New found hatred for Western Digital....

[Topcat]

Bought 2x 160gb WD1600AAJS hard drives, these are the WD "BLUE" series. One for a GX620USFF and the other for a customer's iMac G5 that had a bad hdd.

Neither of those systems will detect these pieces of shit. I plug them into a sata2 controller (3ware), they detect fine. Both the GX620 and the iMac are both sata1 interfaces, but the drives claim backward compatibility.....but still don't detect. Pretty pitiful. Yes, I tried the jumper trick to lock in the speed at 1.5, no avail. I connected a 500gb WD Black to each machine mentioned above, drive shows up and no problems.

Doesn't anyone make a good hard drive anymore?!?

[weirdlookinguy]

Re: New found hatred for Western Digital....

I guess not :/

I've also heard that their Caviar Black drives are rigged to drop out in RAID (they want you to buy their more expensive RE RAID-specific drives). There used to be a utility that would change a setting on the drive to make it not drop every few days but the newest Caviar Blacks are locked and the utility doesn't work anymore.

Damn shame really. My older (can still be unlocked with the utility) Caviar Black 640 has been great so far.

[Agent24]

Re: New found hatred for Western Digital....

I just upgraded the HDD in one of my PCs to a 500GB Seagate.

The motherboard is an Asus K8N with nForce 3 250 and only supports SATA-150

When I put the new drive on in the board wouldn't complete the POST, blank screen, CD-RW drive light started flashing really weird then it would reboot itself and finally complete POST and detect the drive.

I jumpered to 1.5Gbit and it worked fine, just like the WD 250GB before it.

I hear now that some drives you must configure the settings in firmware with the manufacturer's own special software. Though I think this may have been for Samsung or Hitachi. Either case, maybe worth looking into?

[Topcat]

Re: New found hatred for Western Digital....

^
That makes sense to the point of when I plugged in a 500gb SATA2 WD black, all test systems that wouldn't see the 160gb WD blue WOULD see the 500gb WD black, and the blacks were actually newer drives (manufactured about a month after the blues were). Both were SATA2. The black didn't even need to be jumpered. This whole thing is just retarded. I'll save the drives for use on something that likes them, but there's no excuse for this crap.

wierdlookinguy,

You're referring to the TLER (or lack of it) on the blacks that were "RE". That too was another sham by WD to stick it in people's asses. For running TLER drives in a RAID, you could simply run a utility and disable it in the firmware, and shazam, no timeouts or broken arrays. The RE and non-RE drives were IDENTICAL, except this little piece in the firmware. WD got wise to this hack and removed the ability to disable TLER from non-RE drives all together, so they could charge more for the "RE" drives. If that's not ripping people off, I dont' know what is.... Like I said, ridiculous.

TLER is time limited error recovery. When a sector becomes unusable, the drive will reallocate it to a spare. This tends to happen on dying drives (obviously if sectors are failing). If the reallocation would take more than I think 8 seconds, the drive would drop from the array and break it. If you are running them in a stripe, that's the end of your RAID and your data. If in a mirror, you'd have to rebuild the array every time it hiccuped. I too learned the hard way about this some time back. I only buy enterprise grade stuff for my personal use now.

[Agent24]

Re: New found hatred for Western Digital....

Well I wouldn't want a drive with any reallocated sectors anyway.

I got burned once when a 40GB drive had 1 reallocated sector. Worked fine for a few months then one day Windows took forever to boot, eventually hung and upon restart the drive failed to read anything.

Head failure or preamp etc, I don't know. Then I read that study Google did, how after the first reallocated sectors drives are significantly more likely to die in 30 days following the error than keep working reliably.

That was the end of it for me, I now have a personal zero-reallocated-sector policy which is why I replaced the 250GB Western Digital, it has 2 reallocated sectors...

Still trying to decide if I should try to RMA or just sell it in "possibly about to die" condition (I'm guessing WD won't accept RMA unless the drive is actually dead\on fire\sounds like a bench grinder)

[Topcat]

Re: New found hatred for Western Digital....

Well I wouldn't want a drive with any reallocated sectors anyway.

Absolutely correct, that's what I was getting at actually....when the drive is reallocating sectors, its on its way out....but in the case of TLER and RAID, it would kill the array before you had a chance to do anything about it (especially a stripe). I'm not a fan of stripe arrays at all, too risky regardless of what drives they are. However, when shoddy drive design will cause a mirror array to drop out and end up in a degraded state, it totally defeats the purpose of having the raid in the first place. The array will break before you even know that you have sectors crapping out.

[Agent24]

Re: New found hatred for Western Digital....

But what is the advantage (in RAID 1) of the RAID edition drives with TLER?

If the drive is failing and you replace it and then you have to rebuild the array anyway, right?

If I am understanding this wrong please correct me...

