Google Cloud NetApp Volumes is going large and adding cost-optimization

We’re excited to announce the general availability of two features that many of our customers have been asking for: petabyte-scale volumes and auto-tiering. These features have been in preview and are now generally available for users of Premium and Extreme service levels across 15 regions supported for production workloads. 

Google Cloud NetApp Volumes is a fully managed, secure, and performant file storage service that supports SAP, Microsoft, and Linux-based applications. It has added support for large-capacity volumes that provides up to 1PiB in storage capacity and up to 12.5GiB/s of throughput in a single volume. With large volumes, you can quickly and easily deploy a variety of workloads that need to store and process petabyte-scale datasets, including HPC, EDA, AI/ML training, and content repositories with high throughput and concurrent access.  

Large volumes are now generally available in all 15 regions where Premium and Extreme service levels are offered. With large volumes, customers can provision a single volume starting at 15TiB that can be scaled up to 1PiB dynamically in increments of 1GiB. The performance of the volume scales linearly with provisioned capacity and the service level selected up to a maximum of 1 12.5GiB/s. Additionally, large-capacity volumes running on the Extreme service level can use multiple endpoints (IP addresses) to load-balance traffic and achieve higher performance, ideal for workloads that require highly concurrent access to a large dataset.  

Large volumes allow customers to bring petabyte-scale datasets used in EDA workloads, AI applications, and content data repositories to benefit from data management features like NetApp® Snapshot™ copies, clones, and cross-region replication on a single volume. Before the availability of large volumes, customers had to select performance- or capacity-optimized options for workloads with petabyte-scale datasets or run their workloads by partitioning their data across multiple volumes.  

For more information about large volumes, see the product documentation.  

When enabled for a Premium or Extreme volume, auto-tiering identifies and transparently moves infrequently used data from existing high-performance hot tiers to a managed cold tier at a significantly lower cost with a lower performance profile. Infrequently accessed data is determined by a configurable cooling policy. Cold tier storage can cost up to 80% less than hot tier, resulting in overall savings of 30% to 50%, depending on the percentage of cold data stored. Auto-tiering can be enabled on a per-volume basis for Premium and Extreme service levels, automatically tiering infrequently accessed data to a fully managed lower-cost storage tier in Google Cloud without requiring any changes to applications or requiring third-party tooling.  

For Premium and Extreme service levels, auto-tiering identifies and moves infrequently accessed data to low-cost cold storage. It’s integrated into Google Cloud NetApp Volumes to provide uniform file access via SMB or NFS for both hot and cold data. When enabled, the cold tier storage is automatically set up and included. Storage pools can contain a mix of volumes with auto-tiering enabled on a per-volume basis. 

With auto-tiering, customers can automatically tier off infrequently accessed data to low-cost storage without manual intervention, changes to applications, or changes to how data is accessed. Types of data that can benefit from auto-tiering include: 

• Infrequently accessed files. Large data repositories that users seldom access. Auto-tiering automatically tiers data off to cold storage until it is requested. 

• Short-lived data cycles. Shared files are typically active for only a month and are then rarely accessed. Auto-tiering moves data to a cold tier after a month of no access. 

• Large dataset retention. EDA and AI training datasets have a 6-month active time, but they must be retained. Auto-tiering moves them to cold tier after 6 months of inactivity. 

• Backup or archival storage. Historical backup and archival data need immediate retrieval at a lower cost. Auto-tiering moves data to low-cost tier after a set period of time. 

For more information about auto-tiering, see the product documentation. 

We are enthusiastic about all the new features we’ve introduced since Google Cloud NetApp Volumes was introduced last year, and we’ll continue to keep up the momentum. To find out more, schedule a personalized 1:1 demo or visit the Google Cloud NetApp Volumes web page. 

This blog is co-authored with Rukmani Gopalan, Group Product Manager, Google Cloud NetApp Volumes, Google Cloud.