Self-Monitoring, Analysis and Reporting Technology
also called SMART or S.M.A.R.T.
You should do a SMART test on all your HDDs and SSDs before you put any data on it and save the values somewhere for later comparison.
Interpretation
| 5 | Reallocated sector count | The sectors which are reallocated to another place on the disk. On HDDs, this might mean that they fail very soon. |
| 192 | Power-off retract count | Instances where the drive didn’t properly shut down or lost power unexpectedly. Happens when you "pull the plug", may also happen when you hot plug HDDs which aren't hot-pluggable. |
| 197 | Currently pending sectors | The sectors which are reallocated due to errors reading them. A sign for a failing HDD, whereas reallocation in SSDs is common as some cells fail earlier than others. |
HDDs
Hard disk drives usually report pending sectors or reallocated sectors before they fail. As soon as you see the number go up, you should back up the most important files first. Monitor SMART data while you do this.
| 3 | Spin-up time | The time in milliseconds the drive took to bring the platters up to the correct speed (usualls 5200, 5400 or 7200 1/min). If the number is more than 500 higher than where you got the drive, the motor might be failing. |
SSDs
Don't trust any metric as indicator for failure on SSDs as they can die spontaneously, even more so than HDDs can.
| 241 | Lifetime Writes in GiB | The amount of data in GiB written to the drives. If that's higher than the TBW specification for the SSD, failure might be imminent. |
Windows
| Crystal Disk Info | graphical application to view SMART data |
Linux
| gsmartcontrol | graphical application to view SMART data |
| smartctl | CLI tool from the smartmontools suite |
[1]
Total Bytes Written or TeraBytes Written, found in the spec sheet of the drive, usually between 100TiB and 300TiB