sbp commented on issue #214:
URL: 
https://github.com/apache/tooling-trusted-releases/issues/214#issuecomment-3543801148

   QuickLog2 belongs to a large family of audit log schemes that rely on a 
forward-resistant MAC. The first such scheme was, I believe, the 
Schneier-Kelsey one from 1998. QuickLog2 and its successor, Nitro, are 
optimised for high frequency logging in the Linux kernel. We can batch in the 
ATR, so our logging would be low frequency. Therefore, one of the other audit 
log schemes in the same family is likely to be more suitable.
   
   All of these schemes share the downside that the initial seed, _S_, must be 
stored on a trusted verifier server, _T_. There have been alternative schemes 
designed with asymmetric cryptography such as LogFAS, but they are typically 
unimplemented and, as in the case of LogFAS, catastrophic vulnerabilities were 
soon found. The idea of using Rekor in conjunction with a Schneier-Kelsey style 
scheme was that Rekor becomes a kind of surrogate _T_, but this sort of 
construction is unimplemented, and its security properties may be non-obvious, 
and non-trivial to study. This is exacerbated by the fact that Rekor itself 
allows a range of signature types.
   
   The Schneier-Kelsey scheme is quite simple at core, but we should try to 
identify an existing conservative and ideally audited implementation. We could 
in parallel try to find an asymmetric sealing scheme; there are a couple of 
potential candidates in the literature, but it is clear that this is a 
difficult domain.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to