Is it necessary to run hard drive test before using the drives?

T

theinv

Limp Gawd
Joined
Apr 17, 2008
Messages
216
Ok, I got two Hitachi 5300 series drives and went to their website to download the Hitachi Drive Fitness Test program.

But the program only runs in DOS mode, definitely a little outdated in the modern world. The worse yet is it will take more than 40 minutes (or even hours) to run the test. This makes me don't want to run it because I can't use my computer at the same time.

Is it really necessary to run hard drive test using the manufacturer's official tools before using the drives? Can I just fill the drive with data and if no problem occur then it means the drive is in good health?


Any other tools I can use?
 
Last edited:
The Hitachi DFT is highly recommended if you care to know if the drive is working properly or not - you cannot use S.M.A.R.T. status ever as an indicator of a drive's well being, so don't even bother. For those that might dare say I'm wrong with that, I've got 9 Western Digital WD800 80GB IDE drives from a few years ago in a shoebox here, all 9 of them are completely pooched/toasted/useless/defective and all 9 of them have "perfect" S.M.A.R.T. status, just as a case - ok, 9 cases - in point.

The DFT takes so long because it's thorough. People seem to think having a physical hard drive means shit works like an SSD and is blazing fast, but they're wrong: physical hard drives are mechanical and as such it's going to take time to process every usable sector on the drive for the test (and no, a quick/short test really isn't going to cut it).

If you care about the drives, and you went through the trouble of making this thread with that post above, you already know what you're supposed to do.

Hell, the time you've "wasted" wondering about the right thing to do would basically be almost done with one of them since you made the post ~30 minutes ago according to the time you opened this thread. :)

Make the bootable CD, run the test on both drives, then you'll know for sure you're dealing with good solid storage media, in under 2 hours, and you're done.

As for it being "absolutely necessary" to do the testing, no, of course it's not. You can plug 'em in, partition 'em, format 'em, install the OS and use 'em without issues 99.999% of the time. But it's that 0.001% of the time that causes people to kick themselves the hardest with that "Dammit, I should have tested these drives, I knew it... I knew I should have tested them..." frustration.

Better safe than sorry. Run the tests, get 'er done, that's about it.

As for any other tools, the Hitachi DFT is the only tool I'd dare recommend for proper diagnostics on Hitachi drives, just as Seagate tools for Seagate works, and Western Digital for Western Digital hardware, etc. There is no better tool - suck it up, run the tests, move on. Get a decent DVD movie to watch about 1.5 hours long, grab some sodas or whatever, kick back and watch the show while the test runs.

I mean... really?
 
Proper testing of these disks is incredibly important.

I just did an order of several 5K3000 2TB for a college radio station.

Turns out one of them is mislabelled - it's only a 1TB.

Just glad I caught that before throwing it into an array.
 
I test each new drive for a few days before putting it into regular use, despite no HD I've bought ever failing on me (but I've returned a few Samsungs for excessive vibration), unlike memory modules. At this moment, I'm testing a 2TB Hitachi, which takes almost 5 hours to scan.

For the last couple of years I've been relying mostly on a self-booting diagnostic called MHDD because it not only indicates what sectors are completely defective but also which are marginal and need multiple attempts to read correctly. Apparently manufacturers think it's acceptable to require 15 attempts to read any sector.

If you don't want to tie up your computer testing new drives, there's a Windows version of MHDD, called HDDscan, but it gives a lot of false positives, maybe due to Windows overhead, and its SMART testing doesn't work right with some drives and probably no drives bigger than 1TB. The Windows versions of WD's Data Lifeguard Tools and Seagate's SeaTools diagnostics will work with any brand of drive, and then there's HDtune, a free benchmark & diagnostic.

I think I've had much worse luck with memory modules because unlike hard drives, most retail memory is made with reject parts (RAM chips with no manufacturer listed).
 
Last edited:
Ok, thanks all!
Just finished testing one 2TB 5k3000 drive with the Hitachi Drive Fitness utility, took 6 hours, found 0 errors. That's a lot of time, but I think it is worth it. Maybe in the future, I will have a low spec machine dedicate to hard drive testing so it doesn't tied up my pc.
 
You should probably run the factory disk tool if you can. 40 minutes is not really much of an investment to catch potential problems. I recently purchased 5x2tb drives and ended up with 1 dud. Luckily I was able to identify the problem before I was dealing with real data.

Just like you, I was irked at the fact that it was DOS only. I didn't have a floppy drive (or disk) either. I managed to find a 'beta' CD image of the tool but unfortunately it didn't work with my configuration of hardware for some reason. After a night of google searches I ran across some forum posts of other users in the same situation as me who hacked together a bootable usb stick with the disk tool based on the beta cd image. As luck would have it, I ended up damaging my USB stick while trying to make it bootable. After that I gave up and just ran HD Tune on all 5 of my new disks. In a round about way I was able to catch that one of my drives was bad.

I ran full sector tests on all 5 and got a clean bill of health. Then just for fun I did a quick read/write benchmark test on each one and noticed in the graphs that one of them had a strange drop in performance towards the end of the disk -- where it spikes down and then back up.

It seemed really odd so I repeated the performance benchmark a few times and kept getting the same results -- this time noticing a grinding noise that would only show up when the sectors at the end of the disk were being accessed. At this point I was worried that my drive was bad so I re-ran the sector test and to my surprise it came back clean again. I definitely heard the grinding noise again so put it back through the sector test again. Finally after about 10 passes with the sector test it started reporting bad sectors towards the end of the disk... and lots of them.

I can't say for sure since I was never able to get the mfg's disk tool to run, but I suspect it would have detected the problem right from the start and saved me hours upon hours of testing.

Had I not done any testing at all I probably would not have run into the problem until much later when the disk was mostly full with real data.
 
The writer of UnRAID has a utility called pre_clear that does a highly thorough fitness test as well.

Basically there's a bathtub curve where a large number of drives are going to fail in the first three months. This tries to exercise the drives enough to precipitate any failures before you put data on them.
 
I like Hitachi DFT, and it works on any manufacturer's drive. It has been successful in determining failures for any manufacturer, by my experience. Some of the other 3rd party tools out there are a pain.

You could always do the drive test while you sleep. You're not really using the machine then. :)
 
iroc409 said:
You could always do the drive test while you sleep. You're not really using the machine then. :)
Click to expand...

That's a good tip. :)

I tested my 2nd drive while I sleep. It is now more fun to test the drives, gives me more confident.

One more question, after you tested the drives using the manufacturer's utility, do you still test it for a few more days by moving actual data in/out of it before put it into actual use?
 
You must log in or register to reply here.
Back
Top