Replies inline.

On 7/10/20 7:50 PM, Brian Toresdahl wrote:
> Your approach and goals don't seem to make sense to me.

TL;DR: The customer is always right, and the customer sees DKIM being used 
regularly to authenticate leaked emails - if
old not-in-use keys are public, anyone can sign anything they want, and 
suddenly you can't authenticate mail with them,
at least after-delivery, that is.

> DKIM works (I'm summarizing a lot) by signing mail, and the public key to 
> check that signature is placed in a "selector"
> record in DNS (selector1._domainkeys.example.com 
> <http://domainkeys.example.com>). If you want to rotate DKIM keys, you
> can immediately start signing new mail with with another selector 
> (selector2._...), keeping the selector1 around long
> enough for the old mail to be delivered and checked. Are you asking how long 
> selector1 needs to stay around? For some
> sort of 95th percentile, probably less that an hour.

Yes, that was my question, thanks :). Sadly, I don't think and hour would work 
- plenty of time to timestamp the emails
in question and suddenly the email is non-reputably signed again. Back to the 
drawing board, I suppose.

> But this seems excessive. The time time to brute force a 1024 key
> is in the years (someone correct me), so to get faster an adversary has to 
> get inside your systems and steal your
> private keys directly, and this conversation is then a moot point.

The goal is to actually publish the private key, no need to wait for someone to 
brute-force it :)

> But realistically, most folks in practice don't rotate selectors often. 
> Anecdotally, I remember a certain John L.
> rotating them every month, and that's the most frequently I've heard 
> personally. When I managed this for Amazon SES, I
> think the policy was 6 months or a year.
> 
> On Fri, Jul 10, 2020 at 4:39 PM Matt Corallo via mailop <[email protected] 
> <mailto:[email protected]>> wrote:
> 
>     For various reasons, DKIM's non-repudiation property has always prevented 
> us deploying DKIM signing on our mail. The
>     obvious fix for this is to roll DKIM keys aggressively (eg every few 
> minutes) and publish the private keys for revoked
>     keys as you go. Given relay times for mail through various hosts, how 
> aggressive could one theoretically get without
>     having keys time out before mail is verified (ie how long must DKIM keys 
> remain valid after mail has been signed by
>     them?).
> 
>     Separately, is there any work towards making DKIM support non-RSA keys? 
> Sure, one could always just generate a fresh key
>     every minute and publish all old ones, but ECC would be nice as it would 
> allow simpler hash-based key revocation where
>     you can reveal every previous key with just one key.
> 
>     Thanks,
>     Matt
> 
>     _______________________________________________
>     mailop mailing list
>     [email protected] <mailto:[email protected]>
>     https://chilli.nosignal.org/cgi-bin/mailman/listinfo/mailop
> 
> 
> 
> -- 
> 
> Brian Toresdahl
> 
> Product Management
> 
> [email protected] <mailto:[email protected]>
> 
> 
> 

_______________________________________________
mailop mailing list
[email protected]
https://chilli.nosignal.org/cgi-bin/mailman/listinfo/mailop

Reply via email to