I was thinking of maybe making a RAID 1 array for myself one day but using standard drives (because they're cheaper!)

[Topcat]

Re: New found hatred for Western Digital....

But what is the advantage (in RAID 1) of the RAID edition drives with TLER?

If the drive is failing and you replace it and then you have to rebuild the array anyway, right?

If I am understanding this wrong please correct me...

I was thinking of maybe making a RAID 1 array for myself one day but using standard drives (because they're cheaper!)

Well, it depends on the controller, but in some cases when this happens it hoses the array completely. This is more true with onboard nvidia controllers/chipsets, I've seen it happen a time or two. However, on a good true hardware raid controller such as an adaptec or 3ware, a raid1 will maintain the integrity of the remaining drive on a mirror even if the other drive drops out of the array, giving you time to save your data and/or correct the problem....but on low-end controllers, even though it's a raid1, once the controller sees the array is gone, it whacks everything...including your data. There's nothing to rebuild at that point. If the data on it was critical, you're stuck trying to recover it forensically or with some other recovery software, whereas on a good controller, you replace the defective drive, rebuild, and you're up and running again.

In the sense of whether the array will fail with or without TLER, no, there's not much difference as far as knowing that you'll end up rebuilding the array when there's truly a failure....but TLER has been known to cause drives to drop out of arrays for no particular reason on earth, hence killing the array for nothing. My last run-in with TLER did just that. One of the drives would drop out of the array from timeouts, had no bad sectors at all.

[Colt45ws]

Re: New found hatred for Western Digital....

I think what hes saying is the drive would drop out before you could get a spare or make sure the data is backed up. I know you are talking about a 1 but how long does a RMA take? If it can recover quick enough the RAID does not drop out you have time to react to say an automated SMART checker email or something...
I have 5 Seagate 7200.11s with 4 of those in a RAID 5. Ive had 3 die and last time I pulled the lone one out of the machine it was in so it could rebuild. Ive decided the warranty replacement is going to sit in its box in my closet until the next one inevitably fails. I wont use it for anything bt a cold spare..
Ive noticed the RE vs Standard thing with WD too. It annoys me but they will probably get my business for the replacement drives for this array after I get tired of replacing these Seagates. Probably go with RE too. fuckers.

Well, it depends on the controller, but in some cases when this happens it hoses the array completely. This is more true with onboard nvidia controllers/chipsets, I've seen it happen a time or two. However, on a good true hardware raid controller such as an adaptec or 3ware, a raid1 will maintain the integrity of the remaining drive on a mirror even if the other drive drops out of the array, giving you time to save your data and/or correct the problem....but on low-end controllers, even though it's a raid1, once the controller sees the array is gone, it whacks everything...including your data. There's nothing to rebuild at that point. If the data on it was critical, you're stuck trying to recover it forensically or with some other recovery software, whereas on a good controller, you replace the defective drive, rebuild, and you're up and running again.

In the sense of whether the array will fail with or without TLER, no, there's not much difference as far as knowing that you'll end up rebuilding the array when there's truly a failure....but TLER has been known to cause drives to drop out of arrays for no particular reason on earth, hence killing the array for nothing. My last run-in with TLER did just that. One of the drives would drop out of the array from timeouts, had no bad sectors at all.

Now that I know about it, this is probably what killed my first and probably last FakeRAID on a AMD chipset. It was just randomly screwed to hell. All the drives checked out good.
I purchased a Highpoint 3510 after that and its been good to me. Makes a heck of a screaming racket when a drive drops out, too.

[AudioCraZ]

Re: New found hatred for Western Digital....

In all honesty, it is the pickiness of the mobo. You can look to see if there is a flash update for the system if it wont read the HDD. The problem is, with newer drives, using parallel platter design, there has to be support by the controller to use them. Everybody knows that Apple would rather you buy a new system, then keep an older one alive more than 3 years, hence why they won't offer any updates, where as standard PC boards will offer updates via bios flashes.

[kaniki]

Re: New found hatred for Western Digital....

^
That makes sense to the point of when I plugged in a 500gb SATA2 WD black, all test systems that wouldn't see the 160gb WD blue WOULD see the 500gb WD black, and the blacks were actually newer drives (manufactured about a month after the blues were). Both were SATA2. The black didn't even need to be jumpered. This whole thing is just retarded. I'll save the drives for use on something that likes them, but there's no excuse for this crap.

A few months back i had a friend that needed a new drive for his gateway laptop. From the factory it had a sata 2 drive in it. I got one of those sata 2 white label ones that was a relabel from a WD drive. Had the same exact problem. It would test fine in one system with a full read scan and all, but not in the system I got it for. Ended up getting a seagate for it and that worked with no problem.

[mockingbird]

Re: New found hatred for Western Digital....

