Well, it depends on your clients. If they don't like .0 or .255, you would have to have a rather large amount of ranges.
E.g. range 10.1.1.1 10.1.1.254; range 10.1.2.1 10.1.2.254; ...... If OTOH you don't have any of those clients, other factors like hashing algorithms and sizes come into play. This was recently discussed on the list, so there should be information about the optimal way to slice the address range from that point of view in the archives. On 28/09/10 1:08, Warren Kumari wrote: > On Sep 27, 2010, at 6:55 PM, Sten Carlsen wrote: > >> While a single zone is perfectly fine from a standards point of view, "some" >> clients might be served addresses they don't like 10.x.x.0 and 10.x.x.255. >> > But that would be DHCP config, no? > > >> Just a reminder that this could be a reason if something appears weird. > > Fair 'nuff, > > W >> On 27/09/10 23:07, Chris Buxton wrote: >>> On Sep 27, 2010, at 1:03 PM, Christopher Cain wrote: >>> >>> >>>> Hi all. >>>> >>>> I am setting up a new appliance-based DNS solution that will contain a >>>> fair number of separately managed Windows DNS slave servers (in addition >>>> to the DNS appliances that will handle the . >>>> >>>> Currently there are just over 8000 host records that resolve to IP's in >>>> the 10.x.x.x space. I am wrestling with whether or not I should create a >>>> single 10.in-addr.arpa zone or if I should create 256 /16 zones (i.e. - >>>> 0.10.in-addr.arpa to 255.10.in-addr.arpa). >>>> >>>> The reason I want to encompass the entire 10 space is so new arpa zones >>>> will not have to be defined on all servers (specifically on the Windows >>>> slaves) if a new part of the 10 space is used at some point. >>>> >>>> Any recommendations or comments would be greatly appreciated. >>>> >>> There's nothing wrong with a single 10.in-addr.arpa zone. If you need to >>> break it up amongst different master servers, a 10.in-addr.arpa zone can >>> still be used to delegate child zones to their respective servers. >>> >>> You might break it up if, for example, the DDNS traffic from DHCP clients >>> across the enterprise would be too much for one master server to >>> accommodate. The BIND name server writes to its journal file synchronously, >>> for every update, and this can be quite a bottleneck. (The same is true for >>> slave servers, which keep a journal file for zone transfers in order to >>> service IXFR requests sent to them.) >>> >>> Regards, >>> Chris Buxton >>> BlueCat Networks >>> >>> _______________________________________________ >>> bind-users mailing list >>> >>> email@example.com >>> https://lists.isc.org/mailman/listinfo/bind-users >> -- >> Best regards >> >> Sten Carlsen >> >> No improvements come from shouting: >> >> "MALE BOVINE MANURE!!!" >> >> _______________________________________________ >> bind-users mailing list >> firstname.lastname@example.org >> https://lists.isc.org/mailman/listinfo/bind-users -- Best regards Sten Carlsen No improvements come from shouting: "MALE BOVINE MANURE!!!"
_______________________________________________ bind-users mailing list email@example.com https://lists.isc.org/mailman/listinfo/bind-users