Is this possible

Danny Mayer mayer at gis.net
Wed Mar 24 16:32:30 UTC 2004


At 12:33 PM 3/23/2004, Kevin Darcy wrote:
> >  The
> >contents of the "SOA" resource record should be treated as private to each
> >particular replication mechanism, and one must not expect different DNS
> >database replication mechanisms to use all of the fields in the same way, or
> >in a way that is compatible with one another, or even to use them at all.
> >
>The standards dictate that the serial number is incremented when the
>zone contents change. How hard is that to get right? Why would you
>defend such total incompetence on the part of Microsoft? (Actually, I'm
>being generous to Microsoft here; some would accuse them of deliberate
>sabotage). As long as the serial number is handled properly, and zone
>transfers work, then all of these other "private" replication schemes
>interoperate with the standards-based one.

While I agree that this is bad, the problem that Microsoft has is that
the SOA record is special in the sense that when doing replication
the CONTENTS of the record needs to be examined. Replication
usually just involves figuring out which record is more recent when
there are different servers with a changed record. That's what's
causing the problem. It's easy to accuse Microsoft of underhandedness
or incompetence but in this case it's one of those overlooked issues
when they don't see the problem on their own internal systems which
are all multimastered. It's actually a hard problem to solve. I believe
that Barry mentioned that they had solved it.

Danny




More information about the bind-users mailing list