Re: Pseudorandom Flow Labels

[Thomas Narten <[e-mail]>]

Case in point about how we are being *extremely* loose in using the
term "pseudo random".

If you look at draft-gont-6man-flowlabel-security-01, it proposes two
different algorithms for generating Flow Label values.

Unless I'm missing something, neither of them actually provides
"pseudo randomness". Both have subsequent Flow Label values simply
increment a counter to get to the next value. What these algorithms do
do is make it very hard for an off-site attacker to guess what Flow
Label values are being used. That is probably a good thing. But that
is not the same thing as saying that the Flow Labels that are
generated are "pseudo random", which is what the current Flow Label
documents are saying we need.

Part of my objection to the term "pseudo random" is that the term has
not been defined within the context of the Flow Label.

RFC 4086 "Randomness Requirements for Security" talks quite a bit
about pseudo randomness. I don't think we need that for the Flow
Label. But without carefully defining terms, that is what is
presumably being required...

Thomas