Disk Utility "Restore" Failing

Question

Level 1

0 points

Disk Utility "Restore" Failing

I use a couple of firewire partitions (on different drives) to keep bootable
clones of my system, dumping the system alternately to one or the other
partition. Both partitions are set up to be marginally larger than my internal
HD. To do this, I boot to one of the current clones. Then I use the Disk Utility
"restore" tab, erasing the destination, to copy the internal HD to the alternate
clone partition.

Up till now, it has been working OK. However I am having troubles backing up
now. When I ran the restore process to create the clone, it was failing with
"Operation not permitted", about half an hour into the copying process. This
happened I think three times. So I ran Disk Utility / First Aid on the source
disk. The disk verified OK, but did come up with one file permission error,
which it repaired (though this is puzzling in itself, I would have thought file
permissions would be ignored in cloning, since it does a block copy)? Disk
Utility and Disk Warrior now both report the internal drive to be perfectly OK.
(they also give the empty firewire partition a clean bill of health). Anyway,
my subsequent attempt to clone failed twice in a different way:
Restore Failure
An error (5) occurred while copying. (Input/Output error).
Any suggestions on what this might mean? Bad firewire controller? Bad
sectors on the disk? In either case, is there any way to get a more informative
diagnostic?

1.33GHz G4 Powerbook, Mac OS X (10.4.7), Toshiba 60GB internal, Sarotech 120GB firewire, Sarotech 300GB firewire

Posted on Sep 3, 2006 5:54 PM

Reply

Answer 1

Best reply

Dr. Smoke

Level 9

56,586 points

Sep 3, 2006 11:10 PM in response to Robert McKay

Hi, Robert. Welcome to the Discussions.

1. You wrote:

"Restore Failure
An error (5) occurred while copying. (Input/Output error).
Any suggestions on what this might mean"

Just what it says. 😉 Unfortunately, those error messages are almost entirely undocumented.

2. You wrote:

"Bad firewire controller?"

2.1. Do the FireWire drives operate nominally otherwise?

2.2. Tried a different FireWire cable or different FireWire port on your Mac?

3. You wrote:

"Bad sectors on the disk?"

That usually produces a very specific, audible clue. See the "Bad Sectors" section of my "Resolving Disk, Permission, and Cache Corruption" FAQ. The terms "bad blocks" and "bad sectors" are used synonymously, but it is the sectors that go bad. Blocks are comprised of sectors.

There could be bad sectors on the FireWire partition. Try "zeroing" one of the FireWire partitions as described in my FAQ.

4. You wrote:

"...is there any way to get a more informative diagnostic?"

Maybe:

4.1. Immediately after the problem occurs, check the ends of these Console logs for clues.

4.2. TechTool Pro 4.5.x has some FireWire tests, but these are not for peripherals, but more related to the FireWire bus in your Mac. May not be definitive.

4.3. To rule out other hardware issues, run the Apple Hardware Test (AHT) that came the affected computer. See my "Apple Hardware Test" FAQ for comprehensive advice on using the AHT in troubleshooting.

5. IMO, the backup method you're using is very time consuming. As you know, each time you use the Restore approach, the entire drive is copied, which can take more than a few hours if you're backing up a good bit of data. You'd be better served by considering a proper Backup and Recovery solution that can both create bootable Duplicates and perform incremental backups: in second an subsequent backups of a given source drive, only the changes are copied, syncing the source drive with the backup. For advice on the backup and recovery solution I employ, see my "Backup and Recovery" FAQ.

6. FYI: Here's a tip: When asking multiple questions, it helps to number them individually. Then someone can respond "by the numbers" vs. having to quote your text to assure a match between your question and the answer.

Good luck!

😉 Dr. Smoke
Author: Troubleshooting Mac® OS X

---
Note: The information provided in the link(s) above is freely available. However, because I own The X Lab™, a commercial Web site to which some of these links point, the Apple Discussions Terms of Use require I include the following disclosure statement with this post:

I may receive some form of compensation, financial or otherwise, from my recommendation or link.

Reply

Answer 2

Robert McKay Author

Level 1

0 points

Sep 5, 2006 6:28 AM in response to Dr. Smoke

Firstly, thank you for the very helpful information. It gave me a lot to think about (and check), but unfortunately hasn't (yet) got me to a solution of the problem. I'm still chasing some of the directions you suggested, but thought I should at least give some response rather than leaving things up in the air. However I've already gained one benefit, in finding out about Troubleshooting OS/X, of which I wasn't previously aware. Nice stuff.

1. You wrote:
"Restore Failure
An error (5) occurred while copying. (Input/Output
error).
Any suggestions on what this might
mean"
Just what it says. 😉
Unfortunately, those error messages are almost
entirely undocumented.

