Re: Algorithms: who's at fault here?

Simon Josefsson <[email protected]> Fri, 26 Aug 2022 11:01:43 +0200
Newsgroups gmane.ietf.secsh
Message-ID <[email protected]>
Mouse <[email protected]> writes:

> I recently tried to ssh to a recently-installed Linux machine at work.
> Algorithm negotiation failed.  On turning on verbosity, this turned out
> to be because, to edit the log down to relevant lines:
>
> remote banner string: SSH-2.0-OpenSSH_8.9p1 Ubuntu-3
> ...
> ssh: my algorithms:
> ...
> 	hk: ssh-rsa ssh-dss
> ...
> ssh: peer's algorithms (%=unrecognized, *=disabled):
> ...
> 	hk: %rsa-sha2-512 %rsa-sha2-256 %ecdsa-sha2-nistp256 %ssh-ed25519
>
> Looking at 4253, the only algorithm for this list I see as REQUIRED is
> ssh-dss (though ssh-rsa is RECOMMENDED).  I've gone through the updates
> I can find to 4253 (6668, 8268, 8308, 8332, 8709, 8758, 9142) and I
> don't find anything removing ssh-dss from REQUIRED status.

RFC 9142 seems like the right document to deprecate ssh-dss but it was
forgotten.  If we would update RFC 9142 now, I think it should be
modified as follows:

* MUST NOT ssh-dss

* MUST rsa-sha2-256

* MUST curve25519-sha256

* MUST ed25519-sha256

I think it makes sense to have at least two MUST algorithms for a
security protocol that are based on different mathematical properties.
RSA-SHA256 and 25519-sha256 seems like popular and proven algorithms.

/Simon
signature.asc (application/pgp-signature, 255 B)
-----BEGIN PGP SIGNATURE-----

iIoEARYIADIWIQSjzJyHC50xCrrUzy9RcisI/kdFogUCYwiL9xQcc2ltb25Aam9z
ZWZzc29uLm9yZwAKCRBRcisI/kdFouPJAP9A6t6fsJpz2pOx3gd8ClV6SusPsLOI
ASrFw5aSZ1DMIAEAt9wX6yHCECheCcJ6mzal5+p6WEylqk8Y5HiPhl1kjAo=
=nM5/
-----END PGP SIGNATURE-----