Yea I have had problems with SATA 300 drives being recognized on old SiS Sata chipsets, but now you're saying this is also a problem with ICH5? I thought Intel would be better than this.

For a good read on the TLER debacle, check out this thread over at Storage Review.

Yea Western Digital are scum and they have always been scum. Their Caviar drives sucked, and their "AB/BB/JB" drives series always had bearings that made the drive sound like a jet engine after a few years of use.

It's a bloody shame Fujitsu was screwed over By Cirrus Logic. Their drives were mechanically superior to anything else out there in every way - albeit always a little slower than the competition. But that Cirrus Logic chip on the later AT series and on the AH (7200RPM) series, would crash and burn after a few months, now Fujitsu only does SCSI drives.

[NxB]

Re: New found hatred for Western Digital....

I dunno, I've had good experiences with WD. Even recovering drives that tested bad. Then again I don't run raid. The jumper to sata 1 always worked for me.

[Agent24]

Re: New found hatred for Western Digital....

It's a bloody shame Fujitsu was screwed over By Cirrus Logic. Their drives were mechanically superior to anything else out there in every way - albeit always a little slower than the competition. But that Cirrus Logic chip on the later AT series and on the AH (7200RPM) series, would crash and burn after a few months, now Fujitsu only does SCSI drives.

Their HDD division was sold to Toshiba a year ago...

[Sudum Gumudu]

Re: New found hatred for Western Digital....

^
You're referring to the TLER (or lack of it) on the blacks that were "RE". That too was another sham by WD to stick it in people's asses. For running TLER drives in a RAID, you could simply run a utility and disable it in the firmware, and shazam, no timeouts or broken arrays. The RE and non-RE drives were IDENTICAL, except this little piece in the firmware. WD got wise to this hack and removed the ability to disable TLER from non-RE drives all together, so they could charge more for the "RE" drives. If that's not ripping people off, I dont' know what is.... Like I said, ridiculous.

TLER is time limited error recovery. When a sector becomes unusable, the drive will reallocate it to a spare. This tends to happen on dying drives (obviously if sectors are failing). If the reallocation would take more than I think 8 seconds, the drive would drop from the array and break it. If you are running them in a stripe, that's the end of your RAID and your data. If in a mirror, you'd have to rebuild the array every time it hiccuped. I too learned the hard way about this some time back. I only buy enterprise grade stuff for my personal use now.

Why is it, when you need to do your homework, you fail to. Had I known about this, I would never have purchased four WD Black 640GB drives for my nwe computer build to use in a RAID 10 array on the MB onboard Intel ICH10R.
There has been nothing but problems. Constant rebuilds, which take forever on such a large array, and always a SMART event which needs to have the disk status reset. I'm sick and tired of it.

[Junk Parts]

Re: New found hatred for Western Digital....

I just got a PC with SATA onboard. Now I need to read up on this RAID stuff. Here at my house all we have ever used RAID for is to kill ants.

[c_hegge]

Re: New found hatred for Western Digital....

I had topcat's problem with Sata300 drives not detecting on a Sata150 controller. I got it to work using the jumpers to force it to 150. I also had the problem with an 80GB Seagate HDD. It only seems to happen on certain controllers. I was having the problem with the onboard SATA on VIA KT600 motherboards, but the same drives worked fine without jumpers on Intel onboard SATA150.

[AudioCraZ]

Re: New found hatred for Western Digital....

In all honesty, WD is my favorite drive. I am currently running 750 Blacks in my system, and all of my WD's, including my 36 raptor lasted way past the warranty. Personally I have never found a system that wouldn't accept one, but I know there are going to be problems with all drives with some controllers, it just happens. There are so many brands of hardrives, and models, that once and a while, support doesn't get written into the boards for every drive. It just happens.

I would say that Seagate also makes a good drive, and Maxtor have always been the tanks of the industry.

[Agent24]

Re: New found hatred for Western Digital....

I had topcat's problem with Sata300 drives not detecting on a Sata150 controller. I got it to work using the jumpers to force it to 150. I also had the problem with an 80GB Seagate HDD. It only seems to happen on certain controllers. I was having the problem with the onboard SATA on VIA KT600 motherboards, but the same drives worked fine without jumpers on Intel onboard SATA150.

The VIA VT8237 southbridge which would most likely be on your KT600 board are known for problems with SATA-300 drives.

[weirdlookinguy]

Re: New found hatred for Western Digital....

Why is it, when you need to do your homework, you fail to. Had I known about this, I would never have purchased four WD Black 640GB drives for my nwe computer build to use in a RAID 10 array on the MB onboard Intel ICH10R.
There has been nothing but problems. Constant rebuilds, which take forever on such a large array, and always a SMART event which needs to have the disk status reset. I'm sick and tired of it.

How long ago did you buy your drives? If it was a few months ago you may be able to run the TLER patch on them.