WD Green Drives no good for QNAP's??

[occamsrazor]

Thanks for the reply Jason.
I don't have the NAS yet but my existing drive reports itself as WD20 EADS-32S2B0
Any idea about that model version? This one was purchased in the Middle East.
Thanks,
Ben

[QNAPJason]

Hi Ben,
May I know if you're experiencing slow performance on WD20EADS-32S2B0 ?

[occamsrazor]

Jason, I don't have a NAS yet, I was looking to buy one, probably a TS-239 Pro II or TS-259.
At the moment the drive is only being used in a single USB enclosure....

[mr_density]

Hi,

I'm new here and recently bougth a qnap 219p (very happy with it), just discovered there is a lot of talk about the lcc and the WD15EARS HDD.

My SMART INFO:
ID Description Value Worst Value Threshold Raw value Status
193 Load_Cycle_Count 199 199 000 3971 OK

Was wondering if the issue corresponds to the numbers above?

In an other thread I did read the support for WD15EARS HDD is coming in the new firmware update pretty soon. That would be a good.

[mr_density]

I think I solved the load cycle count issue, after a few hours running and looking at the load cycle count, it isn't incrementing as fast as it did. I took out the WD15EARS drive and added a jumper to enable SSC on the drive. Afther that I did a full reinitialization of my qnap 219p and updated the firmware to version 3.2.4 Build 0315 and formated my HDD in ext3.

I'm still going to upgrade to version 3.2.5 when it is available because qnap is saying this upgrade really supports the WD15EARS drives.

[Avatar]

I have a TS-439 Pro with four WD10EADS-00L5B1 (1Tb each), they have been removed from the compatibility list for the problems addressed here. As I said in another thread, one of them has started to work badly (>138.000 read errors by now), the HD are continuously working even with the minimum of processes running and the process who takes more CPU cycles are the kernel modules md*. Maybe the array is continuously rebuilding? The web interface says nothing is wrong. I do not think at all that this malfunctioning of one of the HD is directly related to LCC etc. because the others three HD works well and they are identical and have the same lifetime. It is defective, that's all. But it still works, the read errors are greatly increasing only when I run the SMART test on that HD, otherwise they do not increase. The test always terminate with an HD reset after an hour... My point is, what can I do now? I am not going to substitute an unsupported HD with the same model, can I install a Seagate Barracuda ST31000528AS which has similar features (Buffer, speed, etc) and is supported in its place? I am on an RAID5 array. Can I substitute the 1Tb HD with a 1.5Tb and continue to use the array until I replace the others three? In that case I will end up with an increased capacity and a more stable system, not too bad...

Any help from QNap would be greatly appreciated.

[loekf]

In my case one WDC20EADS drive just died.

I made the mistake to mess with the volume and RAID settings in the hope the "read & write" errors the 419P reported were just an incident. I tried to rebuild the RAID5 array with the ejected drive, but no luck. Kept getting read/write errors.

I ran a bad block scan and it reports more than 100 bad blocks (sectors), so my conclusion is that the drive has gone "bad". Funny enough the SMART data suggests there's nothing wrong with the drive.

I have RMA'ed the drive to Western Digital. I replaced the drive with a new WDC20EARS (64 MB cache). In the shop I heard that the "EADS" series has become "end of life" (= is not manufactured anymore). The main difference is cache size, which is irrelevant for a NAS. I have to add that WDC's website suggests the EADS series is still in production.

I should just have swapped the drive and let the NAS rebuild its RAID5 array, now I have to copy all data back again.. sigh

[occamsrazor]

So my TS-239 Pro II has arrived, though I haven't set it up yet. I have two drives I'd planned to use with it:

WD20EADS-00R6B0
WD20EADS-32S2B0

My understanding is the 2nd one is slightly newer, I could be wrong. I've read a zillion forum threads on the high LLC counts, using WDIDLE.exe utility, etc etc and am now completely lost as to what is what. Before I set it up is anyone able to answer the following:

1. Is one or the other of the above considered to have "fixed" the problem, are there any known differences?

2. Does the high LLC count problem only occur when used in a RAID situation? Because I'm planning to use them as individual volumes....

3. Has anyone here used the WDIDLE.exe utility to diable or increase the head park time? Should I try to run this?

There's so much confusion out there over problems with these drives...

[petur]

Somebody *please* explain to me why people keep buying WD drives? Samsung makes some very good drives for a nice price. Run rock solid in my boxes...

[loekf]

Somebody *please* explain to me why people keep buying WD drives? Samsung makes some very good drives for a nice price. Run rock solid in my boxes...

Ehhh..... /EMBERASSEMENT MODE ON

Guess it's the brand name.. personally I associate Samsung more with TVs, a fridge, semiconductor parts etc..
Western Digital is supposedly a house-hold name for hard drives.

Maybe my next NAS will have Samsung drives. They are definitely a lot cheaper than Seagate and WDC.

/EMBERASSEMENT MODE OFF

[sl1000]

Somebody *please* explain to me why people keep buying WD drives? Samsung makes some very good drives for a nice price. Run rock solid in my boxes...

Buy first, read later (when in trouble)

[hammo]

Somebody *please* explain to me why people keep buying WD drives? Samsung makes some very good drives for a nice price. Run rock solid in my boxes...

Because a lot of people have been trusting WD with their data for over 20 years? I've owned a gajillion drives and have had precisely one WD fail - a Caviar 340MB.

[Avatar]

Yep, but nowadays... my mind goes (with a great embarassment at both thinking and saying it) to Toyota... I will replace my HDs with Samsung ones one by one hoping in the best... Never done before, I am a newbie about arrays stuff and the docs are not so explicative... To stay safe I think I need to make a complete scan for bad blocks on the good drives only then and if all is ok I must put the bad one in degraded mode, turn the NAS off and substitute it, then turn the NAS on again. Is that right?

[sl1000]

No need to turn the nas off and on with the TS-439, you can hotswap them.

[STurtle]

I took out the WD15EARS drive and added a jumper to enable SSC on the drive.

Why would the spread spectrum clocking affect the load cycle count? I thought SSC is mainly used for stress tests of electromagnetic interference?

Or do you mean the Jumper between 3-4 instead, which somehow affects Power Mangement in a way that is unclear to me? Has anyone here any experience with that jumper?


BTW, here are my WD15EADS stats after half a year:

Code: Select all

9    Power_On_Hours      096   096   000   3640    OK
12   Power_Cycle_Count   100   100   000   51      OK
193  Load_Cycle_Count    189   189   000   34605   OK