This overview of SMART attributes in SSDs explains how organizations can put them to good use. SMART SSDs provide peace of mind about device life span.
Most SSD users understand that SSDs are subject to wear. However, few realize that SSD wear statistics and other important parameters are constantly measured by the SSD itself.
This information is at the fingertips of the user, thanks to the availability of SMART SSD attributes -- self-monitoring, analysis and reporting technology.
What SMART means
SMART attributes are a subset of the internal statistics that the SSD's internal controller maintains for management of its NAND flash chips. The controller manages the NAND flash wear through wear leveling and bad block management. It has its own detailed understanding of how much traffic the SSD has handled and how much more it can take before becoming problematic.
A useful consolidated subset of the data is available to interested parties through the SMART SSD attributes.
A brief history of SMART attributes in SSDs
SMART attributes started in HDDs as early as 1995, but their popularity increased significantly with the advent of the SSD. HDD statistics were difficult to understand unless admins had a thorough understanding of an HDD's internal workings and how they were likely to respond to various wear mechanisms. SSDs are somewhat simpler, since their wear mechanisms are relatively straightforward: Users can count writes and keep a tally of how many of the SSD's spare flash blocks have failed.
In the early days of widespread SSD use, which occurred shortly after NAND flash prices fell below those of DRAM in 2004, HDD makers and SSD makers were locked in battle. SSD makers criticized HDDs and questioned the endurance of mechanical systems in general. HDD makers responded fairly by pointing out incidents of SSD failures caused by the poor NAND flash management in use at that time.
To improve their sullied reputation, SSD makers worked together to improve their wear management algorithms and created more resilient SSDs. At the same time, many of them added SMART attributes to enable their customers to monitor the SSDs' wear themselves and understand how the SSD and their workloads interacted with one another.
The most important benefit of the SMART SSD is that it provides the data to monitor SSD wear, as long as users understand how to digest it.
Benefits of SMART SSD use
The most important benefit of the SMART SSD is that it provides the data to monitor SSD wear, as long as users understand how to digest it. This data is important for high-reliability applications, such as industrial control.
For example, once an organization installs a production robot arm, it may remain in use for decades. The SSD should not fail unpredictably in this kind of environment. With SMART attributes, users can set the controller software to interrogate the SSD every month to examine its health and then to report back to the operator well in advance if it's necessary to duplicate and replace the SSD.
The same is true in less demanding high-reliability environments. Transaction processing is a high-write load application, and high write loads accelerate wear. A data error can rapidly cause millions of dollars' worth of faulty transactions, so the SSD cannot fail during its usable life. SMART attributes enable the system to monitor its own health and alert the repair crew when an SSD is likely to fail, which enables them to replace it before it causes a catastrophe.
While these examples are extreme, SMART attributes can also give PC users greater peace of mind that their SSD is unlikely to suddenly fail any time soon. Some PC users say they have found that their SSD's SMART attributes have consoled them by indicating that the PC's SSD is unlikely to wear out within the next 100 years.
Shortfalls to SMART attributes
There are few standards governing SMART attributes, so each SSD tells its story differently from the next.
Vendors share their own list of attributes in their own favorite format. Some vendors provide different sets of attributes in different formats. They also provide attributes for different models within their own product offering. The user must examine these attributes with a good knowledge of how to read them for the exact SSD that the system uses.
For the data center, this is an acceptable situation, since the same attribute set may apply to tens of thousands of SSDs, in which case a little coding would pay back with solid feedback on a large squad of devices. PC users are likely to find this challenging since they need to find the appropriate decoder for whichever SSD is in their PC.
Considerations for use
Since SMART SSD information is buried deep within the hardware, the easiest way to access it is through a tool designed for attribute access. For example, Intel provides such tools, and Kingston offers a tool for its own SSDs.
There are many other freeware tools for SMART access, but SSD maker Crucial, which is owned by Micron, says that users can't trust all these monitors and provides a tool for its own SSDs. Other SSD vendors have similar tools.
Whatever the need, though, rest assured that there is a way to understand well in advance whether an SSD will last a few more days or years based upon its current workload. It may take some effort to put the right tools in place, but SMART attributes provide users with an understanding of the load that the SSD experiences and how long it will last under those conditions.
