The Device Id shown here is your N .
A hardware RAID controller manages its own disks and typically only presents a "Virtual Disk" to the operating system. When you run smartctl -a /dev/sda , you are asking for the health of the virtual drive, which doesn't have SMART data. To get the health of the physical drives behind the controller, you must use the -d (device type) flag to bypass the virtual layer. Step 1: Identify Your Drive IDs (N)
The key is always: with -d megaraid,N .
smartctl -a /dev/sg0 --device=megaraid,2
smartctl & megaraid: how to find the right device node for an adapter # The Device Id shown here is your N
| Controller | Syntax | |------------|--------| | Dell PERC (MegaRAID) | -d megaraid,N | | LSI MegaRAID | -d megaraid,N | | HP SmartArray | -d cciss,N | | 3ware | -d 3ware,N /dev/twa0 |
# For most modern PERC cards: smartctl -a /dev/sda --device=megaraid,2 To get the health of the physical drives
Dell PowerEdge servers use cards (e.g., PERC H310, H330, H730p, H740p). Almost all PERC controllers from the 10th generation onward are based on LSI MegaRAID chipsets.
If a physical drive is not part of a Virtual Disk, /dev/sda cannot access it. You must use the device ( /dev/sgX ). Almost all PERC controllers from the 10th generation