Almost Constant Disk Errors, why?

Question

I am not sure what is going on, but in the last two weeks, I have had to re-start my MAC Pro 5 times with my Snow Leopard Program disk and use the Utility in that disk to correct disk errors. The problem always is the same "Invalid volume directory count". There are times when my machine locks up for a period of time and I can't do anything with it, and then, it starts working again, and nothing seems to be going on when it locks up (nothing else is running in the background). I also get frequent error code 10810 "The operation can't be completed" messages. I have run Tech Tool, and there are no apparent errors. I have not been able to create these problems, they seem to have a mind of their own. Has anybody else run into this?
Thank you.

Answer 1

SMART status of the drive after one of these events may give you more insight into what is happening.
Hard Drives have redundancy bits in each block and error correction logic, and will also try to re-read bad blocks many, many times in hopes of getting correctable data the next time. This can cause the huge delays you are seeing. These errors will NOT be seen if you do a Disk Utility ( Repair ) or a Tech Tool Repair. These options check ONLY the directory for consistency, and do not read the actual data blocks of your files.
The test that will demonstrate these errors is a "Surface Scan", which reads every block on the drive (and may take hours to complete). On certain drives with certain bad but not deadly errors, the problem shows itself as the program appearing to stall, seemingly "stuck" on a particular block as it reads it again, and again, and again. Eventually it may get good data and go on to the next block, or may give up on that block and increment the Bad Block counter.
With this many errors, your best bet at this point may be to backup everything and re-initialize with Zero all Data option. This coaxes the drive to substitute its reserved spare blocks for any found bad in the process.
If it finishes successfully, your drive has 100 percent good blocks again.
If it gets more than ten bad blocks, you may get "Initialization Failed". You can try it again, but there is a limit to how much you can trust a disk that is developing so many bad blocks.
Extended tests done by Google researchers on a huge sample of consumer level "always on" drives used in Google's server farms indicated that having at least one Bad Block occur in use was a strong indicator that the drive was likely to develop more Bad Blocks and fail, as much as 40 percent more quickly than drives with no Bad Blocks.
Drive temperature was NOT a leading cause of death. They do not generally cycle the power on their drives nor spin them down, so those did not figure in their potential causes of death.