Apparently, it's not to be trusted; see Trends in a Large Disk Drive Population (Pinheiro, Weber and Barroso 2007).
Are my disks failing?
The one-line executive summary is available for supported disks via the -Hc option. grep for "^SMART overall" and verify PASSED as opposed to FAILING. If you're already FAILED, the command will generally fail in and of itself.
Hardware RAID Controllers
- Dell PERC -- See Linux on Dells. Example:
[wopr](0) $ lsscsi -l [0:0:32:0] enclosu DP BACKPLANE 1.06 - state=running queue_depth=128 scsi_level=6 type=13 device_blocked=0 timeout=0 [0:2:0:0] disk DELL PERC 6/i 1.11 /dev/sda state=running queue_depth=128 scsi_level=6 type=0 device_blocked=0 timeout=90 [wopr](0) $
- "Monitoring Hard Disks with SMART", Linux Journal 2004-01-01
- SMART on SATA drives is covered in more detail on the SATA page.