bind-9.3.2 / CPU issue.

[Daniel Ryslink]

Hello,

We run several BIND 9.3.2 caching nameservers in a query-intensive 
environment (ISP), and have experienced exactly the same problem you 
describe.

I have found a workaround by setting the max-cache-size parameter to a 
very low value (something around 150MB). That way, the server falls 
periodically into that state when almost 100% of CPU is used by BIND, but 
there is no slowdown on the DNS responses.

This workaround has issues as well, though, because records are discarded 
from cache prematurely.

Best Regards
Daniel Ryslink

On Fri, 11 Aug 2006, Wesley Hof wrote:

>
>
> Hi list,
>
> I currently run bind version 9.3.2 on a machine :
>
> Mem:   3108468k total
> Swap:  2097136k total
> Cpu:   dual Intel(R) Xeon(TM) CPU 2.80GHz
>
> I have four of these boxes running as caching nameservers for my customers. I get about 600 queries/second on each machine. After x amount of time my cpu usage peeks to 100%
> resulting in _real_ slow responses, the problem doesn't go away from itself, I need to restart named to get this fixed. All four machines have this issue. I tried several upgrades,
> tuning config options like max-cache-size, max-ncache-ttl ... still no result. I do use an ACL to restrict users based on ip, tried to disable that to for a while, but still with no result.
>
> Is there someone with the same issue ? Or some good tips ? :)
>
> Thanks,
> Wesley.
>
> --
> (o_  Wesley Hof
> //\  UNIX System Engineer
> V_/_ Scarlet Belgium
>
>