[hipl-users] Re: HIP base exchange times
- From: Miika Komu <miika@xxxxxx>
- To: hipl-users@xxxxxxxxxxxxx
- Date: Wed, 28 Jun 2006 23:58:41 +0300 (EEST)
On Wed, 28 Jun 2006, Väisänen Teemu wrote:
Has someone seen similar behaviour in HIPL:
A <----> B
A is an initiator and it has lower processing power than B that is a
responder. HIP base exchange times are smaller than in a similar
situation with the same machines and connections, but where the B is
the initiator and it has more CPU processing power than A that is the
responder.
What might cause this?
Thank you for any answers!
Could it be somehow related to this:
nosig/hmac --- I1 ---> no verifying
verify signature <--- R1 --- signature
sign and hmac --- I2 ---> verify signature + hmac
verify signature and hmac <--- R2 --- sign and hmac
===============================sum============================
1xsig, 1xhmac, 2xverify sigs, 2xsig, 1xhmac,1xverify sig
1xhmac verify 1xhmac verify
Remember also that dsa signing is faster than dsa, but veryfying is
faster. Or maybe it was the other way around. In addition, Tobias said
that HIPL signatures might slower than expected (buzgilla id 197) due to
some byte order swapping.
--
Miika Komu http://www.iki.fi/miika/
Other related posts:
