Hard Drive Testing

[Zaish]

Hey all,

I am being told that running your normal hard drive tests ( Read, Verify ) will damage a well working hard drive. I have searched google to find a answer and either this is not a normal question or I am not phrasing it correctly to find the answers. I was wondering if anyone has came accross an article to find this information or knows from personal experience if running such tests will make a perfectly working hard drive to fail all of a sudden.

Thanks again.

 

[NYJimbo]

Seriously ?

 

[anonymous Mac Tech]

Hey all,

I am being told that running your normal hard drive tests ( Read, Verify ) will damage a well working hard drive.

If that's the case that drive is on its way out anyway. For instance, zeroing a drive is considered a good fitness test but in no way should it "kill" a working drive.

 

[NYJimbo]

If that's the case that drive is on its way out anyway. For instance, zeroing a drive is considered a good fitness test but in no way should it "kill" a working drive.

Exactly. If testing a "perfectly good" drive kills it, then it was not a perfectly good drive. If these test were killing good drives people wouldnt be using them.

 

[angry_geek]

Okay, I'll bite. If the hdd is operating normally, you do any test you want to it and it won't fail. Hdd's fail for a myriad of reasons, ie: over heating, shock, wear and tear, surges/brown outs. They can get bad sectors or clusters and continue operating for 10 years. They are definitely a conundrum. I can't tell you how many times a client has asked me how their hdd failed and I have to say I don't know. Now, if a hdd is on its way out, you can kill it with a chkdsk, or other intensive tools. I've seen this happen.

PS You need to go introduce yourself. This way we know how better to respond to your questions. It seems like you may be a student, which is fine. We just need to know.

 

[Zaish]

In responce to angry_geek.

I have worked in this field for a while. 4yrs, even had a bussiness running for a while, but due to the economy I found I was making about the same working in a tech shop. So I am the primary tech here.

I know it sounded like a stupid question, he is a technician himself and said he found the information online about it. I cant seem to find it or I may not be phrasing it the same way he has. I know that it is not a perfectly working drive if it fails, however my question was to explore the possiblilty my understanding was flawed. I know there are many reasons why a drive would fail.

He had brought me a drive that he claimed was working perfectly, so before putting it into a customer machine I test it out, and it failed. He said that my stress testing it caused it fail. So just to cover all area's that perhaps I am in fact doing something I asked here. You never know

 

[NYJimbo]

...... he is a technician himself and said he found the information online about it.......

....."He had brought me a drive that he claimed was working perfectly, so before putting it into a customer machine I test it out, and it failed. He said that my stress testing it caused it fail. "

Since you did not explain what you did to "stress test" it I assume you didnt beat the hell out of it for 24 hours or something.

If he brings you a drive and you test it with perfectly acceptable testing software and it fails then the drive is no good. End of story. The reason for the testing software is exactly what you experienced.

He's not much of a technician if he blames you for doing something we all would do and then cites a reference that he cannot reproduce.

Frankly I would be ****** and him and try to point out how he is wrong, unless that would cost you your job.

 

[Zaish]

While I would find much excitment in saying he is wrong. He is the manager here, he works on computers still but not as much since he runs the store now(customer relations end). He said he found some information online a while about about the tests doing that, so I was just checking because you never know I could have been wrong. Alot of times I will research something I am being told because you never know it might save you time in the future or better understanding.

As far as the time I ran it for was about 5 minutes after a quick reformat of it.
I didnt figure running it for 5 minutes would cause it to do this but than again I dont claim to know everything and I could have been wrong.

 

[Zaish]

To answer what he considered a stress test was: Extended Smart test along with a full read/verify to every sector to make sure everything was okay. I have had Smart test tell me it worked great but found alot of bad sectors before.

 

[angry_geek]

Your boss is way wrong. You are in the right here. I've run DOD scrubs for hours on end on 3 year old drives and was still able to use them in whatever case I wanted. I have several file servers in my shop and house using old drives. As long as you didn't drop it on the floor or beat it with a sledge hammer, you did nothing wrong. It was just a bad drive. If you want, drop a hint to your boss to check the forum. I fear for him, though, if he drops a line on here like he did on you.

 

[usacvlr]

There used to be a virus that simply traversed the drive heads back and forth all the way madly to burn out the drive. Any major reading and writing on a drive over a short period or long period of course will stress it when normal usage will not if the drive is not already heavily fragmented. You could argue that keeping a drive defragged will keep it alive longer.

 

[purple_minion]

Unfortunately I can't cite a source to my info, I have horrible detail retention. However I do remember reading an article stating that defragging versus no defragging didn't harm the drive either way. I think they broke it down into a file spread all over the drive versus defragging it once, consolidating the file, and then accessing it. They were showing which is harder on a drive, small random writes to get the file, or the chugging done to defragment and then access the file. I believe they came to the conclusion the neither did much to the drive.

Also I believe google's study found that temperature made the biggest difference to longevity.

I have used PC-Check to really screw with a drive, letting it run two days in a loop between destructive writing, reading, and mechanical stress testing (which moves the heads back and forth from inside to outside repeatedly for an hour) and it works fine.

Also if a linear read killed the drive it was going to go out very soon anyway. I would rather it go out before you put someone's valuable data on it, or take the time to install an OS, etc. I also remember reading that electronic components either die of old age, or infant mortality. Obviously old age from wear and tear, but you also want to weed out the manufacturing defects. A graph looks like an inverted bell curve, or like a cross section of a bathtub.

Remember suprise HD death is what causes issues, if the drive tapped you on the shoulder, told you "Hey, I'm scheduled to choke out on 7-15-09 better be ready!" A lot of issues would be solved.

 

[Zaish]

I had the manger read some of the comments earlier. He never came back with a continued argument to his claim that it will mess up a hard drive. So yes it was fun to point to him and say " Ha! I told you!"

The bad thing is, he is the one who cuts the check. I sure hope its right

 

[angry_geek]

I sure hope its right

Not to worry. We're kings of geekdom, and we're right

You did nothing wrong; that drive was already on its way out.

 

[purple_minion]

I had the manger read some of the comments earlier. He never came back with a continued argument to his claim that it will mess up a hard drive. So yes it was fun to point to him and say " Ha! I told you!"

The bad thing is, he is the one who cuts the check. I sure hope its right

Sometimes people need to be called out. I usually add a little disclaimer to things I say along the lines of "I'm not an expert but here's what I think, I could be wrong." I don't think any of us are experts, more of a jack of all trades.

Also, who knows... maybe you accidently shocked the drive, computer has a marginal powersupply, the planets aligned just right.

Also if it's not a new drive (not that that would make it not fail) you're going to have a much better chance of it failing, as everything fails eventually. I've had perfectly good drives chugging away in my HTPC, and the next time I reboot the drive will not read the firmware off the platters. It worked fine the whole day/week/month but once it was forced to reread the fireware with a reboot I saw that there was a lot of problems.

Also a lot of OS mask problems, I have put HD's in pc's and looked at the linux system logs to find it screaming at me with issues... however in use surfing the web nothing is wrong. And linux screams, windows doesn't seem to say much unless it's really really gone. I'd get pages full from dmesg, while I might have found one or two events in the logs on a windows machine. Granted I have limited experience with only my home machines, and therefore a small working set to draw conclusions.

Also everyone learns something new every day, even your boss. I learned the other day that the mold on fruit like lemons or muskmelons is penacillin (sp, need to install firefox).