a maximum disk size of 16 TB. The underlying disk type can be a single EBS storage type or a tiered storage configuration that includes an EBS performance tier and an S3 capacity tier.
In Azure, an aggregate can contain up to 12 disks, with a maximum disk size of 4 TB. The underlying Azure managed disk type can be either Standard Storage (HDD) or Premium Storage (SSD).
You choose the disk type and the default disk size when you deploy an ONTAP Cloud system. For more details, refer to the following:
|The total amount of storage purchased from AWS or Azure is the raw capacity. The usable capacity is less because approximately 12 to 14 percent is overhead that is reserved for ONTAP Cloud use. For example, if Cloud Manager creates a 500 GB aggregate, the usable capacity is 442.94 GB.|
Data tiering in AWS
With ONTAP Cloud for AWS, you can reduce storage costs by using EBS as a performance tier for "hot" data and AWS S3 as a capacity tier for "cold" data.
Tiering data is a good choice if your data changes frequently or if you use ONTAP Cloud for disaster recovery or long-term retention:
After an aggregate has reached 50% capacity, ONTAP Cloud tiers Snapshot copies of read-write volumes to cost-effective S3 storage, freeing up the EBS performance tier for "hot" data.
When you replicate a volume for disaster recovery or long-term retention, data for the destination volume resides in the S3 capacity tier. If you activate the destination volume, the data gradually moves to the EBS performance tier as it is read.
Note the following about the EBS and S3 tiers:
An ONTAP Cloud system’s capacity limit is spread across EBS and S3 storage.
The EBS tier can be General Purpose SSDs or Throughput Optimized HDDs.
An ONTAP Cloud working environment uses a single S3 bucket for tiered data. This includes an HA working environment.
Cloud Manager creates the S3 bucket and names it fabric-pool-cluster unique identifier.
|You do not need to install a feature license to enable data tiering.|
Setting up a tiered storage configuration
You simply need to create a VPC Endpoint to S3 and then enable S3 tiering when creating a volume. For details, see Tiering data in AWS.