Low Level format tool prob

[buster80]

trying to blitz a couple of maxtor 80gb SAS drives

Low level tool + Wipe tool thinks they are a little bigger than they are and thus does not format them

Windows sees them fine.

any alternative?

thanks

andy

[Vulcan]

trying to blitz a couple of maxtor 80gb SAS drives

Are those the drives attached via the RAID HBA? If so, have you tried using a dumb SAS HBA instead? Or have you checked the RAID HBA BIOS - some which I've used have erase functionality built-in. The problem that you mention would fit with these drives being attached via the RAID HBA, which "reserves" some space on the disk, and does not allow OS access to that space.

[Edited to add: Other people have mentioned using Windows diskpart to erase disks - I've not tried this, as I use *nix, but you may want to investgate that. Alternatively, you can use the Windows port of dd or even dd under *nix to achieve the wiping of the OS-accessible part of the disks.]

[Michael.Reilly]

Could it be possible that a HPA is set, and your low level format utility is checking for total LBAs, but some of your LBAs are hidden in the HPA.

If that's the case you can remove the HPA using hdparm

[loki]

You could try winhex to write 00 to the drive sectors - just a thought.

Diskpart uses the secure erase command built into the drive which I belive became standand in drives since 2001 but is still worth a try.


Loki

[Vulcan]

@loki:

Diskpart uses the secure erase command built into the drive which I belive became standand in drives since 2001 but is still worth a try.

Interesting - as I mention above, I haven't used diskpart, but are you sure that it uses the "secure erase" command? That's not what MS seem to say:

http://support.microsoft.com/kb/300415

Use the clean command to remove partition or volume formatting from the current in-focus disk by zeroing sectors. By default, only the MBR or GPT partitioning information and any hidden sector information on MBR disks is overwritten.

That suggests a zero-fill. No mention of "secure erase" by MS.

Also, FYI, the standard "secure erase" command is not available in SAS / SCSI / FC disks (only (S)ATA) and the OP says this is a SAS disk Furthermore, I have not seen any RAID volume which responds to a "secure erase" command issued to the volume (which could be what is presented to Windows and hence what is available to diskpart). Therefore IMHO it makes sense that diskpart does a zero-fill, in order to maintain compatibility with any attached disk device.

What evidence is there that diskpart uses "secure erase"? Thanks.

[loki]

It could be possible that the wording secure erase used by general users is different to what it means in DR terms.
If that is the case then I conceed

http://www.techguider.com/tutorials-cle ... 67978.html

Clean All A Disk With Diskpart Command
How to Clean or Clean All a Disk with the Diskpart Command. How to use the clean command on a selected disk to delete all of its MBR or GPT partitions, volumes, and any hidden sector information on MBR disks is overwritten. The data on the HDD is not written over using the clean command. It is only marked as being deleted and can be written over when new data is written/saved to the HDD next. You can use the clean all command (secure erase) to do the above and also have each and every disk sector on the HDD written over and zeroed out completely to sanitize and delete all data on the disk. This option takes longer to do, but is the best option to help ensure that the data on the disk has been removed and not just marked as deleted.

http://www.sevenforums.com/tutorials/52 ... mmand.html

This will show you how to use the clean command on a selected disk to delete all of it's MBR or GPT partitions, volumes, and any hidden sector information on MBR disks is overwritten. The data on the HDD is not written over using the clean command. It is only marked as being deleted and can be written over when new data is written/saved to the HDD next.

OR

You can use the clean all command (secure erase) to do the above and also have each and every disk sector on the HDD written over and zeroed out completely to sanatize and delete all data on the disk. This option takes longer to do, but is the best option to help ensure that the data on the disk has been removed and not just marked as deleted.

Loki

[Vulcan]

It could be possible that the wording secure erase used by general users is different to what it means in DR terms.

Thanks for the links & quotes from those websites. My interpretation of them is that they are saying the result of a diskpart "clean all" action is a secure erase by overwriting, but is not saying that the "secure erase" command is specifically used, as far as I could see.

However, they are certainly rather ambiguous in their language, and IMHO it is likely that they are unclear themselves, since from the outside, unless there is an HPA (and perhaps in some other corner cases), the result can look the same between a zero-fill of the user area and a secure erase (i.e. they can both look like a disk which is zero-filled). Of course for a typical SSD, the internal result will be very different.

For the reasons I gave about compatibility, I doubt that "diskpart" uses the (S)ATA "secure erase" (strictly called the Security Erase Unit) command. There's yet another reason why diskpart is not likely to that IMHO - most BIOS send a Security Freeze Lock command before the OS boots, which (deliberately) makes it very difficult to later send a Security Erase Unit command, requiring an ATA hard reset (or, for some disks, requiring a power-cycle), before a Security Erase Unit would be accepted.

I'll keep my eyes open for any opportunities to find out the result conclusively

[guru]

built in secure erase is not safe at all

[buster80]

active@ killdisk for windows seems to be doing the trick, it was beginning to do my head in.

Tell me this, if it does a zero fill, why the need for a 3 pass?

thanks for all help

andy

[buster80]

yip, first pass on the free version all zero, why the need for multiple?

[Vulcan]

active@ killdisk for windows seems to be doing the trick, it was beginning to do my head in.

Which is doing basically the same thing (with the same limitations) as what I suggested in my edit on Aug 14th (but you didn't reply to that ).

Tell me this, if it does a zero fill, why the need for a 3 pass?

No-one said there was a need. I'm fairly sure that one of the various government specs required a 3-pass overwrite as a minimum. However that requirement was likely based on the same, now obsolete, disk technology as the original recommendations in the famous Gutmann paper on the subject (which has since been updated, to explain how the original recommendations were mis-interpreted and have become outdated, with new disk technology).

[buster80]

Which is doing basically the same thing (with the same limitations) as what I suggested in my edit on Aug 14th (but you didn't reply to that ).

Sorry bro, my forum email alerts were going straight to spam for some reason

thanks again for your help

andy

[buster80]

likely based on the same, now obsolete, disk technology

makes sense, cheers