Sector read/write wear on rotating media

J

Jim

Hello,

I am working on an embedded system that uses a hard drive (rotating
media, laptop drive, not CompactFlash).

My application will write to a particular small range of sectors and
then read them back. This will be done using C code that does low
level sector reads and writes, without going through a file system.

With CompactFlash, I would worry about wear leveling and limited
writes. But this is rotating media, not CompactFlash.

Should I be concerned that rotating media will also degrade at some
point if I hammer the same sectors over and over, with other sporadic
seeks to other cylinders from time to time? Or should it just run
forever (within the nominal failure rate) no matter whether I
write/read the same small set of sectors over and over?

Thanks.

Jim
 
A

Arno Wagner

Previously Jim said:
I am working on an embedded system that uses a hard drive (rotating
media, laptop drive, not CompactFlash).
My application will write to a particular small range of sectors and
then read them back. This will be done using C code that does low
level sector reads and writes, without going through a file system.
With CompactFlash, I would worry about wear leveling and limited
writes. But this is rotating media, not CompactFlash.
Should I be concerned that rotating media will also degrade at some
point if I hammer the same sectors over and over, with other sporadic
seeks to other cylinders from time to time? Or should it just run
forever (within the nominal failure rate) no matter whether I
write/read the same small set of sectors over and over?

I think HDD wear is not area specific, so it should live
as long as if you were using another access pattern.

Arno
 
F

Folkert Rienstra

Arno Wagner said:
I think HDD wear is not area specific,

That's usually the idea. But not his.
so it should live as long as if you were using another access pattern.

Depends on whether the drive gets the opportunity to place the heads
in a different area from time to time or that the usage pattern forces
them to stay in the same area and heat up that particular area.
 
J

Jim

Arno said:
I think HDD wear is not area specific, so it should live
as long as if you were using another access pattern.
Depends on whether the drive gets the opportunity to place the heads
in a different area from time to time or that the usage pattern forces
them to stay in the same area and heat up that particular area.

That's interesting. I hadn't thought of heat buildup if I stayed on
the same track all the time. In my embedded application, there will
be occasional seeking to other cylinders for other I/O that does not
involve the few sectors that are repetitively written and read as an
IPC mechanism.

I guess what I'm really wondering is if there is a reliability issue
with writing excessively to the same few sectors, such that the
magnetic properties would somehow weaken and become unreliable. If
this doesn't really happen with rotating media, I have no worries. But
if drive life assumes broad distribution of writes on average, then I
wonder if hammering the same sectors over and over would result in
reduced drive life, i.e., would statistical field failures increase.

Thanks.

Jim
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top