Internal hard drive failing to mount during restart

There are several threads mentioning this issue with certain very large drives.

When the Mac checks them, they may be showing as 'Not ready' and are not just left unmounted, but are not made available for MacOS to use at all. Readers have observed that when this happens, Disk Utility is unable to detect and mount the drive on command.

My theory is that some of these large drives do not spin down fast enough to actually be Ready after a Restart. The only work around I have seen is my own: Retrain yourself. Don't do a Restart, do a Shutdown-and-Start.

If your setup allows it, you could move that drive into an External enclosure, and see if you get different results.

hopefully the drive manufacture [...] will come up with a fix soon.

I would not count on that. Those big drives may be marketed as Server drives, and are operated by most customers on a 24/7 basis.

¿Spin it down? Only to swap it out for a new one.

All sounds good.

Can you remove and reseat it and shuffle the order of the drives to see what that does.

Also try the drive in an external case, or Voyager like dock, and see whether it has the same problem there. It may just be the drive.

Peter

That fix is probably never going to happen. And it isn’t completely clear where the problem is, given that some folks have reported that Snow Leopard can restart and mount large drives without a problem.

The 6TB Toshibas work properly inside my 2010 Mac Pro.

10.6.8 may have different amount of time from chime to looking at additional drives available. The traditional way to do this was to focus on the Boot drive ONLY, until the system was fully up and running. Only then did you look at the other drives and attempt to Mount them. Two things have changed since then.

The first is the addition of Fusion Drive. Now you may need a second physical drive sooner, to combine it with the boot drive to make a Fusion drive. So the timeframe for looking at what used to be strictly "secondary" drives has changed, because they may be more important early on.

The second is the use of Core Storage data structures, most often used for encrypted Volumes, but also allowing the use of Volumes than span physical drives. It is not unreasonable to think that today's MacOS would be looking at the drives much sooner than 10.6.8 would. And more time from chime to look at Drives means more likely to be Ready and to Mount.

SMART status was thought to be really interesting when it was introduced. But is started to become discredited because most drives showed "Verified" until they were dead.

Closer inspection of the individual Counters inside SMART status has shown that there is valuable information lurking there, especially in the two counters you cited, which indicate Bad Blocks are accumulating, and should be dealt with. It appears that Tech Tool has finally awakened to the fact that SMART consists of actionable information other than just "Verified".

If you can get it to tell you How Many blocks are in those two categories, you can better decide what to do about it.

--------

Drives have always been built with spare, unallocated blocks added on. These are only available as substitute blocks. The controller is perfectly capable of permanently substituting a spare for a block that has gone bad, but can only do so when new data are provided for the Bad Block, since the current contents of a Bad Block are unknowable -- it fails to read as good, even after error-correction. Everyone can understand that suddenly inserting a substitute block of Zeroes into one of your files is not the right thing to do.

For a handful of Bad Blocks, re-writing the file or rewriting the entire drive will fix the problem for a while. For hundreds, the drive has already passed its prime, and will likely continue to decline. It should be retired, or used as an secondary Backup.

Tech Tool Pro also contains a disk SCAN function, where it will go out and read every block (it takes many hours). If Bad Blocks are encountered, it will look up was show a list of what FILEs are involved, and list them for you. This can allow you to recover form a handful of Bad Blocks without having to completely ERASE and re-write the entire drive.

An additional wrinkle is that NAS drives seem especially likely to not mount on restart, but some non-NAS WD Black and some HGST helium drives also have been reported here to be problematic. I have an earlier 4TB WD Black that worked fine, but later models were reported here to not mount on restart.

But the 6TB Toshiba in my Mac Pro mounts correctly in Snow Leopard, Mavericks, Yosemite, and Sierra.

Apple software has always used 10 blocks as the maximum number of spares that could be substituted in one pass. However, that number has always seemed somewhat arbitrary to me, and another pass would often complete the sparing, making the drive 100 percent good again.

Google did a very large scale study on their server farms of consumer-grade drives. They found that once a drive accumulates more than a handful of errors, it will (on average) continue to accumulate errors so that it will be unusable in about six months of 24/7 operation, requiring replacement.