Speed of BIND vs. W2k DNS

[David R. Conrad]

Hi,

At 05:51 PM 12/23/2000 +0000, alevey wrote:
>Unsecure updates work fine between BIND & W2K-DNS

Nice to hear.  It would be nicer if MS would support HMAC-MD5 TSIG, the 
mandatory TSIG algorithm, so the updates could be secured.

>You missed a HUGE aspect of W2K DNS... multi-master. No single point of
>failure, unlike BIND.

You're right, forgot about that.  It would be interesting to see how W2K 
DNS's multi-master approach deals with network partition scenarios where 
the masters can't sync.  I suspect, but don't know enough about W2K DNS to 
know for sure, that they have not solved this problem, just moved it 
someplace else.  If they have not solved that problem, then they are 
violating the DNS specifications (and also setting the stage for strange 
results).

>People have seen problems using BIND and DDNS in a W2K enviro. Mostly with
>BIND not taking the updates properly. A records I think....

Casting the problem as BIND not taking the updates is pre-supposing that 
W2K is sending correct updates.  In any event, this is the first I've heard 
of such a problem.  More detail would be helpful.  We did find problems 
with BINDv9 and W2K DNS due to Microsoft sending two bytes of garbage (not 
surprisingly, the ASCII characters "M" and "S") in zone transfers (or 
something like that, I've forgotten the details).  We solved the problem by 
relaxing BINDv9's conformance to the specs in that area...

Rgds,
-drc