Re: [hiprg] use of sha-1 in the DHT interface draft

["Henderson, Thomas R" <[e-mail]>]

> -----Original Message-----
> From: Ahrenholz, Jeffrey M
> Sent: Friday, October 21, 2011 8:00 AM
> To: Henderson, Thomas R; [e-mail]
> Subject: RE: use of sha-1 in the DHT interface draft
> 
> > However, my understanding is that the DHT server does not enforce the
> > use of SHA-1; modified clients (that are HIP-aware) could use SHA-256
> > truncated to 160 bits.
> 
> The DHT server does use SHA-1 internally to process removes. If the
> client uses another hash function, the secret given in the DHT remove
> will not match the SHA-1 hash used in the put.

I did not catch that detail since it is not very clear from the API document, but it makes more sense that it operates that way.

> 
> The draft does not currently require a modified OpenDHT/Bamboo server.
> It states in sections 3 and 7 that the server MAY/SHOULD be modified
> for signature and host ID verification (it looks like the "MAY" in
> section 3 should be changed to "SHOULD".)  Use of another hash
> algorithm would require further server modifications (and specifying an
> interface that deviates from OpenDHT.)
> 
> > - keep the draft as is, citing that OpenDHT interface prescribes SHA-
> 1,
> > and that the other usage of SHA-1 is not trying to obtain any crypto
> > properties.  However, the implication of this is we will need to add
> > more text in the security section and elsewhere to defuse this "red
> > flag"
> 
> I would prefer the above option. The security considerations section
> can include text about deviating from the OpenDHT interface in order to
> support other hash functions. The original intent of the draft was to
> re-use the freely-available OpenDHT interface in an interoperable
> fashion.
> 

In light of the above, I would lean towards your recommendation as well.  

- Tom