Re: [DNSOP] DNSOP Digest, Vol 139, Issue 7

Viktor Dukhovni <viktor@dukhovni.org> Thu, 07 June 2018 16:19 UTC

From: Viktor Dukhovni <viktor@dukhovni.org>
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
Mime-Version: 1.0 (Mac OS X Mail 11.4 \(3445.8.2\))
Date: Thu, 07 Jun 2018 12:02:35 -0400
References: <mailman.568.1528386376.5672.dnsop@ietf.org>
To: dnsop@ietf.org
In-Reply-To: <mailman.568.1528386376.5672.dnsop@ietf.org>
Message-Id: <E39C1AB4-24B3-414B-846E-64B03601DCD1@dukhovni.org>
Archived-At: <https://mailarchive.ietf.org/arch/msg/dnsop/0vQ-IPOSXKH2L7_rVVuMM-uqiec>
Subject: Re: [DNSOP] DNSOP Digest, Vol 139, Issue 7
Precedence: list

> On Jun 7, 2018, at 11:46 AM, dnsop-request@ietf.org wrote:
> 
> This particular author believes that the DNSSEC should move to ECC,
> so there’s a high cost associated with KSK algorithm rollover. So, if people
> are going to change to “stronger” (whatever this means in DNSSEC context)
> algorithm they should be strongly encouraged to change the algorithm
> to ECDSA256 (for now).

Just wanted to add, that I don't think of 10 as stronger than 8,
there's no evidence of weakness in SHA2-256 any time soon.  I
rather view 8/10 as being two sides of the same coin, RSA with
the two un-truncated SHA2 algorithms, with which to use a "toss-up".
So there's no real advantage to moving to 10, modulo a small
performance improvement in verifying the signatures of "large"
RRsets.  And the fact that RSA-SHA512 is can't be used with
512-bit keys is perhaps a feature.

-- 
	Viktor.

Re: [DNSOP] DNSOP Digest, Vol 139, Issue 7 Viktor Dukhovni