hardware problem

kate · 12-05-2006, 02:50 AM

guys, besides the /var/log/messages are there any where else we can get a description on a certain part of hardware failure in the system from the OS?

btmiller · 12-05-2006, 06:06 PM

Depends on the type of hardware. For hards drives there's smartctl if the drive is S.M.A.R.T. capable. You might be able to run stress tests (e.g. Prime95) to determine if a component is bad. In general, however, the kernel doesn't have any ways of determining hardware health, it just sees the symptoms when things go wrong.

hollywoodb · 12-05-2006, 06:51 PM

Most distributions also have /var/log/dmesg, and if not, just use the `dmesg` command from console or terminal.

matthewg42 · 12-05-2006, 07:01 PM

There are some utility programs for this:
lshw, lspci spring to mind.

kate · 12-07-2006, 09:56 PM

Thanks guys, but my system doesn't have the smartctl command, plus, i already looked into the dmesg, it doesn't show anything wrong...
Funny thing is the problem looks like an I/O error. which is tied to a particular disk section.. but the dmesg has no errors at all on this.

btmiller · 12-07-2006, 10:27 PM

Try booting into single user mode (or off a Knoppix disk, which is the safest way) and run badblocks on your hard drive. if there is a bad sector, it should detect it. Don't run badblocks while your system is fully booted up and file systems are mounted read/write.

You should also fsck the partition in question (fsck /dev/whatever) after finishing the badblocks scan,

kate · 12-11-2006, 11:50 PM

Thanks BT! appreciate it!