6

I would like to use SMART to perform a health check of my hard disk. I have an HP server of 4 years ago I believe. Every time I try to use the following command

 smartctl -a /dev/sda

I got an error message saying the following

 Device does not support SMART

 Error Counter logging not supported
 Device does not support Self Test logging

Does it really mean that there is no way I can use SMART on these disk, or is it just because I am missing some driver or my BIOS is too old?

According to dmidecode I have bios 2.7 so it should be not too old.

I am using CentOS 6.4.

6

The most likely reason is that your server came with a hardware raid controller and sda is not an individual disk, but a logical drive.

HP provides tools to monitor your hardware, the typical Linux utility to monitor HP raid controller status and the drives contained would be hpacucli found here.

Some useful commands would be:

hpacucli> ctrl all show config 
hpacucli> ctrl all show config detail
hpacucli> ctrl all show status

hpacucli> ctrl slot=0 pd all show 
hpacucli> ctrl slot=0 pd 1 show
hpacucli> ctrl slot=0 pd all show status
  • Yes. There is RAID. Why this RAID makes sMART complicated to use? I read about this difficulty elsewhere. Is this going to complicate and make impossible to use SMART or it is really just a matter of using the right tool? – Abruzzo Forte e Gentile Nov 28 '13 at 12:13
  • 1
    SMART is a set of disk properties, a logical drive consists of multiple disks, so that doesn't align. The RAID controller could interpret the status of the underlying disks and present a SMART like interface, but typically the RAID driver and/or manufacterer tooling provides better or more relevant status information. – HBruijn Nov 28 '13 at 12:43
  • 1
    @HBruijn This is correct. I often say that S.M.A.R.T. is not the final word in storage monitoring. In the HP Smart Array case, the controllers use SMART amongst a number of other methods to determine drive and array health. – ewwhite Jun 29 '14 at 21:07
5

To add to what Aleksandr said (I can't comment yet, unfortunately):

smartctl -x -A -d sat+cciss,0 /dev/sda

The sat+ bit gets you more than just cciss on its own - in particular, you should be able to access the individual attributes, not just the overall SMART status and computed values.

2

Device does not support SMART

This can mean one of two things:

  1. The named hard disk indeed does not support SMART. In that case, you are out of luck with that particular drive.

  2. smartctl has no clue how to talk to the disk. If that is the case, you may have a little bit of luck with carefully trying various parameter values for --device; see the smartctl man page for details.

BIOS version is completely unrelated to whether the disk drive supports SMART or not. Whether a drive supports SMART is a property of the drive, or more accurately the drive firmware.

2

You can also use repos as write here - http://sysadmin.te.ua/linux/hpraid-monitoring.html

echo 'deb http://hwraid.le-vert.net/ubuntu precise main' > /etc/apt/sources.list.d/raid.list

This repo has almost all raid monitoring utils for all type of raids! Also article shows how to monitor HP raid using 'smartctl' as well.

smartctl -a -d cciss,0 /dev/sda

Regards.

  • actually this should be the accepted answer – kiltek Mar 3 '17 at 13:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.