SNIA Releases Enterprise SSD Testing Spec

The Storage Networking Industry Association (SNIA) released on Monday a new specifications designed to measure the performance characteristics of solid-state products.

The Solid State Storage Performance Test Specification (SSS PTS) was developed by the SNIA Solid State Storage Technical Work Group as a means to benchmark performance in new solid-state storage (SSS) and solid-state drives (SSDs). SNIA is an industry association comprised of trade vendors and universities.

Until now, there has been no widely accepted industry standard test methodology or test environment for measuring solid-state storage device performance. This created conflicting, vendor-specific performance measurements that didn't accurately indicate the performance state, read/write mix, or data transfer size, according to SNIA.

"The issue was that people, without doing their own tests and coming out with their own form of tests, couldn't compare performance using the numbers from data sheets," said Paul Wassenberg, chair of the solid-state storage initiative, in an interview with InformationWeek. "And when you use a HDD benchmark for SSDs, you get unrealistic results. So that led to the creation of the SSS PTS."

To avoid confusion, the SNIA will release two versions of the test specification: the enterprise SSDs spec this week and another for client SSDs, to be released in the third quarter of this year. The Enterprise Performance Test Specification) defines a series of device tests and methodologies for comparable testing of SSD devices in enterprise systems, such as storage arrays.

"Bringing clarity to an otherwise confusing marketplace, the SSS PTS will lead to greater confidence in solid state storage solutions and ultimately higher adoption rates of this very beneficial technology," according to a white paper by SNIA.

The SSS PTS was developed by representatives of more than 40 SSD and storage system companies including Samsung, Intel, Toshiba, IBM, Seagate, EMC, Hitachi, and Western Digital.

One of the primary issues with SSDs is that performance decreases after initial use. Once data is written to SSDs, the processor functions more slowly as data is moved around through reading, writing, and erasing data. This slows performance down relative to its Fresh Out of Box State (FOB)--a term coined by the SNIA.

After a certain amount of use, the SSD settles into what SNIA calls the "steady state," or a state where performance can be measured and standardized. Certain details dictate the behavior of a device including the cost of the device, amount of NAND Flash over-provisioning, intended workload/use case, and the device write history dictate what type of Steady State behavior a Flash SSS device will exhibit.

"The duration of the FOB can vary considerably depending on the drive design," Wassenberg said. "We have a way of preparing the drive by purging it, preconditioning it, and taking measurements only when the drive is in a steady state, a very consistent repeatable level of performance, a mode that is carefully defined within the SSS PTS."

The following is the PTS test sequence:

-- Purge: Put the SSD into a near-FOB state by erasing data

-- Precondition: Put the SSD into a steady state

-- Test: Take measurements when the SSD is in a steady state

The PTS identifies and recommends test platforms and software tools for this testing but does not require them in this iteration of the SSS PTS. Any tool capable of meeting the requirements of the SSS PTS may be used. Calypso, IOmeter, and Vdbench are all cited as having tools necessary to conduct SSS PTS tests.

IT teams areas are packing more information on fewer devices, delivering faster throughput while using less space and power, and managing the needs of more applications with fewer people. Our new report shows how smart CIOs will accelerate this trend by adopting new multipurpose arrays and converged networks. Download our report here. (Free registration required.)