Re: [Nolug] DNS/Reverse Lookup and a Host file.

"Wimprine, Thomas" <twimprine@stei.com> writes:

> We have a management/monitoring station running HP Openview that's managed
> my a contractor that supports all our routers for the company (550+). They
> maintain a host file on that station (NT4) to assist in resolving the names
> of the routers. We were not interested in having all that data in DNS. I
> brought up a new DNS server on Linux running BIND and it's been running
> flawlessly (probably because nobody knows how to mess it up!). Now they
> wanted me to take that host file and import that data into DNS nightly.
> Internally we have two domains and old one we are migrating from
> (stewartenterprises.com) and the new one we are migrating to (stei.com).
> These records need to be consistent across both domains, and the reverse
> lookup.

Are you using Bind9 views? This comes in really handy if you want to
split your DNS and have a more detailed internal DNS view and an
external view for the rest of the world. Of course I prefer djbdns
for doing this, but I have implemented Bind9 also.

Also, have you considered putting these particular entries from the
hosts file in a subdomain? Maybe later the architecture changes (you
use Dydns or something else) so you could keep this isolated from the
parent domain. something like routers.stei.com. That way you can
maintain a separate config area in BIND and separate zonefile
information.

> Solution (not totally implemented yet)
> I have an event run at night that copies that host file from the NT system
> to a samba share on the DNS server. From that file I have a cron job that
> copies that to /etc/hosts. There is a utility that is downloadable from the
> Oreilly ftp site called h2n. This will parse the hosts file rearrange it so
> it's readable by BIND, however it overwrites the file that was previously
> there and I need to share it between two domains. It outputs db.stei and

This is a big problem actually. Since h2n is meant for primary
imports, it doesn't increment the Serial # in the zone file. This
means that requestors won't always realize that there's been an update
and you'll get stale information. I'm simplifying here, but basically
you need a way to increment the serial # each time you update your
zone via cron. Modify h2n to get the current serial # first,
increment it, and that write out that number in it's update.

-- 
Scott Harney<scotth@scottharney.com>
"...and one script to rule them all."
gpg key fingerprint=7125 0BD3 8EC4 08D7 321D CEE9 F024 7DA6 0BC7 94E5

___________________
Nolug mailing list
nolug@nolug.org