The Gebesse Blog

Thoughts from the world of technology and business

Product review – Supplier-provided disk repair programs.

I recently had what I think is the very first Blue Screen of Death I’ve seen since I started using Windows 7. It was not caused by the operating system, but seemed to result from a problem with the computer’s video hardware. Whatever the fault was, it caused the computer to reboot without warning. As I had several programs running at the time there was the real possibility that one of them had been writing to a hard disk at the instant of failure so checking disk integrity was called for. Windows can’t check the boot disk while the system is running (because there are always programs and services running with open files), so when you ask for a check you get this:

Chkdsk warning

The next time you start the computer it checks and repairs the disk before Windows starts. The trap here is that a 500 gigabyte disk with more than half a million files on it takes a long time, so you end up looking at a screen like this for a couple of hours.

Chkdsk running

When I finally regained productivity I thought it might be possible to find something that would do the job in less time. I have disks from Seagate and Western Digital, so they looked like the obvious places to get what I wanted. Both have free disk inspection and repair software, and here are the stories.

Seagate SeaTools

Seagate have a very attractive web site with a nicely understated design. I found the relevant page and after agreeing to Seagate’s user agreement I was told that I could download the software. The only problem was that there was no link on the page to actually start the download. Nowhere. I even looked at the page source to see if I could locate the link, but it was what could politely be called a dog’s breakfast. Eventually I located a link at download.com and installed it.

The first thing that appeared when I started the program was this informative screen:

SeaTools licence?

Almost immediately followed by:

SeaTools crashing

Remarkably, clicking on the Cancel button did nothing except make the dialog box go away. I was then able to accept the mystery box and the program started.

SeaTools starts

Do you notice how there is no serial number or firmware revision for the last disk in the list? They were there the last time I ran this program. And you know how Windows uses drive letters? Reading from the top, the drive letters are C:, D:, W:, F:. That makes sense – who cares about alphabetical order.

When you click on a disk in the list the “Basic Tests” icon lights up, and when you click on that you get:

Basic Tests - options

But what do those options mean? Let’s click on “Help”.

SeaTools Help?

This is really starting to make me feel confident. I tried “Short Drive Self Test” on one drive but cancelled it after a very long period of inactivity. I then tried “Short Generic” and it ran in a few minutes but didn’t tell me anything other than it had finished. I have no idea how to actually run a test that will fix errors, and I don’t think I want to try. My data is too important.

And one last thing. You can see above that there is a “System Tools” icon. The first option this gives is to check the status of your antivirus software.

Anti-virus?

I could have sworn that I renewed my licence for Norton AntiVirus just a few days ago. Perhaps I should check.

Why, that looks like an anti-virus program!

Seagate SeaTools appears to have no use at all except to show that bad programming isn’t confined to amateurs and small companies that don’t have an international presence and a reputation for manufacturing high quality electronic components.

I just thought – who writes the drivers and firmware for Seagate disk drives? Perhaps I should be worried.

Western Digital Data LifeGuard Diagnostics

At least I was able to get this from the Western Digital web site, so it is already ahead of the Seagate product. This is what you see when you run the program:

Western Digital Data LifeGuard Diagnostics

Again the drives are listed in C:, D:, W:, F: in the top panel, although they are listed alphabetically in the bottom panel. Why this should be so is a mystery. You might notice that the two fixed drives have a S.M.A.R.T. status of “Not available”, but the two USB external drives have passed the test. (Self-Monitoring Analysis and Reporting Technology is a system developed by major disk drive manufacturers to improve reliability.) Remember this set of options for the Seagate test program?

Basic Tests - options

The “S.M.A.R.T.” option only appears when one of the internal drives are selected, and when you run the test you see:

Oh, look - S.M.A.R.T.

So the Western Digital program thinks that the USB drives have S.M.A.R.T. capability and the internal drives don’t and the Seagate program thinks the opposite. (And did you notice that the Serial Number and Firmware Revision for drive F: have come back?) The WD program returned what appeared to be valid S.M.A.R.T. configurations for both USB drives, but somehow I don’t feel inclined to believe it.

When selecting the tests to run on a drive the following box appears:

Western Digital tests

At least the Help worked so I could find out how what the options meant, but the Western Digital program inspires no confidence. When repairing disks it stops on every error and if the user says “Fix that” it simply overwrites the damaged bits with zeros, thus destroying damaged files completely. This is not what the user really wants. The “Write zeros” option (which overwrites everything on the drive) might be useful if a drive is to be scrapped, but there are better ways of making a drive unreadable.

The verdict.

Zero stars!Neither program is of any use if you want to detect and correct errors on your hard disks. Neither of them do anything beyond scanning the surface and any repair is rudimentary and possibly disastrous. I’ve added their scores together and the total is zero stars.

It looks like I’m back to Chkdsk. I’ll just have to keep some books handy to read while it does its work.

In case anybody cares:

Seagate web site

Western Digital web site

 

If you liked this post, please subscribe to our RSS feed! You can also follow us on Twitter here.

7 thoughts on “Product review – Supplier-provided disk repair programs.

