restructuring/redesigning BIND (was Re: 9.2.5 db causes high cpu?)
paul at vix.com
Tue Feb 22 22:21:42 UTC 2005
> > named in BIND9 starts up pretty quick, if you mean "starts to read its
> > zone files", or even "starts to answer queries" if you don't have a lot
> > of zone-data to read in.
> Well.... yeah, ok. But I still think it takes too long to load zone
> files and answer queries.
so do i. and we're working on that. but moving logic into DSO's won't help.
rather, we need offline precompilation of zones into binary files, like NSD.
that way we won't spend a lot of nameserver CPU cycles parsing RFC1035 zone
> The fact that named blocks completely until all zones are loaded is
yes, it is. maybe you can help me understand the alternatives better? when
we get a query for a zone we havn't loaded yet, should we answer SERVFAIL or
should we just silently drop the query?
> > now that the apache2 people have done the hard work of figuring out now
> > to do "DSO"'s on many different systems, i am ready to agree with you
> > for things like DLZ. "define a plug-in architecture" was on the roadmap
> > draft i last saw.
> Neat. Is that on the ISC site somewhere? :-o ;-)
the roadmap is available to members of the BIND Forum. i've been told that
sgi is considering rejoining. note that the place i saw it was a draft, and
would have to be seen and discussed by the forum before it could become real.
More information about the bind-workers