Toshiba 4tb HDD problem?

[teknow]

Hello everyone,

last week I bought a Toshiba MD04ACA400 4tb HDD and I'm noticing an issue with its SMART readings that worries me. The G-Sense Error Rate keeps increasing on it, apparently while it's being written to. After first completing a full surface scan the raw value came up to 4, now after roughly 1tb of data copied to it, it's well above anything I've ever seen on any of my drives that tend to remain at 0 G-Sense errors even after years of usage. As you can see in the screenshot all the other parameters are perfect and even the G-Sense Error normalized value is still at 100 (perfect score), but the anomaly is troubling nonetheless.

The G-Sense errors are definitely not caused by an outside force (there aren't any present), but the above average loudness of the drive's reads/writes and noticeable rattling that can be felt while it's reading/writing, actually makes me think the drive is "G-Erroring" itself with the jerks of its heads moving across platters. I have no experience with Toshiba drives of this size, so I don't know whether this is deemed acceptable behavior for them, or whether I received a faulty piece. Being brand new the drive works perfectly fine for the time being, but the described issue imho doesn't bode well for its longevity. Should I have it replaced with a new one, a different model altogether or is everything just fine? Thanks in advance

 

[Elixer]

Where do you have the drive mounted in?
A HD should never rattle, which is why that indicator is increasing...

 

[teknow]

Thanks for replying. Maybe we understand rattling differently, because as far as I'm concerned every hdd rattles somewhat when reading/writing, except if it's strictly a sequential, or very short read/write where the heads don't have to move a lot and there's no back and forth movement. By rattling I mean the short bursts of vibration that you can hear and if you press your hand against the case also feel. Even laptop hdds "rattle" the way I understand this word, but obviously their rattling is extremely faint. Rattling to me is a defining characteristic of a hdd in action, as opposed to an ssd that's totally silent with no moving parts. So to rattle is normal as far as I'm concerned, it's how vigorously and loudly this drive rattles that is above all my other current hdds. Although back in the days hdds would indeed rattle even more, so without the G-Sense error rate increasing I'd probably think that this is just a loud(er) drive.

To answer your question, the drive is mounted normally just like every other drive I've used, perfectly flat in its default orientation. This cannot be causing my problem.

 

[Elixer]

So, the drive cage/bay has no rubber vibration dampeners, and you are using 4 screws to tighten it down? Is this a cheap case?

For what it worth, vibrations are common, and more so in cheap cases, and you can fix it if you have some rubber washers.

If the HD is making strange noises, that is a completely different thing.

I bet from Toshiba's point of view, nothing is wrong with the drive, it all checks out, and it is within specs.

 

[teknow]

It's a Fractal Design Mini, so a pretty decent case I'd say and the drive bays have rubber grommets, although much stiffer than my other Antec case. However there's been no hdd I ever came across that didn't rattle at least a tiny bit when reading/writing. Exceedingly faint if suspended in a well isolated case or in a laptop but still there if one pays attention. This new Toshiba of mine is just noticeably louder than what I'm used to with current drives. Coupled with the abnormal G-Sense error rate it makes me think there might be an issue with it.

Btw I don't see how even badly mounting a hdd could be the cause of G-Sense errors if there are no external forces. AFAIK the G-Sense parameter is there to track errors caused by bumps, drops and similar external influences (mainly relevant in laptops), but a sub-optimally mounted hdd is still perfectly still and shouldn't register any G-Sense errors imo. IOW the drive's internal forces shouldn't be such to cause a G-Sense error no matter how it's mounted, provided it remains still.

The only strange noises I hear from this drive are similar, maybe equal to those during writes just shorter, but they happen when the drive is idling and should supposedly only emit the spindle noise (which is also louder than in my other drives but since the Toshiba probably has more platters in it, it's not that surprising). I don't like that it's doing this, but I don't know whether it's a problem. I know some models do or used to do some "mechanical gymnastics" while idling that you could hear, but it's been a while since I came across a drive like that (Seagate 7200.7 was the last one like this that I owned).

 

[Elixer]

I suppose if that keeps going up from that last screen shot, then, RMA to the dealer I suppose, since I don't know if Toshiba ships back a new HD or not.

 

[teknow]

Thanks for the help Elixer, I'll do that.