  • Nick Andrew says:

    CHKDSK checks the filesystem integrity; the WD and SeaTools programs check the disk drive sectors directly, i.e. they detect only physical errors reading or writing the disk.

    What does that mean? If you have a bad sector then the WD or SeaTools programs will find it. Disk drives have a set of spare sectors and when you write data to a bad sector, the disk drive will re-map that sector address to a new location somewhere on the disk, effectively repairing it from the computer’s point of view (but not necessarily yours).

    Writing zeroes is all that program can do, to fix a bad sector. If there’s a read error then whatever was in that sector is gone and you will never get it back. You might as well write zeroes onto the sector, because the file or directory that used that sector (if any) is now corrupt.

    CHKDSK as I said earlier, checks your filesystem integrity – it understands the layout of a FAT or NTFS filesystem and checks that directories have valid contents and the file allocation table contains reasonable data and so on. It can’t check the content of your files at all.

    Since you had a software error which caused a crash, it is highly unlikely (dare I say impossible) for that event to have caused a physical error on your drive. So CHKDSK should have been your first and only port of call.

    SMART when enabled will regularly monitor your drive and detect problem scenarios (i.e. imminent failure). It’s not perfect, but it can give you some assurance that your data’s still intact when you use it on really large drives.

    Most filesystems DON’T CARE MUCH about data integrity. If you had a bad sector and it was one of those assigned to a file (say your resume) and you used SeaTools to fix your error by writing zeroes into that sector, your FAT or NTFS filesystem has no idea that anything went wrong, because it has no checksum of the data in that file, and no redundancy in its data storage from which it could recover that file. YOU are expected to keep backups of all your files, but how are you to know that that particular file has been corrupted if the system doesn’t tell you? You could easily not notice, and maybe overwrite your backup of that file eventually.

    For the above reasons I recommend use of RAID-1 whenever possible: use 2 drives and they contain the same data. If one drive fails then you still have your data and replace the failed drive as soon as possible which will copy all the data onto the replacement drive. RAID-1 can also repair sector errors by copying the original contents (from the other drive) rather than zeroes. It’s not a guarantee that your data won’t be lost by disk drive failure, but it’s a lot better than the default situation, which is to do nothing and hope daily that your drive doesn’t die, taking all its files with it.

    I don’t know what different filesystems are available (if any) on Windows, but on Linux a couple of quite sophisticated filesystems are available: ZFS (a commercial development by Sun Microsystems before Oracle bought them), and btrfs.

    ZFS protects against silent data corruption with self-healing, and has inbuilt RAID, snapshots and clones, subvolumes, deduplication, encryption, copy-on-write and more. Due to the licensing it can’t be distributed with linux, but can be downloaded separately.

    btrfs provides some similar capabilities, including device balancing, inbuilt RAID, copy-on-write, transparent compression, checksums on data (so you’ll know which files are corrupt, at least!), subvolumes and more. It’s part of the linux kernel.

  • Harry Phillips says:

    Check out Spinrite, I have had much success with it.

  • @Harry

    I’ve been using SpinRite for years. Unfortunately in this case Murphy’s Law struck and it wouldn’t run from the bootable CD I had. I suspect that FreeDOS wasn’t happy with my hardware configuration because everything stalled at a display of “InitDisk”.

    I’ve now got a bootable USB thumb drive with Win98 MS-DOS on it and that works properly. I’ll be running SpinRite on all my drives tonight after I’ve finished doing all the things I need a computer for.

  • @Nick

    I tested these two programs because they were specifically recommended on several web sites as alternatives to CHKDSK. I didn’t believe it then and I don’t believe it now. I always assumed they would be rudimentary surface scanners with no knowledge of the actual file system and I didn’t expect them to be able to fix the sort of integrity problems that CHKDSK can work on.

    My concern is that people might believe what they see on “answers” and forum sites, use these programs, and end up worse off than they were before. I have to admit that even with my low expectations I was surprised at how useless they are.

    My approach is always CHKDSK first, then SpinRite. Yes, SpinRite is essentially a surface scanner but it seems to have a remarkable ability to recover stuff so I assume it uses checksums and any other clues it can find. I’ve used it to recover disks that CHKDSK refused to look at and claimed were not formatted.

    The file system I use is NTFS which is almost certainly the most used system of its kind in the world. It’s been around since the early 90s and has evolved over the years. It’s the only choice for Windows, but Linux and OSX can use it too.

  • Nick Andrew says:

    I think you’re right to use CHKDSK first; it’s kinda scary to realise that the vendor-supplied diagnostic/repair tools could repair bad sectors on the disk and leave you with no clue which specific file was corrupted.

  • Andy Farkas says:

    Chkdsk *can* check the content of your files!

    -Start an administrative command prompt (right-click the icon and click Run As Administrator…)
    -Type in “chkdsk c: /f /r /v” and answer yes to schedule. Reboot.

    Chkdsk will now go through 5 stages, the additional 2 stages being to read and check file content and then check unallocated space.

  • Andy – same thing from the dialogue box if you get there by right-click on drive, select Properties then Tools and Check Now and tick both boxes. Mine went through the five stages.

Leave a Reply to Peter Bowditch Cancel reply

Your email address will not be published. Required fields are marked *