Lilo makes a lot of disk noise

Question

After upgrading to Debian Wheezy, the lilo command causes my boot disk to make a lot of noise. I am not aways of any real problems, just the noise sounds somewhat scary, especially since this wasn't the case earlier.

When I run lilo -v4 I can see that each noise corresponds with one of these lines:

Mapped AL=0x00 CX=0xbc27 DX=0xbee0, LBA=12500007
Mapped AL=0x00 CX=0xbc28 DX=0xbee0, LBA=12500008
Mapped AL=0x00 CX=0xbc29 DX=0xbee0, LBA=12500009
Mapped AL=0x00 CX=0xbc2a DX=0xbee0, LBA=12500010
Mapped AL=0x00 CX=0xbc2b DX=0xbee0, LBA=12500011
Mapped AL=0x00 CX=0xbc2c DX=0xbee0, LBA=12500012

there are 27779 such lines in the verbose output.

The noise does not sound unhealthy, more like a regular head movement. It sounds like for each of these lines lilo does a seek and the head happens to be in a distant position.

Does anybody know what might have caused the change from silent operation to noisy operation?

The -v4 was only to see what it is doing. The noise occurs with or witout it — Martin Drautzburg, Sep 02 '13 at 14:06
Not necessarily - if all drivers you need to mount `/` are built-in (they usually are when you are building custom kernel). — peterph, Sep 03 '13 at 18:25

score 1 · Answer 1 · answered Sep 13 '13 at 21:32

I don't know much about lilo, but I can tell you how to figure out if the hard disk is OK. Use smartctl as root to query its status:

smartctl -a /dev/sda | less

In rare cases, the SMART overall-health self-assessment test result line will tell you whether your hard disk is failing, but in my experience, this predictive component of SMART seldom works. Instead, scroll down to the block beginning with SMART Self-test log structure; most probably, it will look like this:

SMART Self-test log structure revision number 1
No self-tests have been logged.  [To run self-tests, use: smartctl -t]

If that is the case, you will need to start a self-test to figure out whether the hard disk is OK. This will take a few hours (depending on the hard disk size). Don't worry, it's non-destructive and it won't block the hard disk, so you can just run it in the background while the filesystem is mounted:

smartctl -t long /dev/sda

Wait until the test is finished (it will tell you how long it will take when you start it), and then re-query the SMART data:

smartctl -a /dev/sda | less

Now look at the self-test block. If it says "Completed without error", your hard disk is fine:

SMART Self-test log structure revision number 1
Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Extended offline    Completed without error       00%      2262         -

If it reports an error, your hard disk is broken and you will need to backup your data immediately, if you haven't already done so. We're using this self-test feature on several thousand hard disks and so far it has been 100% reliable - if it says "Completed without error", the disk is definitely OK, and if it reports an error, it's definitely broken.

Lilo makes a lot of disk noise

1 Answers1