bind 9.3.0 performance problem on x86 Linux

[Brad Knowles]

At 11:00 PM +0800 2005-01-10, yan, tommy (CSG-MKG-BJ) wrote:

>  I suspect either the redhat kernel has a scalability
>  problem, or SMP architecture has trouble handling mass
>  dns requests.

	Keep in mind that the Itanium architecture is better designed to 
handle SMP, whereas the Intel x86 architecture has to work harder to 
do that.  There may be kernel support for the hardware architecture 
to make better use of the features available, which could explain the 
performance difference you're seeing.


	Rick Jones (from HP) is the acknowledged master in this field, 
but I've also done some benchmarking in this area myself (see 
<http://www.shub-internet.org/brad/papers/dnscomparison/>), and your 
numbers are about the highest I've ever seen for BIND 9.  Indeed, 
your numbers are the highest I've ever seen for any nameserver on any 
platform (including Rick's testing of Nominum's CNS and my own 
testing of NSD).

	Can you share more details of your testing configuration with us? 
Are you and Rick going to publish another white paper on nameserver 
performance measurement on HP hardware?

-- 
Brad Knowles, <brad at stop.mail-abuse.org>

"Those who would give up essential Liberty, to purchase a little
temporary Safety, deserve neither Liberty nor Safety."

     -- Benjamin Franklin (1706-1790), reply of the Pennsylvania
     Assembly to the Governor, November 11, 1755

   SAGE member since 1995.  See <http://www.sage.org/> for more info.