Mac Pro hard drives going bad in one bay

Have you tried some data recovery software? You can use the Google search tool in this regard. .My Mac Book Pro hard drive has crashed and I'm in the process of replacing it. However, I had a lot of important data on that hard drive that he needs recovered. Then I found uFlysoft data recovery software that recovers data from corrupt/bad hard drives. And fully supported for NTFS/FAT32 mac files.

There are stories about extra large drives that do not mate with the backplane harness and are invisible. There are stories about extra large drives not initializing properly.

There was recently a story about what appeared to be bad backplane connectors for Bay 3 and 4. (That turned out to be loose connection where the SATA harness plugged onto the motherboard.)

But there are no stories I remember about drives in a bay that worked for a while, then stopped, that did not turn out to be Bad Drives.

So I expect you are the victim of remarkable Bad Luck.

What is symptom of the bad drives?

- Can you hear/fell it spinning?

- Does it show in Disk Utility?

- Does it show in Profiler under SATA?

Your power supply is not overloaded. Green Drives are not appropriate for daily use for anything except Backups.

Clicking sounds from a desktop drive are sometimes caused by a "Recalibrate" operation. This should not be routine, but it could be happening under certain circumstances.

Data in data blocks on a drive are written with some extra bits, in a semi-redundant way. The extra bits included so the a very small burst of errors in a block can be dealt with readily. Each and every block is read back through error-correction circuitry, which can tell whether there has been an error, and whether it has been corrected. Uncorrectable errors are dealt with by re-trying the read, often as many as 100 times. This generally does not move the drive head, and is fairly fast and noiseless. The error is reported up the software chain, and it will re-try again, often another 100 times (so this may now be 100*100 re-trys). The error is then reported further up the software chain, and often a recalibrate command, essentially seek track 0, is issued. Track 0 on many drives has an independent sensor that reports "I am at track 0 now". Everything is now in a known state, and the re-trys begin again.

Recalibrate, Seek track 0, is the thing likely to cause the 'click'.

Blocks that are Bad after this many retries may not come clean with further re-trys. The drive has a mechanism to substitute spare blocks, but it can only be invoked when NEW data are supplied for the Bad block. But they generally produce "I/O Error" (which is not just a random error, but means the data could not be read after thousands of Re-trys.

Back in the day when drives cost over US$1000, you would attempt to rescue an errant drive by writing Zeroes to every block. This provided new, predictable data. Any block that needed error correction immediately after writing was declared Bad by the drive controller, and a block from the substitutes pool (the drive controller's private stash) was substituted in an essentially permanent way. If the drive passed this test with none or only a few substitutions, it was declared 100 percent good again, and could be re-purposed for serious work again.

A drive that used too many (Apple defined this as more than 10) substitutes got "Initialization failed!". You could run the write Zeroes one more time if needed, but beyond that, it was not Prudent to put that drive back in general use. Perhaps it could be used to store a Backup once in a while. Otherwise it could not be trusted with your precious data and would be retired.

Note that Write Zeroes (one click off the default in "Security Options") to a normal-sized disk today takes all afternoon. But if you run it in a Running System (as opposed to Recovery Mode) it can work in the background. (Also, Disk Utility can work on multiple drives if you merely open a new Window inside Disk Utility.)

Spare blocks and that method of defect mapping and substitution has been in Hard Drives since they were first manufactured. It was originally used only for "factory" defects, and it was invented to avoid having to manufacture a drum or disk surface that was absolutely perfect. A few small defects could be dealt with by substituting spares, and a 'pretty good' disk platter could be made acceptable instead of scrapped.

Manufacturers quickly realized that a few blocks might also go bad in the field, and the 'substitute spares' mechanism was extended to include "User" defect mapping as well.

A new drive just purchased may report (if you were to read out the drive controller's defect map) that it has some substitutions as issued by the manufacturer. That is still considered a top quality drive in perfect condition.

A drive that has just been issued should have no defects that develop in the field. If the drive cannot be made right by writing Zeroes during the warranty period, that is Definitely a warranty claim. If it does develop defects during the warranty period, that might be cause for a warranty claim.

Most Users are not going to scrap a drive with a bad Block or two. They will Zero the drive (invoking the substitute spares mechanism) and see if it will "come clean" and continue to work.

Manufacturers will replace Bad Drives. But it can take several weeks. Re-writing with Zeroes can be done in an afternoon.

>>erasing with zeroes is all I want to do.<<

Correct.

Hard drive technology pushes the physics of magnetic regions right to the edge, than backs off a tiny bit until they can hit their reliability numbers. (If they thought they could get more bits in without completely losing reliability, they would push closer to the edge).

Very subtle changes in the magnetic material over time, as it heats and cools, can cause the magnetic regions to squirm around a little and can cause a few bits to flip polarity, causing errors. The built-in error-correction can compensate for a few bits, but occasionally too many errors accumulate in a block and it goes Bad. Re-writing it with a different pattern or the same pattern is sometimes all it needs to be good again for a long time.

Although it seems odd that you have more than one drive go bad, maybe these are from the same batch, and there were some impurities in the magnetic stuff that day. Or maybe some of the other chips are out of spec and causing marginal data to show up as bad.

The manufacturer cranks these drives out by the thousands. You have a sample size of only a handful. But you cannot really draw ANY conclusions from your small sample size except that you have some drives that are going wonky on you. When it takes all afternoon just to write data to the drive, manufacturers are not doing the long burn-ins that used to cull "infant mortality" out of the product.

You should think of each drive as a coin-toss. It is no more or less likely to be good or bad based on the previous one. If you think it might be a bad batch, you could try a different source for the drives, or try a different brand.

What are the odds on 3 of the six drives being bad and all of them on the same drive bay?

The Mac uses drive names, not drive Bays as the identifier. So you can power down your Mac, shuffle the drives around randomly, or slide each drive to the right one bay, and power up with no change in anything.

If the bad drive follows the drive bays, you may have a marginal connection. The wiring to those bays is on a wiring harness that plugs in to the motherboard at one end. But this is almost unheard of. Typical problems reported here are drive does not work (at all) in any Bay, or Bay 3 does not work (at all) with any drive.

Drive errors are unlikely to be anything except drive problems. Bay problems seems to cause complete failure.