Hard Drive Failure - Happy Ending

[Leonard Grey]

In general I seem to be able to avoid the problems that people post
about here...but I'm not immune to PC trouble.

Last night my computer did not boot. I discovered that the BIOS was not
detecting either of my (ancient IDE) hard drives.

The BIOS was fine: It was detecting my other hardware and the clock
showed the right time. I could feel gentle vibrations from both hard
drives and the cooling fans were working, so my power supply was okay.

I swapped my IDE cable for a new one...no help. Motherboard damaged?
Perhaps, but not likely, unless my UPS failed.

That left the hard drives themselves. I removed my boot drive and
started the computer with only my storage drive and - voila - it was
detected. That meant the electronics in my boot drive had died, and
since it was first on the ribbon cable, neither drive was detected.

The rest was easy: Buy a new drive and restore the old drive from a disk
image made only 12 hours previously.

So what is the point of this post, besides boasting? The point is that
even a non-geek like me can solve most any computer problem with
methodical troubleshooting and a current backup.

 

[Daave]

In general I seem to be able to avoid the problems that people post
about here...but I'm not immune to PC trouble.

Last night my computer did not boot. I discovered that the BIOS was
not detecting either of my (ancient IDE) hard drives.

The BIOS was fine: It was detecting my other hardware and the clock
showed the right time. I could feel gentle vibrations from both hard
drives and the cooling fans were working, so my power supply was okay.

I swapped my IDE cable for a new one...no help. Motherboard damaged?
Perhaps, but not likely, unless my UPS failed.

That left the hard drives themselves. I removed my boot drive and
started the computer with only my storage drive and - voila - it was
detected. That meant the electronics in my boot drive had died, and
since it was first on the ribbon cable, neither drive was detected.

The rest was easy: Buy a new drive and restore the old drive from a
disk image made only 12 hours previously.

So what is the point of this post, besides boasting? The point is that
even a non-geek like me can solve most any computer problem with
methodical troubleshooting and a current backup.

Yes, backups are a must. And a validated image can certainly save one's
bacon!

 

[Unknown]

You fail to recognize the fact that not everybody has methodical
troubleshooting ability.
One must have SOME knowledge of computer architecture to perform trial and
error,
or as you say, 'methodical troubleshooting'.

 

[DavidM]

Yes, backups are a must. And a validated image can certainly save one's
bacon!

So what's the best way of validating an image if you don't have a
spare drive laying around? I keep my C: images on a D: drive, so
certainly don't want to use that!

 

[Brian A.]

So what's the best way of validating an image if you don't have a
spare drive laying around? I keep my C: images on a D: drive, so
certainly don't want to use that!

When a user creates/schedules a backup/image/clone job in their application,
the user has the option to either verify or validate that the information
written to the Destination matches the information read from the Source once the
actual backup/image/clone process has completed.


--

Brian A. Sesko
Conflicts start where information lacks.
http://basconotw.mvps.org/

Suggested posting do's/don'ts: http://members.shaw.ca/dts-l/goodpost.htm
How to ask a question: http://support.microsoft.com/kb/555375

 

[DavidM]

When a user creates/schedules a backup/image/clone job in their application,
the user has the option to either verify or validate that the information
written to the Destination matches the information read from the Source once the
actual backup/image/clone process has completed.

All very fine, but that doesn't validate that the cloned image can be
restored and will boot - that's what I mean by validation, ie the
whole process. An image that can't be restored, or doesn't work, is
worse than usless, cos you don't know you have a problem until it's
too late.

So I don't see a way of testing (fully) that it works without having a
completely spare drive, but I'd be happy to be wrong about that?

 

[Daave]

All very fine, but that doesn't validate that the cloned image can be
restored and will boot - that's what I mean by validation, ie the
whole process. An image that can't be restored, or doesn't work, is
worse than usless, cos you don't know you have a problem until it's
too late.

So I don't see a way of testing (fully) that it works without having a
completely spare drive, but I'd be happy to be wrong about that?

Nothing in life is 100% foolproof. The term "validation" has a specific
meaning (performing what is called a consistency check). You don't need
to fully test an image to validate it. If the image is validated, then
there is an *extremely* small chance that it won't be able to be
restored.

For those who are extra concerned, cloning can be done in addition to
imaging for another layer of backup. The advantage of a clone is that it
can be booted from instantly (and thus "validated" as you have been
using the word).

 

[DavidM]

Nothing in life is 100% foolproof. The term "validation" has a specific
meaning (performing what is called a consistency check). You don't need
to fully test an image to validate it. If the image is validated, then
there is an *extremely* small chance that it won't be able to be
restored.

For those who are extra concerned, cloning can be done in addition to
imaging for another layer of backup. The advantage of a clone is that it
can be booted from instantly (and thus "validated" as you have been
using the word).

Well in 40 years of working in the IT business I've seen many
instances where "verified" backups and images have been useless
because the end to end process of restoring a system has not been
thoroughly tested and has failed to work at the critical time, and
this includes PCs up large complex systems.

I think we'll just agree to differ and let people make up their own
mind about the level of risk that they are happy with for their
particular system.

 

[Daave]

Well in 40 years of working in the IT business I've seen many
instances where "verified" backups and images have been useless
because the end to end process of restoring a system has not been
thoroughly tested and has failed to work at the critical time, and
this includes PCs up large complex systems.

I don't doubt this. A very small percentage of a very large number of
instances (40 years' worth in your case) still yields a significant
number of negative events. And this is the reason I added:

So, I agree with you that "having a completely spare drive" that is a
perfect clone which can be booted from and therefore instantly tested is
the easiest way too minimize risk even further. But again, nothing is
100% foolproof and it's possible (though, *highly* unlikely) for this
method also to eventually fail.

I think we'll just agree to differ and let people make up their own
mind about the level of risk that they are happy with for their
particular system.

There's more common ground here than you think. I believe there was a
misunderstanding of terminology.