Hi All, I am currently building a Network Access Control system, and in order to keep it "out of band" (via a layer 3 firewall), I would ideally like to use a DNS redirect to send people to the NAC server from a registration VLAN. I am having issues with doing a redirect with some exceptions (the registration VLAN needs access to the University Shibboleth servers and the IT registration pages which are outside the College network).
Now I realise that I am not the first person to try and do this, so I searched the mailing list for similar discussions. https://www.unbound.net/pipermail/unbound-users/2010-April/001134.html https://www.unbound.net/pipermail/unbound-users/2010-May/001171.html Based on what I found (and read in the annotated unbound.conf file) I realised that something like this should work: local-zone: "." redirect local-data: ". A <NAC server ip>" local-zone: "google.co.uk" transparent This however doesn't seem to work as I would expect it to, as everything is redirected by the local-data to the NAC server ip. (note: changing this to "refuse" rather than "redirect" works as expected, can connect to google.co.uk, get refused for everything else) I thought this might be a version issue, as CentOS 7 is packaged with an older version (1.4.20??) and I know that in recent versions additional options were added for the zone types. So I compiled 1.6.0 from source and experienced the same behaviour, even when attempting to use always_transparent , I tried all sorts of other iterations of options and none worked as I had hoped... Noticing that I can find multiple references to the above example, has the behaviour of Unbound changed? If so how do I accomplish the above, I would expect the "always_transparent" would have been the answer if the local-data was the cause of the behaviour: "always_transparent Like transparent, but ignores local data and resolves normally." But this still doesn't work as expected when using a redirect. Any help would be greatly appreciated! Many Thanks, Simon Wedge St Antony's College University of Oxford (apologies if you get two copies of this message)
