Raw Error Read Rate - WD 1.5 TB

H

Hellhawk

n00b
Joined
Oct 18, 2004
Messages
61
Ok, I need some insight here and perhaps someone here can offer it. I had an HD go bad on my home server (a WD Green 1.5 TB was showing 1445 pending bad sectors in SMART, so I filed for an RMA). So no problems there. However this got me into a rather paranoid state of mind and so I decided to start looking at all the SMART data on all my systems and checking the drives. Long story short, my new 1.5 TB Caviar Black on my i7 system is showing potential errors.

I've been researching this for almost 2 days now and I think I've driven myself to the point of insanity. At this point I may just RMA the drive for piece of mind but part of me is curious if I'm over-reacting.

With that said, how much weight should be given to the fields:

Raw Read Error Rate and Write Error Rate? Both of those fields have increased data values, 5 & 22 respectively. The current and worst values for both (200) is unchanged, but the raw value is 5 and 22. Now the threshold for the Raw Read is 51 and if I understand how this works, the current value of 200 will decrease and if it ever falls to 51 or below, there's a problem? So then, where does the data value come into place? I've done two surface scans and each time, the Raw Read Error Rate has increased (though the drive passes fine).

For better illustration:

ID/Name/Current/Worst/Threshold/Data
01/Raw Read Error Rate/200/200/5/OK

Now, Reallocated Event Count, Pending Sector Count, Offline Uncorrectable all have data values of 0, which makes me think I'm probably looking into this way too much. Information on this is very spread as the relative accuracy and usefulness of SMART itself seems highly debated.

I have 4 other Caviar Blacks (2 640's and 2 1 TB's) that have run for well over a year each and neither have data values for Raw Read or Write above 0.

Anyone have any experience/insight? Feel free to tell me I'm paranoid here if that is really the case...

Any help is appreciated.
 
Raw Read Error Rate and Write Error Rate? Both of those fields have increased data values, 5 & 22 respectively. The current and worst values for both (200) is unchanged, but the raw value is 5 and 22.
Click to expand...

I have not made any useful conclusions from those two SMART parameters (and I monitor the SMART for over 100 drives at work). One reason is the value ranges seem to wildly vary from drive to drive.
 
Interesting link, but yeah, not quite sure of the relevance. Although ZFS sounds fascinating, not quite sure how that helps me out of my current predicament. Though it does give me a potential reason to move away from WHS in the future.
 
In my experience, with WD drives the read/write error rates should normally stay at 0. Since it's just 5 & 22, it's not serious at this point, but may indicate future problems.

I've got a Samsung F1 with some known corrupted files. Every time I try to access those files, the read error rate goes up. It also creates a "Disk" type event in Windows event viewer that states the drive has a "bad block". So it CAN in some cases indicate real problems. However, my HDD does have pending sectors in addition to the read errors.

External events can also cause read/write errors. I've had the write error rate go up when a power outage happened while I was writing to the HDD. A bad cable could probably do it as well.
 
You are correct in that if the value goes down to under 5 then it would be considered 'failing.'

Since its currently 200 and a wost value of 200 I would say its probably fine right now. The raw read/write error rates aren't usually the best thing to go by when a drive is failing. The two biggest things I see on drives failing are the re-allocated sector count and pending re-allocation sector count.
 
Hellhawk said:
Interesting link, but yeah, not quite sure of the relevance. Although ZFS sounds fascinating, not quite sure how that helps me out of my current predicament. Though it does give me a potential reason to move away from WHS in the future.
Click to expand...
The point is that possibly you will encounter corrupt files, without SMARTs alarming. SMART might think everything is good, and suddenly you get corrupt files. If you are paranoid, I just want to tell you that SMART is nothing you can trust on. SMART does not protect against data corruption. And SMART will not know if data corruption ever occured. Just read the link, he encounters data corruption without any alarms - data corruption just suddenly appeared.
 
Thank you very much folks for all the insight. On the basis of the fact that it might indicate future problems, I spoke to WD about the issue. The drive is older than 30 days, so I can't return it to the retailer. However, it only has a few hundred hours of use, and because of this, WD has not only agreed to replace it, but they are also giving me an upgrade to a 2 TB drive. All in all, an acceptable solution.

My other 4 WD drives are clean after years of use, so I suspect this was just bad luck.
 
You must log in or register to reply here.
Back
Top