[bind10-dev] recent performance improvements
Jeremy C. Reed
jreed at isc.org
Mon Mar 19 22:59:11 UTC 2012
On Mon, 19 Mar 2012, JINMEI Tatuya / ???? wrote:
> I don't understand this:
>
> > bind-9.9.0 root-memory-.success 50519.012123 qps
>
> Does this mean you ran bind-9.9.0 (built with threads) without
> specifying -n? Or does this mean you specified "-n 1"? If it's the
> latter I suspect there's an error in the test. If it's the former,
> the comparison is misleading because it uses multiple threads.
I did not specify -n.
> When we want to compare bind9 with a single-process bind 10, we need
> to specify '-n 1' for bind9 (or more preferably, build bind9 without
> threads because it would still use other threads running on other
> cores for network I/O and timer events).
I didn't originally plan to compare BIND 10 with bad performing BIND 9
also. But now I am doing that also.
More information about the bind10-dev
mailing list