feedback requested

Stefan Schmidt zaphodb--bind at
Thu May 6 09:45:59 UTC 2004

On Thu, May 06, 2004 at 01:35:39PM +1000, Mark Andrews wrote:
> 	If you have tried the latest rc/beta could you just drop a
> 	note on how it went.  Silence doesn't alway indicate success.
> 	The output of "uname -a" and whether build were successful
> 	or not is what we are after.
I have 8.4.5rc1 running loadbalanced behind Linux IPVS (direct routing,
weighted round robin) making use of the -4 switch. I can say that it runs
smooth at a sustained peek rate of 430 recursion queries/s and i
am not seeing any differences at interupts, user-, system-cpu or
memory usage compared to 8.4.4-P1 on the other machine.
8.4.5rc1 is just running smooth as far as i can tell if we don't take
the high cpu usage on SERVFAILs into account that is.
The 8.4.4-P1 machine is running Linux 2.6.2-rc1, the 8.4.5rc1 is on
2.6.6-rc2-bk5, both Compaq Proliant DL360 G1 2xPIII 800 running debian
unstable with package libc6-i686 aka glibc 2.3.2 compiled for i686 plus
NPTL support.

"I thought we saw the worst of it; the Shadows, the Vorlons, the war. But
there's something far worse than the Shadows: reporters!"
- Sheridan, "The Illusion of Truth" (B5)

More information about the bind-workers mailing list