Re: SQL 2005 - Searching Encrypted SSN

If you use a fixed salt, then you compromise on security.

See http://blogs.msdn.com/lcris/archive/2006/05/08/592868.aspx for an
illustration.

If you use salting, then the second encryption will produce a different
result, so you cannot use an index over encrypted data.

Raul's blog describes a solution and provides some sample code:
http://blogs.msdn.com/raulga/archive/category/12301.aspx, but the solution
requires some changes on the application side to take advantage of the
index.

Thanks

--
Laurentiu Cristofor [MSFT]
Software Design Engineer
SQL Server Engine
http://blogs.msdn.com/lcris/

This posting is provided "AS IS" with no warranties, and confers no rights.

"Ron Brent" <RonBrent@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:12577962-6176-465E-86F1-5B791A893C78@xxxxxxxxxxxxxxxx
Hello Laurentiu,

Do you think that it is possible to do the following when the encryption
is
salted?
"2) write your own encryption/decryption functions (not too difficult with
a
SQLCLR hosted .NET System.Security.Cryptography namespace)"
What I mean is - how would the second encryption (that is done by the
function within the query) produce identical cipher text results to the
first
encryption given the fact that the encryption algorithm uses a random
initialization vector?

In your opinion, what would be the best way to do what I want as quickly
as
possible and without compromising security?
Are you familiar with any 3rd party database encryption solutions that
uses
different IVs and perhaps can solve my problem?

Thanks,

Ron

"Laurentiu Cristofor [MSFT]" wrote:

Please also note that RC4 use is not recommended. Do not use RC4 just
because it's unsalted and allows you to index the encryption for equality
searches. While it may be convenient, it is not secure.

Thanks

--
Laurentiu Cristofor [MSFT]
Software Design Engineer
SQL Server Engine
http://blogs.msdn.com/lcris/

This posting is provided "AS IS" with no warranties, and confers no
rights.

"Mike C#" <xyz@xxxxxxx> wrote in message
news:uaH7JLgmGHA.4100@xxxxxxxxxxxxxxxxxxxxxxx
That was the kicker. Unless you can specify your own salt and IV (and
I
don't have SQL 2K5 installed here at work, so I can't test it), you'll
either:

1) have to perform decryption on every column and compare,
2) write your own encryption/decryption functions (not too difficult
with
a SQLCLR hosted .NET System.Security.Cryptography namespace), or 3)
Store
a hash of the encrypted data. If you only need the SSN for
identification
purposes, and not for reporting purposes, you could probably hash the
SSN
and forget about encryption altogether.

BTW, according to BOL, SQL 2005 does not salt the encryption performed
with RC4 and RC4_128.

"Ron Brent" <RonBrent@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:82A6C044-195A-403A-8BD5-E9CF2F163035@xxxxxxxxxxxxxxxx
Hi Mike,

According to Laurentiu's blog -
http://blogs.msdn.com/lcris/archive/2005/12/22/506931.aspx -
"The encryption algortihms in SQL Server 2005 are salted. By salting,
I
mean
that the encryption algorithms are always using a random
initialization
vector (IV), which leads to the following property: encrypting twice
the
same
piece of data using the same key will produce two different
ciphertexts."

You say that theoretically I should be able to encrypt the SSN I'm
searching
for and compare that to the already-encrypted columns.
How would the second encryption (that is done within the query)
produce
identical cipher text results to the first encryption given the fact
that
the
encryption algorithm uses a random initialization vector?

Thanks,

Ron

"Mike C#" wrote:


"Ron Brent" <RonBrent@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:1929F115-F507-4448-9AD7-1F900B9A1F3F@xxxxxxxxxxxxxxxx
Thanks Uri!

It will be helpful once we decide to implement the encryption, but
my
question is this -
is it possible to search an encrypted column the way it's described
in
the
blog and get the results quickly (at the same speed it takes to
search
on
a
clear text)?

If you want to search using = in the WHERE clause, you can get
comparable
speed. Just don't decrypt the SSN's in the column. Theoretically
you
should be able to encrypt the SSN you're searching for and compare
that
to
the already-encrypted columns. Of course this won't work if your
column
looks like this:

999-99-9999

And you want to search on last 4 digits or something similar. In
that
case
you need to decrypt and compare row-by-row (unless you split the SSN
up
into
separate columns...).










.

Relevant Pages

  • Re: SSN encryption
    ... > state of issuance of SSN. ... You're overestimating the ability of encryption to solve the problem. ... > the database, that may be much harder. ... hole will break your security. ...
    (sci.crypt)
  • Re: SSN encryption
    ... Andrew Swallow wrote: ... > Assuming the researches are using their own computers and that they will ... > Use AES encryption to encrypt all zeros to produce 128 random bits; ... > using the secret key variable and an IV equal to the SSN. ...
    (sci.crypt)
  • Re: SQL 2005 - Searching Encrypted SSN
    ... specify their own salt/IV values for encryption in the future? ... probably hash the SSN and forget about encryption altogether. ... "The encryption algortihms in SQL Server 2005 are salted. ... for and compare that to the already-encrypted columns. ...
    (microsoft.public.sqlserver.security)
  • Re: Search encrypted column in SQL 2005http://msdn.microsoft.com/w
    ... As HashBytesdefeats the purpose of the Random IV generated during ... encryption process, I'll look at Raul Garcia's proposed MAC indexing method. ... For a full string exact match (like a full CC# or full SSN) you can do as ... Erland suggested and use HashBytes() to store hashes of the encrypted data ...
    (microsoft.public.sqlserver.security)
  • Re: SQL 2005 - Searching Encrypted SSN
    ... don't have SQL 2K5 installed here at work, so I can't test it), you'll ... purposes, and not for reporting purposes, you could probably hash the SSN ... and forget about encryption altogether. ... for and compare that to the already-encrypted columns. ...
    (microsoft.public.sqlserver.security)