Re: [IPsec] revisiting 3DES and -graveyard

[Paul Wouters <[e-mail]>]

On Tue, 21 Apr 2020, Dan Brown wrote:

> [DB] Last week, I looked up what NIST documents say about 3DES.
> https://csrc.nist.gov/publications/detail/sp/800-131a/rev-2/final
> If I read them correctly, this document implies something like:
> - NO: new deployment of 3DES
> - OK: old deployment of 3DES encryption, until 2023, then NO more 3DES
> encryption.
> - OK: old deployment of 3DES decryption (e.g. to decrypt archived stuff).
> Not sure how much IPSec wants to follow NIST.  Presumably they do for 3DES,
> since 3DES is NIST's?
> The text below sounds to me like IPSec is already trying to do something along
> the NIST guidelines. (So, info above I wrote above is already well-known to
> IPSec.)

There is also the SP800-77 rev 1 draft "Guide for IPsec"

https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-77r1-draft.pdf

Which puts 3DES (TDEA) into the "Legacy" category.

It also states:

 	When migrating from IKEv1 to IKEv2, an upgrade of the algorithms
 	used is strongly recommended. 3DES, MD5, SHA-1 and DH Group
 	2 and 5 should not be used.



 	The Triple DES (3DES) encryption algorithm is no longer
 	recommended. It is much slower than AES-GCM and AES-CBC,
 	and it requires more frequent rekeying to avoid birthday attacks
 	due to its smaller block size of 64 bits.

Paul