Hi,

i currently have a few disks i want to check for overall health.
I know of the tools smartctl and badblocks. How do they compare? And if i really want to be sure which do i use?
Or is there even a more simple/quick way to do this?

JJ

Smartctl is an interface for the "SMART" self-diagnostic system built into most hard drives. badblocks is an external program that tests the drive for "bad blocks". Is there any reason not to use both?

I think the simplest way to get information about the overall health of a disk is get reports from S.M.A.R.T. This technology can return status about the health of the disk and warning you about an imminent problem before it turns itself in a real problem.
It can test your disk in background and don't depends on your O.S. or even on the filesystem or partition table you have on disk.

On *nix systems, the smartd daemon can collect info about S.M.A.R.T. status and log any relevant messages to your system log. You just need to inspect this file in a regular basis using a log report tool or whatever.

badblocks on the other hand is your last resource for disk/baseboards not S.M.A.R.T. compatible. badblocks won't do anything else than listing the bad blocks it found on the disk. Is up to you to repair the file system and reserve the bad blocks from allocation and it could be tricky and risky task in some situations. It is better to backup the whole affected partition and create a new filesystem on it using the check flags (for ext2/3 it is -c) and after that, restore the backup.

smartctl uses the built-in diagnostic capabilities of modern HD. Your HD can have a microprocessor inside it. To get a status of the disk and see if it's smart-enabled,

smartctl --all /dev/sda

Read all the output and you can find alot of data including if the disk reported any errors during operation, how many hours, how many start/stop cycles and even the current temperature. If you want to run tests you can do

smartctl -t short /dev/sda
or
smartctl -t long /dev/sda

It will tell you long the tests will run..about 2 minutes for short and 2 1/2 hours for long. It even tells you when the results will be ready! After that, do

smartctl --all /dev/sda

again and you can see the test results.

(Change /dev/sda to whatever drive you want to test)

I could use them both ofcourse.
Right now i started off with the smartmon tools - i have 1 disk reported as bad (having read errors) - and i think i might run a badblocks on that one.
My question was actually which of both should be preferred - i guess smartmon since it is the newest. Does badblocks still work on new drives (i.e. those with smart onboard) btw - is it still able to pinpoint the actual sectors on the disk that are bad?
Since i get read errors on one of the disk including a reference to dma it might as well be an electronics error as actual damage on the platters.