SSD Endurance: Data Integrity is Marathon, Not a Sprint
The tech industry is rightly obsessed with analyzing solid state drive (SSD) performance. As the modern go-to technology for Tier 0 enterprise storage, SSDs house business’s most critical data and are core to generating revenue. Split-second access to this data is important, but if that data should become corrupted, the most rapid access in the world won’t matter.
“Nobody wants to see their data unintentionally altered, but it’s going to have a more significant impact if this critical data is in a large enterprise,” says Teresa Worth, senior product marketing manager for enterprise SSDs at Seagate. “From a reliability perspective, you need to make sure that your storage systems are going to work as promised and not crash. That’s part of the problem with client vs. enterprise devices. Client devices can’t hold up under demanding enterprise workloads. They weren’t designed for those environments.”
All quips about consumer SSD door stops aside, the paramount fact is that enterprises must honor their service level agreements (SLAs), and the more “nines” are involved the more pressure there is to avoid any point of failure. Downtime can literally cost thousands or, in the worst cases, millions of dollars per minute. Imagine Amazon.com going dark on Cyber Monday. (A recent study by The Ponemon Institute for Emerson Network Power found that the average downtime event for data centers lasted 90 minutes and cost over $500,000.) Downtime and data loss can impact everything from a company’s financials to its lasting reputation.
Today’s enterprises need the utmost in storage reliability. In the world of SSDs, this means devices with high “endurance.” At a micro level, endurance denotes the ability of a flash memory block to endure a given number of write cycles. More broadly, endurance implies a storage device operating as promised for the duration of its expected life. This includes not only operating throughout its warranty but operating at optimal performance levels throughout that period.
The last year has seen a rising tide of marketing and ensuing press coverage stating that consumer drives are becoming just as strong on endurance as their enterprise counterparts. If true, this would essentially erase the line dividing the two groups. However, there’s only one problem: it’s not true. If you want to keep your organization’s data safe and obtain the return on investment you expect from your SSD purchases, read on and learn why bona fide enterprise drives are still the only smart choice for today’s data centers.
The Roots of Wear and Tear
At the most basic level, the NAND flash memory used in SSDs is comprised of millions of cells. Each cell is comprised of a control gate resting above a floating gate. This structure sits above a P-type silicon substrate. The gates are separated from the silicon by a thin oxide layer. To program a cell, electrons are sent from the substrate, through the oxide, and into the floating gate. Erasure reverses this process.
The reason why flash memory doesn’t lose its data when you remove system power is due to the strength of that oxide insulator layer. In fact, when new, most flash cells can retain data in the absence of external power for over a decade. When new, the oxide layer is thick (on an atomic scale) in order to hold in the electrons securely over time. However, as flash cells are used, the retention time shortens.