Double IOPS using Seagate Exos with MACH.2 Dual-Acutator Technology
In a hyper-converged world, data is the most expensive currency as cloud customers deploy large fleets of 3.5-inch HDDs in their data centers and manage their application storage nodes across multiple devices, spread across multiple racks. In order to meet their Service Level Agreements (SLAs), these cloud customers need to achieve a certain level of performance, measured in IOPS (I/ per second) and command latency, across the deployed HDD storage capacity. This required performance is characterized as IOPS-per-terabyte (IOPS/TB) at a particular latency.
Seagate’s new 2X14 enterprise hard drive is the first to integrate Seagate’s groundbreaking MACH.2™ multi-actuator technology, and is being tested by customers now for data center deployments. MACH.2 has demonstrated its benefits over drives for many applications including content delivery networks (CDNs), video streaming, mail servers, backup/shuttle services, Hadoop, cloud applications, and more.
MACH.2 solves cloud data center needs like IOPS-per-terabyte and cost-per-terabyte
Data center architects also look at enterprise SSDs as one way to provide high IOPS. That can make sense depending on the balance of various needs for each application including capacity requirements, IOPS, throughput, environmental costs like cooling, and cost. But with their lower cost-per-terabyte and higher capacities, hard drives provide the most effective storage tier for leveraging capacity — if the HDDs can provide the necessary IOPS.
Seagate MACH.2 technology addresses the IOPS/TB challenge by using two actuators that can transfer IOs independent of each other within a single HDD, creating parallelism within a single hard drive that enables up to double the performance. Within a drive, the top half of the read/write heads are addressed by one actuator, while the bottom half of the heads are addressed by a second actuator. Each actuator addresses one half the total capacity of the drive.
MACH.2 prevents "stranded capacity"
a particular HDD device meets or exceeds the threshold IOPS/TB as required by the customer, all of the capacity on the HDD device can be utilized. If the IOPS/TB on an HDD drops below a defined threshold based on application workloads, the customer cannot utilize the drive capacity with the targeted workload — a problem known as “stranded capacity.”
James Borden, cloud product manager at Seagate, explains the dilemma posed by stranded capacity.
“As hard drives get increasingly higher in capacity, up until now the ability to transfer data faster has not advanced commensurately,” says Borden. “Without multi-actuator technology, once you exceed a certain you’re no longer able to deliver the expected quality of service, because you’re limited by each drive’s maximum throughput.”
“For example, in order to maintain the expected quality of service, a 16TB drive requires twice the maximum throughput capability of an 8TB drive. As your application starts to exceed the maximum IOPS capability of a storage device, the only way to solve that is to put less data on the device — and that equates to stranded capacity.”
“One partial solution to the wasted capacity is to use a ‘short stroke’ strategy,” explains Borden. “You put a subset of your data on the outer diameter of a use the internal diameter to store less important data that doesn’t need to be delivered with the same quality-of-service metric. In that way, at least you won’t have a hard drive that’s only two-thirds full. But short stroking still makes data management more complex, and you can’t be certain you’ll always have full utilization of your installed capacity.”
But when deploying 2X14 drives with MACH.2, short stroking is no longer necessary. A data center can install a 14TB dual-actuator drive in each slot. To the application, it looks very similar to the high throughput provided by two 7TB drives — but it fits into a single slot.
MACH.2 technology provides a proven solution to accelerate data transfer while reducing overall command latency. As more and more applications encounter IOPS/TB constraints, MACH.2 technology will resolve performance constraints, reduce TCO, and help customers meet the growing demand for increased performance as defined in SLAs.