Sam Varshavchik wrote:

Rodrigo Severo writes:

Does anybody knows of any suitable solution for this issue besides bind? I want by all means not to go back using bind if at all possible.


Fortunately or unfortunately, bind is the only battle-tested DNS server in existence.

I have just installed bind version 9.2.5. Testing the answers of several repeated queries to this local-caching-only-forwarder-bind the results didn't look random at all (despite being different one from the other) so I looked at bind's documentation. At the end of section 6.2.14.12 I found this:

"BIND 9 currently supports only a "random-cyclic" ordering, where the server randomly chooses a starting point within the RRset and returns the records in order starting at that point, wrapping around the end of the RRset if necessary."

which is a pretty good description of the kind of pseudo-randomness I'm getting from bind9. I am not looking for true perfect randomness. Not at all. My point is that the randomness argorithim bind is using makes almost all answers for a 10 MXs with 2 top MXs have the same order. In fact I believe it makes 90% of them to have the same order.

At ISC's main page <http://www.isc.org> there is a fat warning of the "Hot Topics" section: " BIND4 <http://www.isc.org/sw/bind/bind4.php> / BIND8 <http://www.isc.org/sw/bind/bind8.php> *Unsuitable for Forwarder Use*". So no luck here also.

So I have to ask again, what should I do to get some reasonable randomness in Courier's MX choosing strategy?


Thanks again for any pointers,

Rodrigo Severo



-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
courier-users mailing list
courier-users@lists.sourceforge.net
Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users

Reply via email to