Hmmm. I'm starting to suspect that the underlying problem may be that the error message isn't only undocumented, but misleading. The error occurs (as nearly as I can tell from the progress bar) at the same point every time (sample size around 10/10). That doesn't sound like an I/O error. It sounds more like some sort of problem with reading a specific part of the source partition (it could also be a problem with writing a specific location on the destination partition, but I've subsequently zeroed the partition, so it should presumably have any bad sectors mapped out). However as I mentioned, neither Disk First Aid nor Disk Warrior see any problems with the source disk. Next two checks are:
.apple hardware test
.trying a clone to a spare USB drive I have
will reply with more detail after that.
Thanks and Best Wishes
Bob

12" 1.3G Powerbook Mac OS X (10.4.7) 2*firewire drives (Sarotech) chained (because 12" PB only has one firewire port)

Reply

Answer 3

Dr. Smoke

Level 9

56,586 points

Sep 5, 2006 8:49 AM in response to Robert McKay

You wrote:

"...finding out about Troubleshooting OS/X, of which I wasn't previously aware. Nice stuff."

Thanks! 😉

You wrote:

"That doesn't sound like an I/O error. It sounds more like some sort of problem with reading a specific part of the source partition (it could also be a problem with writing a specific location on the destination partition, but I've subsequently zeroed the partition, so it should presumably have any bad sectors mapped out)."

Any problem reading from (I for Input), or writing to (O for Output), a disk would be classified as an I/O Error (Input/Output Error = I/O Error).

What makes generic messages such as "error(5) (I/O Error)" so frustrating is that they don't provide details, e.g. "Did the I or O fail and why?" For example, a failed attempt to read-from or write-to RAM could also produce an I/O error. It's a generic message lacking additional context. For example, if I were programming and testing my program resulted in an I/O error in a specific line of code, I'd have some context for debugging my program.

You wrote:

"However as I mentioned, neither Disk First Aid nor Disk Warrior see any problems with the source disk."

That means the directory of the source disk is in good shape. Those tools are concerned with the directory structure of the disk, a different issue from bad sectors. To confirm bad sectors, one needs resort to tools like the Surface Scan feature of TechTool Pro, per my FAQ cited in point 3 of my earlier post. If you're not hearing the sounds I described in the FAQ, then it's unlikely there are bad sectors on the source disk, but only a Surface Scan could confirm this.

Obviously, if bad sectors were occurring in the portions of the disk where the directory is written, Disk Utility's First Aid function or Disk Warrior, would fail, but you'd also have the audible indications of bad sectors — repeated chugging or grinding — as the drive attempted to read the bad sectors.

For example, the directory might specify that a file begin and end at certain sectors on the drive. But Disk Utility First Aid and Disk Warrior won't find a bad sector in the middle of that range of sectors. They're concerned only with the directory structure, i.e.the "address book" and "map" of where files reside on the drive, not if the file itself has bad sectors in the middle of its allocated blocks.

Good luck!

😉 Dr. Smoke
Author: Troubleshooting Mac® OS X

Reply

Answer 4

Robert McKay Author

Level 1

0 points

Oct 29, 2006 3:24 AM in response to Dr. Smoke

Thanks again for all the help. I never did hear any repeated seeking (but my hearing is straining to hear the PB disk running anyway). In the end, the problem went away when the 10.4.8 update came out, so I assume that some badly mis-structured file or directory (not detected by disk utility or disk warrior) got overwritten in the upgrade. Anyway, after 10.4.8 installed, disk cloning with disk utility started working again.

Once I had a reliably backed-up system, I was prepared to risk checking out the disk fully with Techtool pro. It found no problems in a surface scan, but it did detect a finder problem with one other file, which turned out to be a system file from an ancient MacOS (7.5.something from memory) that I keep for resurrecting ancient files on a 68K emulator (I still have the odd macdraw and macwrite file from my original fat mac!). I'm guessing that there was some file corruption just before my previous HDD failed, and this got into the backups from which I restored the new HDD, causing the subsequent problems. Needless to say, I'm now checking out my backups much more carefully.

All three utilities are now 100% happy with the disk, my backups are working properly, and I'm feeling comfortable. I guess it's time for something else to go drastically wrong...

What makes generic messages such as "error(5) (I/O
Error)" so frustrating is that they don't provide
details, e.g. "Did the I or O fail and why?" For
example, a failed attempt to read-from or write-to
RAM could also produce an I/O error. It's a generic
message lacking additional context. For example, if
I were programming and testing my program resulted in
an I/O error in a specific line of code, I'd have
some context for debugging my program.

Yes, it does seem pretty strange that more info. isn't given. Especially since, in other contexts, OSX does give more informative I/O error messages. So is this a case of very low-level code generating an exception, which higher level code is supposed to catch and expand on (because it knows more context), but disk utility isn't catching? That also seems a little strange, because if any application ought to catch and explain I/O errors properly, it's disk utility.

Anyway, thanks again for your careful assistance

Best Wishes
Bob

Reply