What is Self-Monitoring, Analysis and Reporting Technology (SMART)?

Self-Monitoring, Analysis and Reporting Technology (aka S.M.A.R.T.) is a feature on many Hard Disk Drives and Solid State Drives (SSD) to monitor the health of the drive with the purpose to forewarn of any failure.

The idea behind SMART is to anticipate a future failure by using statistics and reliability data reported back from the drive. By understanding there may be an imminent failure, the drive can be proactively replaced at the next scheduled system maintenance.

So, now for the tricky part. There is not an industry standard that every manufacturer adheres to. In fact, there are items that may be extremely predictive in a hard disk drive that are not as relevant in solid state drives. It's important to understand the answer to the question, how do solid state drives work?

Temperature is one of these items. A hard drive showing substantial increase in temperature could be signaling a forthcoming failure due to an increase in mechanical motor friction, whereas a solid state drive does not have these mechanical characteristics.

Hard drives do not have a write limitation, but SSD write endurance is finite. In the case of hard drives, the number of writes to a particular block is not relevant, whereas it is extremely relevant to a solid state drive.

It is important to review the literature related to your hard drive or solid state drive to understand exactly what is being reported when you read the SMART data.

As an example, available attributes from the SMART Read Data command for the Cactus 503 Series Industrial CompactFlash card is shown below.

Spare Block Count Attribute - Information about the number of available spare blocks. Can be monitored to see the remaining percentage of spare blocks available.

Spare Block Count Worst Chip Attribute - Info about the spare blocks in the flash chip with the lowest current number of spare blocks.

Erase Count Attribute - Info about the number of flash block erases performed. Can be monitored to see remaining erases available on the card.

Total ECC Error Attribute - Info regarding the total number of ECC errors on flash read commands.

Correctable ECC Error Attribute - Info regarding the total number of correctable ECC errors on flash read commands.

UDMA CRC Error Attribute - Info regarding the total number of UDMA CRC errors on flash read commands.

Total Number of Reads Attribute - Info regarding the total number of flash read commands.

Power On Count Attribute - Info on the total number of power on cycles.

Total LBAs Written Attribute - Info on the total amount of data written in units of 32MB.

Total LBAs Read Attribute - Info on the total amount of data read in units of 32MB.

Anchor Block Status Attribute - Info on the total number of times the Anchor block has been rewritten, either by a repair operation or by firmware updates.

TRIM Status Attribute - Information on the percentage of the card that is in trimmed state.

I hope this has been a useful overview of the SMART feature on SSDs. If you are having specific issues on your OEM application that you would like to address, please contact us.

Steve Larrivee has over 30 year's experience in the data storage market, including 5 years at Seagate Technology and 10 years at SanDisk. He joined Cactus Technologies Limited as an equity partner and Co-Founded Cactus USA in 2007 with partner Tom Aguillon. Learn more about Steve on Google+