I have three machines on the network. Master at the NOC and two slaves at towers. They handle our domains, PTRs, etc. As well as DNS for customers. Recursion is locked down to our address blocks only. I also have an anycast address shared between all three. The infrastructure devices use that for lookups.

Use BIND views to separate things if you're paranoid.

On 9/6/2016 2:26 PM, Josh Baird wrote:
I wouldn't be overly concerned about your recursive boxes being authoritative for your internal (only) zones. You already have mechanisms in place to prevent external clients from using them for recursive services.

On Tue, Sep 6, 2016 at 3:20 PM, That One Guy /sarcasm <thatoneguyst...@gmail.com <mailto:thatoneguyst...@gmail.com>> wrote:

    Im putting our recursive sservers up for our network to use,
    theyre access limited by ACL and external router firewall policies
    to our networks only

    There will be four total servers NS1 and NS2 are our current
    authoritative only servers, they are public facingfor our domains
    and our ARIN allocation

    I read many conflicting best practices, so ...

    NS3 and NS4 I am tempted to make slaves to NS1 (its the master for
    all zones) and put our RFC 1918 space on NS1, however this creates
    a security dilema in that a new bind vulnerability could expose
    our internal space structure, not that its a huge deal today, I
    would prefer to not have made a poor choice for ease today that
    causes a problem down the road.
    Im tempted to delegate a subdomain (infrastructure.domain.com
    <http://infrastructure.domain.com> or whatever) to NS3 for rfc1918
    record, but then that puts authoritative master zone records on a
    recursive server which all the best practices suggest avoiding.

    I suppose i can put forwarders in for this up to NS1/2 on the
    recursive servers and use bind views to limit the internal zones


    What is recommended in this scenario?

    Also, with a set of recursive servers, is it possible to sync the
    cache between the two so I can load balance the servers (we wont
    likely ever have enough load from our network for it to ever be an
    issue)

-- If you only see yourself as part of the team but you don't see
    your team as part of yourself you have already failed as part of
    the team.



Reply via email to