Re: Wiping of data on large storage arrays

["SCHALIP, MICHAEL" <mschalip () CNM EDU>]

Doing a wipe on a complete storage array can be done, but only if you have a week or two to let it run.....possibly 
longer.  In the Fed sector, they sometimes write into the contract that all the equipment *except* for the physical 
hard drives......then they just shred the hard drives.  I've seen one contract where they agreed to provide "like drive 
hardware" instead of returning the hard drives - the contract got brand new drives in return, and we didn't have to 
sanitize anything, (I know - sounds expensive, but the savings in time/labor more than paid for the drives - and the 
level of "sensitive" dictated the extra caution.....).  Another contract just said that the hard drives had to be 
returned - so they put all the drives through a degausser, turned up to "easybake", and called it good.....and the 
contract folks were fine with that.....

Just other options.....

M

-----Original Message-----
From: The EDUCAUSE Security Constituent Group Listserv [mailto:SECURITY () LISTSERV EDUCAUSE EDU] On Behalf Of Jones, 
Dan
Sent: Friday, March 18, 2011 9:38 AM
To: SECURITY () LISTSERV EDUCAUSE EDU
Subject: [SECURITY] Wiping of data on large storage arrays

Hi All, 

I'm searching for options to address the contractual need to purge data at the end of a contract. Many of our contracts 
call for 'secure deletion' of the data owner's data when a contract ends, including issuance of an affidavit to that 
effect. 

A DOD 5220.22-M wipe is simple to do when data is stored on a single disk. Even with early storage arrays, one could 
provision a small set of disks for a project and then wipe the disks at project's end. Similar things apply with backup 
tapes too. 

The verbiage of old-style data destruction requirements does not mesh well when data is stored on more modern storage 
-like an Isilon array (since data will age-out over time and be migrated to slower disks or near-line storage). 

I wonder how others may be addressing this need. 
        - data is spread over so many disks we don't worry about it
        - destroy the encryption keys so the data becomes irretrievable cyphertext
        - something else? 

If the response is significant I'll summarize responses for the list

Vendors need not reply

Thanks,
Dan Jones
ISO
UMass Medical School
dan.jones () umassmed edu


 

--
This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.