Thank you for being a valued part of the CNET community. As of December 1, 2020, the forums are in read-only format. In early 2021, CNET Forums will no longer be available. We are grateful for the participation and advice you have provided to one another over the years.

Thanks,

CNET Support

Question

Does anyone know of a good HDD test for Windows?

Mar 27, 2012 1:57AM PDT

I currently use the HDD test built into PassMark, but it doesn't work all that well.

Does anyone know if a good HDD test I can automate and run in Windows? It needs to be able to start automatically, like through a switch or something and produce a log I can examine afterwards.

Thanks

Discussion is locked

- Collapse -
Clarification Request
(NT) Which windows? And why not the scan mode for chkdsk?
Mar 28, 2012 6:38AM PDT
- Collapse -
Windows Vista/7
Mar 29, 2012 12:28AM PDT

I can't use chkdsk because I can't control how long chkdsk takes to run.

- Collapse -
Sure you can.
Mar 29, 2012 4:20AM PDT

But let me preface that I'm also a programmer and could tell it to stop as needed.

Maybe the usual drive test is the real answer here or you'll have to write your own to meet your requirements.

Say, how about why you need this? Most drives are fine with SMART and such. What's driving you to do this?
Bob

- Collapse -
Sort of
Mar 29, 2012 6:08AM PDT

As far as I know, ChkDsk can't inherently do this. However, I could programmatically just kill the process after the set amount of time.

It's for the company I work for. I wrote an application that run several tests that determine the operational integrity of notebooks. PassMark's HDD test doesn't work well enough. We've seen a lot of false positives. So we need a test that performs better.

I'll have to run ChkDsk on some bad HDDs to make sure it can detect the errors. If that's the case, then ChkDsk will work fine. I'm not sure why I didn't think about killing the process myself Silly

- Collapse -
Tipping my hand.
Mar 29, 2012 6:31AM PDT

One of my consulting gigs involves HDD use in DVRs. It turns out that with what the software is doing inside the HDD (it has software there!) the best tool is the one from the maker and the SMART report. There are folk that think this is false positive because the drive continues to work but once you've dealt with about 100K drives you learn it's not false. The drives do fail later. Not all but most.
Bob

- Collapse -
Right. But...
Mar 29, 2012 6:49AM PDT

That doesn't quite apply to this case. It's not a false failure, which is what you described. PassMark isn't reporting an error and we're saying it's OK. PassMark is reporting no error and our OEM equipment is saying the HDDs are bad.

- Collapse -
Can you tell more about the OEM equipment?
Mar 29, 2012 9:24AM PDT

Here we used the SMART report tool and it is in line with some dedicated tester.

However, I think I'm getting the picture now. You want to know if the HDD is good. Well, it is good enough at the time of the test to pass that test. This is NOT going to tell us it's good in 3 months or 1 day.

As we know, it's not a matter of if a HDD will fail, just when. You may want to collect your data and have a statistician go over it. We found that there is a correlation between hours of use, the SMART report and earlier than what some expected failures.

-> And in closing, about half the bad drives in the return pile are fine. Those are usually user errors, bad formatting or "other." I wish users were a bit more savvy but most folk want toast from the toaster and care little about how it works.
Bob

- Collapse -
Well...
Mar 29, 2012 10:13AM PDT

I work for a repair center. If it legitimately passes today, I don't care when in the future it will fail.

But that isn't the case here. We have drives that are passing PassMark, but when we put it on the OEM equipment, it fails. As for the equipment itself, I'm not very familiar with it. Another department takes care of that. But we have equipment for each OEM; Toshiba, Western Digital, Hitachi, etc.

So all I'm concerned with is making sure we fail drives that are bad now. I'll probably get a chance to run ChkDsk on some HDDs at the beginning of next week. But I'll still look into some other tools.

- Collapse -
What I'd do next.
Mar 29, 2012 11:37AM PDT

Since the tester make, model is not shared here I'd look at the SMART numbers. Even our custom in house software and no I can't share it software has turned to using that.

And yes, Passmark would pass those drives.
Bob

- Collapse -
SMART seems fairly irrelevant
Mar 29, 2012 12:14PM PDT

None of the bad drives so far have triggered SMART. I was told earlier, that after they fail the OEM equipment, we use a third party HDD tester. It's this rack of devices that you plug the HDDs into. I forget what it's called.

