Using smartd to check health of hard-drives?

felixk · 09-23-2009, 07:21 PM

My OS is configured to use <smartd|smartctl> to check the health of both of my hard-drives; one drive(/dev/hdb) is mounted and used routinely. The second drive (/dev/hda) once had a problem - which has been rectified by reformatting. It is now only mounted when being used for storing backups.

However, "smartd" dutifully checks its health every 30 minutes and reports the presence of an unreadable HD sector - despite the fact that this is no longer the case.

Full details are below.

1. smartd checks:

Sep 24 08:16:41 carrot smartd[2717]: Device: /dev/hda, 41 Currently unreadable (pending) sectors
Sep 24 08:16:41 carrot smartd[2717]: Device: /dev/hda, 41 Offline uncorrectable sectors
Sep 24 08:46:40 carrot smartd[2717]: Device: /dev/hda, 41 Currently unreadable (pending) sectors
Sep 24 08:46:40 carrot smartd[2717]: Device: /dev/hda, 41 Offline uncorrectable sectors

2. smartctl test:

[root@carrot felixk]# smartctl -H /dev/hda
smartctl version 5.38 [i686-pc-linux-gnu] Copyright (C) 2002-8 Bruce Allen
Home page is http://smartmontools.sourceforge.net/

=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

While the problem is just an irritation, any advice on how to get rid of the misleading messages will be gratefully received.

Felix

lutusp · 09-23-2009, 07:35 PM

Quote:

Originally Posted by felixk

My OS is configured to use <smartd|smartctl> to check the health of both of my hard-drives; one drive(/dev/hdb) is mounted and used routinely. The second drive (/dev/hda) once had a problem - which has been rectified by reformatting. It is now only mounted when being used for storing backups.

However, "smartd" dutifully checks its health every 30 minutes and reports the presence of an unreadable HD sector - despite the fact that this is no longer the case.

While the problem is just an irritation, any advice on how to get rid of the misleading messages will be gratefully received.

Felix

If it were no longer the case, the message would not be printed. Look into:

Code:

$ man badblocks

Code:

$ man e2fsck

felixk · 09-23-2009, 10:31 PM

Quote:

Originally Posted by lutusp

If it were no longer the case, the message would not be printed. Look into:

Code:

$ man badblocks

Code:

$ man e2fsck

I <hope|think> the penny has dropped.

My backup hard drive :

- still has "bad blocks/sectors" (i.e. reformatting made no
difference); and
- this does not prevent it from being considered "healthy" by smartctl.

Thank you for the prompt response and helpful advice.

Felix

lutusp · 09-24-2009, 03:52 AM

Quote:

Originally Posted by felixk

I <hope|think> the penny has dropped.

My backup hard drive :

- still has "bad blocks/sectors" (i.e. reformatting made no
difference); and
- this does not prevent it from being considered "healthy" by smartctl.

Thank you for the prompt response and helpful advice.

Felix

But reformatting cannot solve the problem. The only way to solve the problem is to directly address it using "badblocks" and "e2fsck".

Code:

$ man badblocks

Code:

$ man e2fsck

felixk · 09-25-2009, 07:18 PM

Quote:

Originally Posted by lutusp

But reformatting cannot solve the problem. The only way to solve the problem is to directly address it using "badblocks" and "e2fsck".

Code:

$ man badblocks

Code:

$ man e2fsck

Unfortunately, it is not that simple.

I use the "xfs file system" and not the <e2fs|e3fs> system.

I do have access to "/sbin/fsck.xfs"; but this merely confirms what I already know - that I have a damaged <block|sector>.

And, based on <warnings|advice> in "man badblocks" (to use the badblocks option as part of a reformat) I am reluctant to experiment with the badblocks command (since mkfs.xfs does not provide a "badblocks" option).

Again, thank you for the continuing intewrest in my problem.

Felix