Re: [v6ops] draft-cc-v6ops-wlcg-flow-label-marking-00.txt

On Fri, Jul 7, 2023, 10:28 AM Dale W. Carder <dwcarder@es.net> wrote:

> Thus spake Brian E Carpenter (brian.e.carpenter@gmail.com) on Fri, Jun
> 30, 2023 at 10:50:24AM +1200:
> >
> > I'd never thought about using a Hamming code to generate flow labels.
> > But since a given flow must use the same flow label throughout, you can
> > only use the regular 5-tuple as input to the Hamming code. I don't
> > believe that a Hamming code will give you a discrete uniform
> distribution,
> > but I'm not an expert in that area. An expert known as ChatGPT says:
> > "No, a Hamming code does not produce a discrete uniform distribution...
> > The non-uniformity of the distribution arises from the fact that Hamming
> > codes prioritize error detection and correction capabilities over
> > uniformity of codeword distribution."
> >
> > So I think that simply using a stateless hash is the best way. These
> > days I like FNV (draft-eastlake-fnv) which is easy and cheap to
> implement.
>
> I think our idea was that it may be possible to explore other encodings to
> do something more best-effort towards a uniform distribution.  As our
> usecase of the flow label is for telemetry, the encoding must be
> reversible,
> which likely rules out a lot of approaches.  It might be good to find
> someone
> with more knowledge in this area.
>

Dale,

I believe what you're looking for is a cryptographic hash function. Encrypt
to map to a uniform distribution, decrypt to get the original bits.
Googling it, I did find some work on using CRC as a reversible hash that
might be relevant.

Tom

> Dale
>
> _______________________________________________
> v6ops mailing list
> v6ops@ietf.org
> https://www.ietf.org/mailman/listinfo/v6ops
>