All but a couple have failed that test too. The strange thing is that PassMark works fine in another business group and fails HDDs accordingly.

- Collapse -
Until we know more,
Mar 29, 2012 2:30PM PDT

I can't guess what they are doing. There is some thought about measuring read/write response times but that means I have to, well we can't get the deep without going to code.

There is a problem here and we faced it at the DVR place is that the drives are too cheap. That is, the hours spent testing were fast approaching the replacement costs.

-> Why not stick with the OEM test results for now and keep looking at other disk test software or begin writing your own?
Bob

- Collapse -
Hm...
Mar 29, 2012 4:13PM PDT

The OEM tests are huge machines. My job is to create and use software the will run on the customer notebook.After repairs are done, we boot the customer HDD and run our software.

The only reason we know these HDDs are bad is because the techs couldn't figure out what was wrong. So they ended up replacing the drives. We double checked the drives on the OEM equipment and found them to be bad.

So that's where I'm at with that.

As for making my own test... that's not a bad idea. It would be difficult and time consuming, but not impossible. I'll look into it. Hopefully ChkDsk will work though.

We'll see.

Thanks for the help so far.

- Collapse -
Re: chkdsk
Mar 29, 2012 5:53PM PDT

Chkdsk is a rather limited tool.
It checks the integrity of the filesystem.
It tries to read all sectors to see if it happens to encounter one that says it's bad.
And that's all.

Any disk test worth the name should at least do a write/read-test at random sectors. And measure the timing of those operations. And of course be able to study the SMART counters and report on them.

Kees

- Collapse -
Re: testing
Mar 29, 2012 6:00PM PDT

If a testing program (even the makers own diagnostics, which should be the best, because they know the internals) says it's OK and your "OEM equipment" (what's that?) says it isn't, one starts thinking that either the errors are transient or your OEM equipment is in error. Did you ever study it's reports to see what's wrong with the disk and to try if it reproducible?

Moreover, why the limitation to Windows software? What's wrong for testing purposes with a stand alone solution (like memtest, running from CD or USB-stick) or Linux?

Kees

- Collapse -
Not Quite
Mar 30, 2012 1:17AM PDT

The "makers" diagnostics are the one that's saying it's bad. That's what I mean when I say OEM equipment. The HDDs are passing PassMark, but when we run the Toshiba HDDs on the Toshiba equipment (OEM equipment), it fails. Since Toshiba knows their HDDs the best, it's safe to assume that their equipment is OK. That and the equipment was calibrated in January.

- Collapse -
We use...
Mar 30, 2012 3:02AM PDT

1. The HDD maker's software. Some are Windows based.

2. We also use some proprietary software (sorry, can't tell you much more because it's the company's secret sauce.)

3. We also use the SMART values.

Try item 1 again. The HDD maker's test software that works on Windows.

I think you had an unwritten requirement that you would be limited to one test software for all drives. That does not exist.
Bob

- Collapse -
I Would If I Could
Mar 30, 2012 4:15AM PDT

I've looked at trying to use the Windows tests from the OEMs, but none have the ability to be automated. I can't start any automatically and none produce error logs.

- Collapse -
Which is why
Mar 30, 2012 4:19AM PDT

Now you know why the company created its own. The problem is a matter of payback.

Will creating your own pay off in the long run? For us it was doubtful, even with some 100K drives a year (it's a DVR system but that's all I can tell.)
Bob

- Collapse -
One more thing...
Mar 30, 2012 6:28AM PDT

Actually, last night I ran SeaTools For Windows and use ProcMon to monitor what it did. It appeared to write some type of log. So maybe I could grab that if SeaTools For Windows is good enough. The only issue is starting it automatically and getting past the license agreement.

- Collapse -
Answer
Just a tool we use.
Mar 30, 2012 6:36AM PDT

Look up AutoIT if you need to automate things such as accepting a license screen, etc.
Bob

- Collapse -
No. Heck no.
Mar 30, 2012 8:49AM PDT

I hate AutoIT. I'm a developer. I can do these things in a real programming language. I just don't want to do that unless I have to.

- Collapse -
Why I thought of this
Mar 30, 2012 8:51AM PDT

Was your last post with "The only issue is starting it automatically and getting past the license agreement."

Sorry but it lead me to think that you needed something like that.
Bob