S.M.A.R.T. Readings

[georgenoise]

So normally just as a baseline on systems I'm working on I run CrystalDiskInfo or GSmartControl, that way if there is a hard drive issue I know before, it actually kind of acts as an up-sell. I've noticed more often than not though that GSmart will show prefailure readings where nothing else, CD/HDTune/Mfg Tools often show nothing. Has anyone else noticed this and if so at what point do you decide to tell the client they have a potential issue?

I'm actually going through it with a Dell system that is under warranty right now. GSmart is the only program that shows anything and it's just a couple of fields that are in prefailure. The machine randomly stalls, sometimes it takes 15 minutes to load from a cold start, and has other random issues. The machine is an i7 w/16GB of RAM so definitely not a bottleneck and after 20+ years in the business I can tell that something isn't right, that something being the HD. The customer doesn't understand that Dell won't warranty the item until it shows up something on their diagnostics but they insist on fixing the problem. I'm pulling hair right now listening to the complaints but there isn't much I can do other than try to stress the drive in hopes that it gets to a point that it fails but I find that a little on the shady side of things.

Anyway, any advice would be appreciated. Thanks.

 

[Larry Sabo]

I guess you could clone the drive to a new/known-healthy drive and see if that clears up the issues. If it does, even though SMART and short/long self-tests show nothing wrong, I'd suggest the client just have you replace it at their cost. If they can't accept that, I'd tactfully suggest they find someone else who's willing to go through the frustration of trying to get Dell to replace the drive.

 

[nlinecomputers]

Just prove it to the client. Clone the drive to a new drive and show it to the client. When they see it is working correctly they can decide to screw the warranty and pay you. If not, show them the door.

 

[nlinecomputers]

I guess you could clone the drive to a new/known-healthy drive and see if that clears up the issues. If it does, even though SMART and short/long self-tests show nothing wrong, I'd suggest the client just have you replace it at their cost. If they can't accept that, I'd tactfully suggest they find someone else whose willing to go through the frustration of trying to get Dell to replace the drive.

LOL great minds....

 

[Larry Sabo]

LOL great minds....

Hahaha. Beat you to it!

 

[fencepost]

The other thing I'd look for is disk, atapi or NTFS events in the Event Log - if Windows is hanging up like that it may be logging things that will help convince your customer.

The other thing you can do (unless this is a mongo storage drive) is take this as an opportunity to sell them on an SSD, automated backup if they don't have it, or both.

"My experience with symptoms like this are that the drive is sending warning signals, kind of like angina before a heart attack. Your Dell warranty will likely cover replacement of the drive (but not data, etc.) once the symptoms get bad enough if they do before something seizes up, but not until things go downhill further. Or, we can just bypass the issue and put in a new SSD that will also be much faster than what you originally got from Dell - and if we do it now while we can easily copy all your data off, then you don't have to go through the hassle of a Windows reinstall, reinstallation of all your apps, and restoring your data from backups."

 

[SmittenmittenS]

We often see this with Lenovo as well. They have a built in Storage test that runs a few tests but never shows a failure or that anything is wrong. We will boot up with Parted Magic and run a health check and we see pre-failure warnings with the same symptoms you described above. When we contact Lenovo we just tell them that we ran 3rd party tools and they showed pre-failure warnings and that business clients with business computers can't afford for the drive to fail completely. Pre-Failure is Failure. They document and still send us out "fresh" drives.

 

[cypress]

The other thing I'd look for is disk, atapi or NTFS events in the Event Log - if Windows is hanging up like that it may be logging things that will help convince your customer.

The other thing you can do (unless this is a mongo storage drive) is take this as an opportunity to sell them on an SSD, automated backup if they don't have it, or both.

"My experience with symptoms like this are that the drive is sending warning signals, kind of like angina before a heart attack. Your Dell warranty will likely cover replacement of the drive (but not data, etc.) once the symptoms get bad enough if they do before something seizes up, but not until things go downhill further. Or, we can just bypass the issue and put in a new SSD that will also be much faster than what you originally got from Dell - and if we do it now while we can easily copy all your data off, then you don't have to go through the hassle of a Windows reinstall, reinstallation of all your apps, and restoring your data from backups."

This is what I would do. Up sell them on the speed of the SSD and having a proper backup in place

 

[carmen617]

Just have Dell replace the drive under warranty.. Get on a chat with Dell, put in the service tag number, and tell them the drive failed diagnostics with error code 2000-0142. Tell them you are a PC tech working with their client's machine and can replace the drive yourself. They will overnight you a drive, you charge the client for the labor, Dell pays for the drive, and you send them the client's drive back after wiping it.

 

[sapphirescales]

Just have Dell replace the drive under warranty.. Get on a chat with Dell, put in the service tag number, and tell them the drive failed diagnostics with error code 2000-0142. Tell them you are a PC tech working with their client's machine and can replace the drive yourself. They will overnight you a drive, you charge the client for the labor, Dell pays for the drive, and you send them the client's drive back after wiping it.

Jeesh, why bother? How much is a hard drive nowadays? $40? I wouldn't want to go through all that BS for $40. Time is money, and wasting my time contacting a manufacturer to get them to honor a warranty is NOT a good use of my time. The client would have to pay me $100 in labor just to get a "free" $40 drive. Better to just either pay the $40 and get a new (and more reliable) drive, or better yet pay $100 and get a faster, better SSD drive.

 

[NviGate Systems]

I agree. If a client wants us to facilitate warranty with an OEM, we charge hourly as we have to be on the phone and follow the "technician/minimum wage script reader's" directions. Also, most OEM's no longer will replace a drive unless it "fails" the diagnostic test they provide. On some Lenovo Edge machines the built in test does not do a full sector test. So unless you have bad sectors in the random areas it tests you will never get warranty on the drive. HP and Lenovo I know for sure on the consumer class require a failure code to get a new part.

We have a grandfathered partner agreement with Acer so I can usually get any part I need even for laptops, they generally don't make an issue, plus all the staff are from Texas, so I can understand them!

I use HDD Sentinel for all my S.M.A.R.T. needs.

 

[PCX]

The "pre-failure" and "old age" in indicators gsmartcontrol are there to tell you what type of failure it would be if the hard drive was in fact failing. You will even see this on brand new hard drives.

You also need to make sure you are doing more than just looking at the SMART data. The SMART data is not an accurate depiction of the hard drive's current state. In order to accurately determine if the hard drive is failing, you need to look at the attributes AND run the SMART tests. The only time we do not run the SMART tests is if the attributes fail. When looking for failing attributes, you want to look at the RAW data. If these attributes are failing, they will usually be marked in red. However, just because an attribute is marked in red, it does not mean that the drive is failing or has even reached the threshold for that particular attribute. With that being said, you will want to come up with your own thresholds for certain types of failures. For instance, in my experience, after about 20 bad sectors, you start seeing a lot more Read Failures when running the SMART tests, even if the threshold according to gsmartcontrol is over 100. This is why we suggest replacing any hard drive that has 20 or more bad sectors.