9.2.5 db causes high cpu? was: Re: BIND 9.2.5rc1 is now available.

JINMEI Tatuya / 神明達哉 jinmei at isl.rdc.toshiba.co.jp
Mon Feb 21 00:44:39 UTC 2005


>>>>> On Sun, 20 Feb 2005 23:47:21 +0100, 
>>>>> Brad Knowles <brad at stop.mail-abuse.org> said:

>> But could you be more specific on what threaded
>> BIND9 can provide on a single CPU?  In my understanding, BIND9
>> carefully decomposes its tasks into small chunks and avoids using
>> blocking APIs so that every single task (in BIND9) will not make
>> others wait for a long period.

> 	My understanding was that BIND9 had all the same blocking issues 
> as BIND8, unless run with threading enabled.

For example? (I'm not very much familiar with the blocking issues of
BIND8)

> Perhaps this is 
> different for BIND 9.2.x versus 9.3.x?  Or maybe I just missed 
> something really fundamental?

>> Regarding the former, I admit my experiences in this area are limited,
>> but from my experiences with {FreeBSD, Linux, Solaris, True64} on
>> {Xeon, opteron, sparc, alpha}, I've almost never seen the case where
>> threaded BIND9 provide better performance than a single process of
>> BIND8...

> 	On a single CPU, yes -- that is to be expected.  On two or more 
> CPUs, a single instance of BIND9 should outrun a single instance of 
> BIND8, everything else being equal.

(I'm definitely concentrating on the case of multiple CPUs here)
Sorry, I'm still not sure whether you are saying

a) the current BIND9 implementation with threading can provide better
   performance than a single process of BIND8, or
b) future BIND9 implementations should provide better performance than
   a single process of BIND8.

I said I had (almost) never seen case (a) with the current BIND9
implementation.

I also said I agreed with opinion (b).

					JINMEI, Tatuya
					Communication Platform Lab.
					Corporate R&D Center, Toshiba Corp.
					jinmei at isl.rdc.toshiba.co.jp


More information about the bind-workers mailing list