 |
 |
Volume failure
|
 |
|
 |
|
Junior Member
Join Date: Feb 2001
Location: Peterborough, Ontario, Canada
Status:
Offline
|
|
The hard drive in my Cube (a Samsung Spinpoint 120 gb) is partitioned into 2 volumes, neither of which had been showing any issues when I decided it was time to do some maintenance. I ran Disk Warrior from Vol 2 on Vol 1, and it showed up all kinds of problems, which it claimed to fix. Checking this with Disk Utility produced some nasty output: invalid node structure; volume check failure; error: the underlying task reported failure on exit. I could run permissions repair, but it took a long time to complete. Further runs with DU and DW continued to report errors, so I finally decided to reinstall the system. The archive and install method reported failure, so I did an erase and install, and this seemed to work. Once I got my system back up to 10.4.6, I ran the utilities again and got more of the same kinds of errors.
So what does this mean? I'm thinking that either there are bad spots on the volume that are not being picked up with the reinstall, or the disk is in the first stages of total failure. If the latter isn't true, I could continue to run from the 2nd volume and ignore the first, or use the first only as a backup. Or I could reformat the volume from Disk Utility and then reinstall the system if this would properly block out bad sectors. I should reiterate that I have had no disk problems other than the bad disk reports. Also, I should mention that we did have a short blackout here a few days ago, and the computer was being used when it occured.
Anyone have a similar experience? Anyone have advice?
(Last edited by mfox; Jun 25, 2006 at 08:59 AM.
)
|
|
|
| |
|
|
|
 |
|
 |
|
Administrator 
Join Date: May 2000
Location: California
Status:
Offline
|
|
What version of OSX are you running, and what version of Disk Warrior? Also, how are your partitions formatted? The more recent options for HFS+ (journalling, HFSX, case sensitivity options, extended attributes) will not be known to older versions of Disk Warrior. Which might cause it to do bogus repairs (Norton does this big time).
Personally, I use Disk Utility to do a Verify on a volume because it's faster than Disk Warrior. If no problems are found, great. If there are problems, I use Disk Warrior to fix them. The latest version of DW is 3.0.3 as of this posting.
In Disk Utility, select the HD icon (not the partition icons) then check the SMART status reported at the bottom of DU. If it says 'verified' then the drive is probably ok on the hardware level. Any other message may mean trouble. 'not supported' would mean the drive is an older one that doesn't have built-in hardware diagnostics.
|
|
|
| |
|
|
|
 |
|
 |
|
Junior Member
Join Date: Feb 2001
Location: Peterborough, Ontario, Canada
Status:
Offline
|
|
I'm running 10.4.6 and version 3.03 of DiskWarrior. The partitions were formatted with HFS+ with journaling. Smart status was verified with DW. I have also now tried making the repair from an external firewire drive; no difference. I even examined and "repaired" the drive from an OS 9 partition (DU and Norton). Both reported fixed bad sectors, but subsequent attempts to put a system (Tiger) onto that volume continued to result in unrepairable disk problems being reported in Disk Utility. Could completely reformatting the disk with DU do anything different than just erasing the volume?
|
|
|
| |
|
|
|
 |
|
 |
|
Administrator 
Join Date: May 2000
Location: California
Status:
Offline
|
|
It does sound like you have developed bad blocks on the first partition.
On SCSI drives, repair utilities handle bad blocks by claiming them. An invisible directory entry is created, which owns each detected bad block. This works, unless you erase the partition, which wipes the directory tree clean and exposes the bad blocks to regular use again. You have to do a low-level reformat of the drive to map the bad blocks out completely.
On ATA drives, repair utilities might handle bad blocks the same way. As for getting rid of bad blocks completely, you can't do a low-level reformat on an ATA drive. Bad blocks are handled by the drive itself, they get mapped out live the next time the blocks are written to. You can force all bad blocks to be mapped out - here is how.
In OS 9 (Drive Setup), check the format options from the menus (low-level format will be greyed out) and checkbox the option to zero all blocks.
In OS X (Disk Utility) go to the Erase tab and click on the Security Options... button. Click the radio button to Zero Out Data. If you suspect some bad blocks are marginal, then try the 7-Pass Erase or if you are really paranoid, the 35-Pass Erase.
Zeroing out a large drive is going to take a long time. Using 7-pass or 35-pass is going to multiply that. Also, the mapping out process has some potential limitations. It works because the drive sets aside a few blocks that are not directly available to the outside world. Then, when a bad block is mapped out, the drive assigns one of the spare blocks to take the place of the mapped-out block. If your drive has had a lot of bad blocks, the spare count may be running low. I'm not sure how to detect this condition, or how to remedy it.
This Apple Knowledge Base article gives some info on how bad blocks are handled on ATA drives.
|
|
|
| |
|
|
|
 |
|
 |
|
Junior Member
Join Date: Feb 2001
Location: Peterborough, Ontario, Canada
Status:
Offline
|
|
Thanks for your suggestion, reader50. It took me awhile to get to it, but I finally followed your advice and used the Zero Out Data option. It didn't work; the reformatted drive continued to be reported as problematic by Disk Utility. I eventually acquired Intech Speed Tools and used it to try something similar, but it has an option to block out bad sectors on a volume level. It claimed to have blocked out three or four bad sectors at the beginning of the volume (and the drive since it was the first volume). But once again after that, I got the same error message from Disk Utility on that volume.
I don't know what it all means since I haven't experienced any problems specific to that volume, but I've given up and replaced the drive with a Seagate 7200.9 160 gb. I don't get any further problem reports from Disk Utility, but I don't notice any difference in performance either. Now I'm not sure whether to treat the Samsung drive as defective and get rid of it, or to use it in some other way.
|
|
|
| |
|
|
|
 |
|
 |
|
Mac Elite
Join Date: Jan 2005
Status:
Offline
|
|
Since the Segate was probably under $60 I would dump the Samsung and not risk any more data... is it still under warranty?
|
|
|
| |
|
|
|
 |
|
 |
|
Junior Member
Join Date: Feb 2001
Location: Peterborough, Ontario, Canada
Status:
Offline
|
|
Exactly what I am doing. The Samsung is under warranty and I just got an RMA to send it back for a replacement. I'll worry about what to do with the extra drive once I get it. And you're right about the Seagate cost. There were a rash of Seagate sales two weeks ago when I was in the States. The 160 I bought will cost me $US40 after rebates! It is surprisingly quiet, even more so than the Samsung it replaced. While the disk stats in xbench suggest that the Seagate is also faster, the Samsung felt faster.
|
|
|
| |
|
|
|
 |
 |
|
 |
|
|
|
|
|

|
|
 |
Forum Rules
|
 |
 |
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
|
HTML code is Off
|
|
|
|
|
|
 |
 |
 |
 |
|
 |
|