Jason binger wrote:
>I am currently doing some work with CF MX 6.1 and was
>wondering if anyone had some information on the
>strength of the CF cookie implementation.
>
>How random the token generation is? How is the
>generation performed?
>What is the range of the generated tokens?
>Has an independent security analysis been performed
>and commented on in a public paper?
>
>I have not seen any vendor supplied information on
>this.
>
>Cheers,
>Jason
>
>
>
Sample a number of userids with WebScarab, allow it to perform the
calculations and graph it. Any non-randomness will show up immediately
as a visual indicator.
Note, if the sessionid is something like MD5(time), the sessionid will
appear random to external analysis, but someone with knowledge of the
implementation would still be in a position to brute force session ids.
Regards,
Rogan
Received on Apr 14 2005