Multinet load balancing - Doesn't seem to work with client running bind 9.1.3rc1

[Mark_Andrews at isc.org]

> 
> Hi there,
> we have a couple of Dec boxes running Multinet v4.1
> and bind version: 
> 4.9.7
> 
> We have a couple of Unix boxes also acting as secondary name servers.
> One of them runs Solaris 2.5 and bind version:
> 4.9.3-P1
> 
> The other runs Solaris 2.7 and bind version:
> 9.1.3rc1
> 
> We have DNS load balacing set up on the Dec servers.
> The zone hosts table on the unix boxes has the following
> entries (as suggested in the Multinet manuals)
> vaxcluster   IN NS ourvax1.foo.com
>              IN NS ourvax2.foo.com
> 
> If I now do (on the Solaris 2.5 box) an:
> nslookup vacluster
> I get the following output:
> 
> Server:  sol25.foo.com
> Address:  999.999.999.10
> 
> Name:    VAXCLUSTER.FOO.COM
> Addresses:  999.999.99.32, 999.999.999.33
> 
> This is exactly the response we want.
> 
> On the Solaris 2.7 box we get:
> # nslookup vaxcluster
> Server:  sol27.foo.com
> Address:  999.999.999.11
> 
> *** sol27.foo.com can't find vaxcluster: Server failed
> 
> Does anyone know why this is happening on the Solaris 2.7 box
> and is there any way to fix this on that box ?

	Turn on debugging.  Find out which query is failing.  Then
	findout why the server is failing for this lookup.  If it
	is a zone that is being served look at the server logs to tell
	you why the zone was rejected.
> 
> Many thanks for any suggestions.
> 
--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews at isc.org