EncryptByCert Problem

I posted this message to the wrong group (sqlserver.programming), so I'm
reposting here. Sorry for the repost... it's late... so here goes:
OK, so I'm testing EncryptByCert with some code like the following:
DECLARE @.v varchar(8000);
SELECT @.v = REPLICATE('A', 117);
SELECT @.v;
DECLARE @.e varbinary(8000);
SELECT @.e = EncryptByCert(Cert_ID(N'TestCertificate'
), @.v);
SELECT @.e;
The function encrypts fine and I'm able to use DecryptByCert to get the
result. Problem is if I change the line SELECT @.v = REPLICATE('A', 117) to:
SELECT @.v = REPLICATE('A', 118);
The EncryptByCert function returns NULL every time with 118 or higher. BOL
states that the result is returned as a varbinary with a max length of
8,000. I keep getting a varbinary with a max length of 128. Can anyone
else reproduce this, or am I doing something wrong?
ThanksYes, this is a restriction of asymmetric key encryption - the actual data
that you can encrypt with one call will depend on the size of the private
key that is used.
For example: 512 bit RSA key can encrypt up to 53 bytes, 1024 bit up to 117
bytes, and 2048 bit up to 245 bytes.
See http://blogs.msdn.com/yukondoit/arc.../24/496521.aspx for a
workaround, if you really want to encrypt more.
Certificates should only be used to protect other keys, not to encrypt data
directly, so this limit does not impact the intended use of certificates.
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@.xyz.com> wrote in message
news:fuaHg.586$k%3.406@.newsfe12.lga...
>I posted this message to the wrong group (sqlserver.programming), so I'm
>reposting here. Sorry for the repost... it's late... so here goes:
> OK, so I'm testing EncryptByCert with some code like the following:
> DECLARE @.v varchar(8000);
> SELECT @.v = REPLICATE('A', 117);
> SELECT @.v;
> DECLARE @.e varbinary(8000);
> SELECT @.e = EncryptByCert(Cert_ID(N'TestCertificate'
), @.v);
> SELECT @.e;
> The function encrypts fine and I'm able to use DecryptByCert to get the
> result. Problem is if I change the line SELECT @.v = REPLICATE('A', 117)
> to:
> SELECT @.v = REPLICATE('A', 118);
> The EncryptByCert function returns NULL every time with 118 or higher.
> BOL states that the result is returned as a varbinary with a max length of
> 8,000. I keep getting a varbinary with a max length of 128. Can anyone
> else reproduce this, or am I doing something wrong?
> Thanks
>|||Someone needs to tell the BOL writers to update the page on this then. It
currently indicates that EncryptByCert can return a varbinary up to 8,000
bytes in length, which is ludicrous if the input is really limited to 421
bytes and the output is limited to 432 bytes. I'm not advocating using
certificates and asymmetric keys to encrypt all data. I am advocating
accurate documentation of system limitations.
"Laurentiu Cristofor [MSFT]" <laur@.nospam.com> wrote in message
news:e2ASWY6xGHA.4336@.TK2MSFTNGP06.phx.gbl...
> Yes, this is a restriction of asymmetric key encryption - the actual data
> that you can encrypt with one call will depend on the size of the private
> key that is used.
> For example: 512 bit RSA key can encrypt up to 53 bytes, 1024 bit up to
> 117 bytes, and 2048 bit up to 245 bytes.
> See http://blogs.msdn.com/yukondoit/arc.../24/496521.aspx for a
> workaround, if you really want to encrypt more.
> Certificates should only be used to protect other keys, not to encrypt
> data directly, so this limit does not impact the intended use of
> certificates.
> 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@.xyz.com> wrote in message
> news:fuaHg.586$k%3.406@.newsfe12.lga...
>|||I'll ask for an update of these BOL topics to mention the restrictions
related to the key lenghts. If you access BOL online, you can also rate the
articles and submit feedback directly. Here are the links for the topics we
discussed:
http://msdn2.microsoft.com/en-us/library/ms188061.aspx
http://msdn2.microsoft.com/en-us/library/ms186950.aspx
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@.xyz.com> wrote in message
news:%23vuAcc6xGHA.3500@.TK2MSFTNGP02.phx.gbl...
> Someone needs to tell the BOL writers to update the page on this then. It
> currently indicates that EncryptByCert can return a varbinary up to 8,000
> bytes in length, which is ludicrous if the input is really limited to 421
> bytes and the output is limited to 432 bytes. I'm not advocating using
> certificates and asymmetric keys to encrypt all data. I am advocating
> accurate documentation of system limitations.
> "Laurentiu Cristofor [MSFT]" <laur@.nospam.com> wrote in message
> news:e2ASWY6xGHA.4336@.TK2MSFTNGP06.phx.gbl...
>|||Thanks. I wanted to get verification and make sure it wasn't a problem with
my installation or how I was using the functions before I claimed errors in
either the application or the documentation.
"Laurentiu Cristofor [MSFT]" <laur@.nospam.com> wrote in message
news:Ovyex47xGHA.1288@.TK2MSFTNGP03.phx.gbl...
> I'll ask for an update of these BOL topics to mention the restrictions
> related to the key lenghts. If you access BOL online, you can also rate
> the articles and submit feedback directly. Here are the links for the
> topics we discussed:
> http://msdn2.microsoft.com/en-us/library/ms188061.aspx
> http://msdn2.microsoft.com/en-us/library/ms186950.aspx
> 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@.xyz.com> wrote in message
> news:%23vuAcc6xGHA.3500@.TK2MSFTNGP02.phx.gbl...
>|||Also, in the meantime, you may want to use the following reference:
http://msdn.microsoft.com/library/d...ryptencrypt.asp
The relevant paragraph extracted from above is:
The Microsoft Enhanced Cryptographic Provider supports direct encryption
with RSA public keys and decryption with RSA private keys. The encryption
uses PKCS #1 padding. On decryption, this padding is verified. The length of
plaintext data that can be encrypted with a call to CryptEncrypt with an RSA
key is the length of the key modulus minus eleven bytes. The eleven bytes is
the chosen minimum for PKCS #1 padding. The ciphertext is returned in
little-endian format.
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.
"Laurentiu Cristofor [MSFT]" <laur@.nospam.com> wrote in message
news:Ovyex47xGHA.1288@.TK2MSFTNGP03.phx.gbl...
> I'll ask for an update of these BOL topics to mention the restrictions
> related to the key lenghts. If you access BOL online, you can also rate
> the articles and submit feedback directly. Here are the links for the
> topics we discussed:
> http://msdn2.microsoft.com/en-us/library/ms188061.aspx
> http://msdn2.microsoft.com/en-us/library/ms186950.aspx
> 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@.xyz.com> wrote in message
> news:%23vuAcc6xGHA.3500@.TK2MSFTNGP02.phx.gbl...
>