Re: [security-services] Fwd: SAML Conformance SSL/TLS requirements

[Eric Tiffany <eric.tiffany@ieee-isto.org>]

On 8/15/05 6:08 PM, "Prateek Mishra" <prateek.mishra@oracle.com> wrote:

> Eric Tiffany had asked the following question about conformance
> specification
> 
> [quote]
>> I have a question about sections 5.1 and 5.2 of the SAML 2 conformance
>> doc. These sections place requirements on "TLS-capable implementations",
>> "FIPS TLS-capable implementations", etc., regarding required cipher suites.
>> 
>> What do "TLS-capable", "FIPS TLS-capable" mean? I know what TLS and FIPS
>> are, but who determines that an implementation is one or the other or
>> both? Isn't the choice of cipher suite more of a deployment issue, and
>> not something that SAML should define normatively? Whether a SAML
>> implementation supports a particular cipher suite would seem to depend
>> on the HTTP/SOAP webserver or appserver, not on the SAML code itself. I
>> think this places some unsupportable requirements on SAML library
>> implementors who may not control how their otherwise-conformant
>> implementations are deployed.
>> [end-quote]

Just to emphasize,  I am really asking two questions:

1)  The terms "FIPS TLS-Capable" and "TLS-Capable" are not defined.  What
does this mean, precisely.

2)  Why are the cipher suites indicated as MUST?

also more below...

> 
> All of Section 5 describes conformant use of SSL or TLS when deployed as
> a part of a SAML implementation. [Rescorla]  has the following text:
> "SSL supports a variety of cipher suites, specifying the set of
> algorithms used for the connection. These algorithms vary from very weak
> exportable ciphers such as RC4 in 40-bit mode to (hopefully) very strong
> ciphers such as 3DES. [...] It is therefore necessary to choose a cipher
> suite commensurate with the value of your data".

And this sort of contradicts your subsequent statements.  If you should
"choose a cipher suite commensurate with the value of your data", then SSTC
should not be dictating a particular set of ciphers which may not be the
best for a particular situation.


> 
> This is precisely the goal of Section 5. We are recommending use of
> specific cipher suites for SSL/TLS that are generally  accepted to be of
> adequate strength. It should be possible to configure SAML products
> using SSL/TLS with these settings, and, indeed a conformance test should
> include such a test.

Well, you are not "recommending", you are REQUIRING, which is my main point.
I believe the language in this section should be relaxed to say these cipher
suites are RECOMMENDED, rather than MTI.

It isn't part of the SAML specification, it's purely a transport layer issue
that (imho) is out of scope for these specs.

ET

> 
> The FIPS piece of it has to do with additional qualification for
> implementations that are required to meet standards set by the federal govt.
> 
> - prateek
> 
> [Rescorla]  SSL and TLS, Designing and Building Secure Systems,
> Addison-Wesley, 2001.
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe from this mail list, you must leave the OASIS TC that
> generates this mail.  You may a link to this group and all your TCs in OASIS
> at:
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
> 

-- 
____________________________________________________
Eric  Tiffany             |  eric@projectliberty.org
Interop Tech  Lead        |  +1 413-458-3743
Liberty Alliance          |  +1 413-627-1778 mobile