WD20EADS is getting about 21,000 disk and NTFS errors in event viewer per 24 hours. (Win7-64)
For the past few days I’ve been moving files off it at a blistering 0.6-3MBytes/s.
This is a completely new failure mode for me. Usually drives just stop spinning or get squillions of bad sectors. This one is actually copying everything successfully bar 1 file - albeit at about the speed of my internet.
For the first 4 months of its life, it was in a 2k8 server as storage and OS. It was quite fast with benchmarks well above 80m/s. As it failed, I noticed it because videos were glitching. I decided a reboot was in order. It stopped booting.
And here we are. I cant help but think this is some form of controller failure.
It’s like the drive has decided that it is permanently in low speed mode.
RMA in probably another week at this rate. Once it’s empty. Sigh.
You are definitely NOT CRAZY… I recently bought two of these WD15EADS 1.5 TB drives for my MacPro, and experienced the same problems as described here, i.e. some use for a while then incredible slow operation. It astounds me that WD is not owning up to this obvious manufacturers defect in this product line.
Ok, I have the same problem, but found a DECENT work-around: use NTFS file compression!
On XP:
On explorer right click over a drive
Select Properties from the menu
Click Advanced
4 ) Compress contents to save disk space
OK
6) Apply changes to this folder only
OK
My theory is that the drive gets confused with the usual writes, on the normal uncompressed mode. For small files, the system writes at least 4KB, leaves slack space. In short, it can be very inneficient. A good drive knows what to do, this Green ■■■■ doesn’t.
When using NTFS compression, the system gathers multiple files and writes larger bundles, This writes files with no slack, grouped together and then compressed for more savings.
Ex: if I’m copying 80 files of 1 KB each, the system makes 80 write requests, leaving 3KB slack space on each. With compression, it makes just ONE if those 80 x 1KB compress somewhat and fit inside a 64KB bundle, or 2 at most.
Of course there may be innacuracies above, but I suspect it gives a general idea
Same problem with WD15EADS-00P8B0. :cry: I send back the store, but he said there was no error. I regret that not buy 2* 750gig drive. Sry for bad english, i’m Hungarian (eastern Europe).
Another 1.5 tb here, same thing - ive ran a chkdsk for 36 hrs and it found a few things, but it didn’t fix it. I am about to just reformat it -or should I just return it under warranty? This is retarded. I just got this piece of ■■■■ 6 months ago, and have been using it for nightly backups.
I’ve had the same drive for around 6 months, and it was always a dog, until today. It was so slow that all I/O basically stopped after a few minutes of writing to the drive. On occasion, Winduhs would complain about unrecoverable write errors. I was about to ask for an RMA, when I figured I had nothing to lose by trying to play with the jumpers, and reformatting the drive. So, I jumpered 5 and 6 (OPT1 enabled), and formatted the drive with 512B sectors. It now sustains write speeds of 70MB per second (that MegaByte) for several hours, with no indication of slowing down. Unfortunately, I didn’t sectionalize my approach, so I don’t know if the jumper did the trick (probably), or the reformat. In any case, that’s plenty fast enough for me on a system that’s a few year old. (Dual core 2.0 P4 with 4 GB RAM). There’s another jumper that can be used to enable (Pins 1 and 2) SSC (Spread Spectrum Clocking). Don’t know what that might do, but since there is no data that I need on the drive now, I can play with it some more. Let you all know if I can improve the data transfer rate.
OPT1 forces the drive to use the slower SATA 1.5Gbits/s (OPT1) interfacem, which is required for certain SATA controllers that do not properly implement the SATA 3.0 Gb/s speed negotiation. So, I guess that this is what you should try. Works for me. Interesting though that other manufacturer’s drives such as an HDT721010SLA360 which use the SATA 3.0 spec work just fine, and at the same time, in the same system.
I am having the same problem. I have 3 of these drives. My computer began to run very slowly and seem to hang. I thought it was the MB or CPU so I RMAed them. But the problem persisted. I swapped drives and the computer was fast again for about 3 months and then it began to slow down again. It looks like it is a problem with all these drives.
I did nothing special with the drive just a quick format and loaded Windows XP.
Is your jumper fix still working? I have the WD15EADS 1.5TB drive and it began to run slower and slower and now it is dead slow. My jumpers are as follows 1&2 SSC 3&4 PUIS 5&6 enabls1.5GB PHY. Nothing listed for 7&8.
Any sugestions as to what I should jumper? My motherboard supports SATA II 3GB/s transfer rate so jumpering 5&6 dosen’t make sense.
I purchased 4 of the WD15EADS drives for my RAID5 setup. brand new, 1 of the drives exhibited problems, whereby first few minutes, it was fast, then it would slow down, then speed up, then slow down (the data transfer rate). IT was causing my RAID5 to be really slow. When checking, the IO would be at 100% causing slowdowns then drop back down. I isolated it to 1 out of the 4 drives. The 1 drive passed all windows and SMART disc tests. I ended up RMA’ing the drive from WD and the new drive was fine.
Now 6 months later, I’m getting the same problem, I haven’t isolated which drive in my RAID5 is causing this, since I now have ~4 gigs worth of data on the RAID5…
this is a real PITA! I might just go buy a Samsung or Hitachi and sell the RMA’d drive.
I’m not sure in everyone’s case, but in mine, out of 2 purchases WD10EADS drives, one of them in fact has 4096 byte sectors. Several linux utilities tipped me off to it reporting 512 logical 4096 physical, but I had just assumed they were “older versions”. The dog poor write performance had me digging further, and finally just to see what would happen, I downloaded and installed BOTH of the Alignment Tools offered by Western Digital. The Paragon and Acronis software both verify that the one drive is a 4096 byte sector drive and will happily realign a partition on that drive. Doing so brings the performance right to where it should be.
I don’t know if this is a case of WD lying to customers, a mix up internally between different drive platter/board/firmware combinations or what, but I would like to hear what WD has to say. If my drives have 4096 byte sectors and the wrong firmware am I missing out on 32MB of cache that may not be enabled by the firmware? Or some other thing that won’t crop up until say a power outage or system crash and valuable files get lost due to a mixup on the drive’s electronics.
I urge you all to download the latest align softwares and see for yourselves.
