reload but the old value linger

[Boylan, Ross]

My fix for the DNS lookup problems I reported a few days ago, based on help here, seems to mostly work.  But there is one oddity.  When the tunnel goes down I comment out the special handling for the zone I reach through the tunnel and reload the server.  But my DNS queries return the same internal IP number I got before, at least for awhile.

Since I can't reach the remote machine anyway, this is probably a pretty minor problem, but I'd like to understand what's going on and how I might fix it.

My theory is that reloading (via rndc reload) does not clear the cache, and that my queries just get the cached value until they expire.  Is that plausible?

If that is the problem, would
rndc flushtree ucsf.edu inside
remove the no longer valid values from the cache?  ucsf.edu is the domain for which I forward, and it is accessible from the "inside" view.

Details:
My main bind configuration includes
view "inside" {
	match-clients { internals; };

	recursion yes;

	// next is only active when vpn tunnel is up
        //when tunnel goes down it is commented out
	include "/etc/bind/ucsf.conf.tunnel";
	
	// allow dhcp to update me
	include "/etc/bind/rndc.key";

	include "/etc/bind/named.conf.default-zones";

	zone "1.168.192.in-addr.arpa" {
# stuff
	}
# and a forward zone
};

lwres {
   view "inside";
   search { "betterworld.us";};
};


And the sometimes included file is
--------------------- ucsf.conf.tunnel----------------
zone "ucsf.edu" {
     type forward;
     forwarders {10.10.10.10;};
     };
------------------------------------