Mailing List Archive


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [tlug] Potentially Dying Hard Disk Questions



On Tue, 03 Aug 2004 20:09:48 +0900
"Zev Blut" <zb@example.com> wrote:

> Hello,
> I wondering if anyone any comments or experience with dying hard
> disks.  I have a, less than a year old, Maxtor 120GB 2MB cache IDE hard
> disk that is acting up on me.
> 
> The symptoms are:
> 
> * Every once in a while the drive starts to make strange noises, like
> speaker clicks or a suction cup being pulled out of suction.  It is
> not the normal hard disk grinding noise I know from older disks.
Strange noises are scary. You probably should get a spare disk before it
dies.
 
> * Less often the system freezes when accessing the disk.  Is it
> common for a Linux system to totally block until the disk comes back
> alive?  Let me rephrase that does the kernel block on some IDE based
> operations or is this a motherboard thing?
For DMA timeouts, there is a kernel block. They are not neccesarily as
evil (can be caused by bad cabeling) and a bus reset usually helps. I get
these errors from IDE cdrom drives as well sometimes.

> * I have also had the drive disappear from the /dev list by the.
> This required a reboot for the disk to reappear.  Are there any
> commands that might be used to get the disk back without a reboot?
If the bus reset doesn't help, then you will need to reboot. I have a
cdrom that does this sometimes (even the lights go out) and hdparm can't
turn it back on.
I had a DMA timeout error a while ago on a HD that brought down a disk in
my raid array, and it turned out that it was a cable error (bad contact).

> Also, once I received I/O errors that prevented me from using the halt
> command.  I think that might be related to it dropping out.
Of course if the halt command needs to be read from that disk, that's bad
luck.


Home | Main Index | Thread Index

Home Page Mailing List Linux and Japan TLUG Members Links