ietf-nntp TLS cipher renegotation to NULL cipher

[Andrew Gierth]

>>>>> "Russ" == Russ Allbery <rra at stanford.edu> writes:

 Russ> Okay, that's a fair cop.  Premature optimization is the root of
 Russ> all evil.  Has anyone actually benchmarked SSL/TLS with a large
 Russ> news server and discovered a substantial performance impact?

in my case (going by openssl's rc4speed.c, I haven't tried this in
anger) it would at least double the amount of frontend CPU power
required, and CPU is the limiting factor there. i.e. I would have to
deploy twice as much hardware for that part of the system.

-- 
Andrew.