GCP – A radically simpler way to plan and manage block storage performance

Earlier this year we announced the general availability of Hyperdisk Storage Pools with Advanced Capacity, that helps you simplify management and lower the Total Cost of Ownership (TCO) of your block storage capacity. Today we are bringing that same innovation to block storage performance through Hyperdisk Storage Pools with Advanced Performance. You can now provision IOPS and throughput in aggregate which Hyperdisk Storage Pools will dynamically allocate as your applications read and write data, allowing you to increase resource utilization and radically simplify performance planning and management. 

You can get started with Advanced Performance Storage Pools in the Google Cloud console today.  Advanced Performance Storage Pools are part of the Hyperdisk family of block storage offerings from Google Cloud, which you can learn more about here. 

The challenge: Provisioning the right amount of performance resources 

Customers have told us that they struggle to balance utilizing all of their block storage performance resources while ensuring their workloads have the resources they need to succeed. The root cause of this challenge is what is often called the “sum of peaks” problem. In order to ensure their workloads are never starved of performance, customers will provision their block storage for the highest performance they’ve observed, however most of the time their disks consume significantly less than that performance. This means that your disks are at chronic low performance utilization. We’ve visualized this phenomenon below.

Achieve 40-50% lower TCO with Advanced Performance Storage Pools

We built Advanced Performance for Hyperdisk Storage Pools to solve this problem. With Advanced Performance Storage Pools you no longer have to choose between ensuring your workloads are successful and driving high performance utilization. In a Hyperdisk Storage Pool with Advanced Performance, you provision performance in aggregate and the Storage Pool intelligently distributes those resources across the disks in the Pool, as needed. All without changing how disks behave, so your applications can run as normal. You can now simplify planning for your performance needs and lower your TCO, without sacrificing application success or resource efficiency.

Let’s look at a few workloads to see how Hyperdisk Storage Pools with Advanced Performance can lower your TCO. Take two workloads, an enterprise application suite that requires 25K IOPS at peak (e.g. when all users are logged in simultaneously) but only 10K at steady state, and a database workload that requires 75K IOPS at peak (e.g. when quarterly reports are due) but only 35K at steady state. Outside of a Pool these workloads would operate at 40-45% performance utilization as their steady state performance would be roughly half of their provisioned performance. However,the dynamic resource allocation of a Hyperdisk Storage Pool with Advanced Performance ensures these workloads run at roughly 80% utilization, enabling the customer to provision significantly lower performance resources and lower their TCO from 40-55% – all without changing their workloads. 

Simplify and plan for block-storage performance

Advanced Performance Storage Pools is a fundamental innovation in how you provision your performance resources. Historically, infrastructure managers are forced to choose between optimizing for performance and accepting low utilization, or optimizing for efficiency and risking application failure. This means that they need to actively manage the performance of each volume and accept high management toil. 

Hyperdisk Advanced Performance Storage Pools makes block storage performance management simple. Disk IOPS and Throughput in an Advanced Performance Storage Pool are “thin-provisioned”. In other words, provisioning performance at the disk level does not consume any resources and you can provision up to 5x the IOPS and throughput in a Pool to the disks in the Pool. This makes it easy to plan deployments (as you no longer need to precisely guess the performance needs of each disk) and achieve high efficiency (as you no longer need to actively manage the IOPS and Throughput allocated to each disk). You can simply create the disks you need, with the highest performance they will likely require, and let the Advanced Performance Storage Pool allocate resources when they are needed. 

Early Advanced Performance Storage Pools customers such as REWE have seen the value of the offering.

“At REWE, it’s a huge pain point for us to have to predict the performance spikes of our applications when planning deployments, and it’s even harder to actively manage performance utilization of storage resources. With Advanced Performance Storage Pools, we can stop worrying about both. We are excited about the savings in TCO, and toil, that Advanced Performance Pools will bring to our infrastructure.” – Jan Rundshagen, Infrastructure Manager, REWE

Get started

You can get started with Hyperdisk Advanced Performance Storage Pools in the same workflow as creating Hyperdisk Storage Pools. Log into the Google Cloud console, navigate to Compute Engine, and look under Storage. Then create your Storage Pool, select its volume type (Balanced or Throughput), Advanced Performance, and set the total amount of capacity and performance you will need in the pool. You can begin by creating new Hyperdisk volumes in the pool that not only share the pool’s capacity as with existing Storage Pools, but also have the added benefit of dynamically sharing performance across your resources.

We hope that Advanced Performance Storage Pools radically improve the way you manage and achieve the performance efficiency of your applications. Hyperdisk Advanced Performance Storage Pools are rolling out to regions and zones now. Get started with Advanced Performance Storage Pools in the Google Cloud console today or learn more in our documentation about how to create, use and manage your Advanced Performance Storage Pools.