Is there a S.M.A.R.T. for Flash Drives?

There's a technology called S.M.A.R.T. that tells you when a HDD is about to go bad. Is there a similar technology for SSD/Thumbdrives?


Solution 1:

SSD drives do support SMART, according to this ServerFault post.

USB thumbdrives may use cheaper memory controllers that do not support SMART, however. It depends largely on the manufacturer and their component choices.

Solution 2:

SMART's a firmware trick, so it ultimately depends on whether or not the person writing the firmware was paid enough to implement it. Odds are good they were not.

Solution 3:

As quack quixote notes SSDs do implement SMART. However, it may not be as useful as it oft claimed. labs.google.com published a white paper about drive failures a while back. The take away for many: SMART isn't all that helpful for individual drives.

From the paper (emphasis added):

We conclude that it is unlikely that SMART data alone can be effectively used to build models that predict failures of individual drives. SMART parameters still appear to be useful in reasoning about the aggregate reliability of large disk populations, which is still very important for logistics and supply-chain planning.

And

Our results confirm the findings of previous smaller population studies that suggest that some of the SMART parameters are well-correlated with higher failure probabilities. We find, for example, that after their first scan error, drives are 39 times more likely to fail within 60 days than drives with no such errors. First errors in reallocations, offline reallocations, and probational counts are also strongly correlated to higher failure probabilities. Despite those strong correlations, we find that failure prediction models based on SMART parameters alone are likely to be severely limited in their prediction accuracy, given that a large fraction of our failed drives have shown no SMART error signals whatsoever. This result suggests that SMART models are more useful in predicting trends for large aggregate populations than for individual components. It also suggests that powerful predictive models need to make use of signals beyond those provided by SMART.

While this may not hold true for SSDs, it's something to keep in mind.