 

[Soulkeeper]

I've noticed 2 gsense errors on my 5TB toshiba.
I never had a gsense smart value increase on a wd or seagate drive, so my only thinking here is toshiba has set them to be much more sensitive.

 

[Data-Medics]

I really wouldn't be too concerned about most S.M.A.R.T. stats except for the ones relating to uncorrectable sectors, and pending sectors. Most of the others are just informational for those who want to collect statistics and track correlation figures.

I think this article nicely explains it in more depth: http://www.computerworld.com/articl...that-actually-predict-hard-drive-failure.html

Half the time SMART errors appearing in software like that are related to incorrect reporting or translating of the raw SMART data and don't actually point to a problem.

Toshiba's are rock solid drives, you should be fine.

 

[Data-Medics]

That having been said, be sure to back up your data. LoL

 

[AlienTech]

Not really sure what that sensor is for. They increase for anything, like powering the drive down. But I have picked up the drive when it was working by mistake and the value did not go up. So like I said, I dont know what it is for as sitting inside the case the value was going up every other day without even touching the case. If it was being dropped I would think it should go up but it was at 0 when I got it and went up after installation. Must be something else then.. The explanation of g sensor could be wrong or toshiba is using something else there..

 

[ArtForz]

Got 4 MD04ACA400 and 2 MD04ACA500 in a Norco-ish 4U box here.

Attr 191 raw sat at 0 during the initial week of burn-in and in use for a few months.
When a 40mm fan in a adjacent switch in the rack went full-on bearing squeal, 191 raw went from 0 to 150-200 on all of them within a few hours. Replaced the fan and it didn't increase ever since.
Makes me think 191 raw on these also counts write retries due to head servo error exceeding safe limits or something similar.

As for the seek noise ... seems normal, they're loud and hot drives - not really surprising considering they're essentially relabeled MG04ACAx00E enterprise drives.

Combine those things and it seems quite plausible that they produce enough seek vibration to make 191 go up all by itself if you have them in soft mounts.


They're slightly odd drives in other regards, too:

1. No full spec sheet to be found anywhere. Since Feb or so there's a 2-page MD04 series product sheet from Toshiba consumer storage, but these drives have been available through OEM channels for well over a year before that.

2. Different linear R/W speeds for the 4 vs 5TB - 185/90 vs. 205/100 MB/s on the outer/inner cylinders and 145 vs. 160 MB/s average.
Haven't seen any other drive family with 1TB platters exhibit something like that, makes me wonder if the 4TB isn't actually a 800GB/platter 5-platter drive.


On the plus side they're really cheap here in .de, support SCTERC and thanks to their enterprise heritage might survive 24/7 operation even if they're not officially rated for it.

 

[AlienTech]

I am getting 222-167MB on the Toshiba 5TB, 208-150MB on the seagate 3TB .. 198-158 on the 2TB seagate. 141-108 on the 2TB greens. 105-81 for the 1TB WDC..

So obviously the higher capacity drives are using more of the platter.. Clearely shown in the difference between the 5TB and 3TB, the 2TB shows less of the platter is being used was supposedly the 1TB standard platters.

I saw the spec sheet for the 5TB enterprise drives not the consumer version They have it as a PDF on the web site somewhere. Since even employees had such a hard time confirming the 128MB cache dram on them it seems they have not published the specs other wise.

 

[gendalf]

Can someone confirm long-term reliability of MD04ACA400 (=PH3400U-1I72)? Did they live through 3+ years and still going with 24x7 usage?

 

[dave_the_nerd]

Can someone confirm long-term reliability of MD04ACA400 (=PH3400U-1I72)? Did they live through 3+ years and still going with 24x7 usage?

200s and 300s in my server at home. Would assume similar reliability from 400s.

YMMV, always back up.

 

[gendalf]

I'm thinking about DT01ACA300 or MD04ACA400 because they're among the cheapest per TB, but reviews vary, also I couldn't find any reviews for over 8760h of usage..

 

[WilliamM2]

I'm thinking about DT01ACA300 or MD04ACA400 because they're among the cheapest per TB, but reviews vary, also I couldn't find any reviews for over 8760h of usage..

I've had my 3TB for over 3.5 years now (over 14000 hours). No issues at all.
It's the retail version, which I notice gets much better reviews than the OEM version, probably due to better packaging, and less DOA drives.