Lowell Gilbert wrote:
Josh Tolbert <[EMAIL PROTECTED]> writes:


Hello,

I'm running DHCP + dynamic DNS here on my home LAN and I've noticed a problem
that needs a manual fix every time the DNS machine gets rebooted. It doesn't
happen very often, but it does happen. :)

My firewall/gateway machine runs FreeBSD-5.4-RELEASE of some patchlevel. It
uses ISC DHCPD from ports to update my DNS server, another FreeBSD machine
(now running 6.0-RELEASE) with new entries when machines register with the
DHCP server. The problem arises because by default named runs -u bind, however
/var/named/etc/namedb/master is owned by root. I believe this is caused by
/etc/mtree/BIND.chroot.dist, since I'm running bind chrooted (the default
setup). When the DNS machine reboots, I have to manually chown
/var/named/etc/namedb/master (or /etc/namedb/master) to bind before updates
will continue, otherwise I see errors such as

named[297]: dumping master file: master/tmp-QQ2UU6pWaZ: open: permission denied

Is there any good workaround for this issue? I'd like to keep bind running as
the bind user as well as keep bind chrooted if possible. I know I could edit
the mtree file on my machine, but that seems somewhat kludgy to me.

Thanks for any help/advice you can give me,


Normally mtree is only automatically run by installworld. Is that what causes the permissions to be reverted?
If so, then change the mtree file (and keep the modifications over
time when you run mergemaster).
If not, then figure out what *is* changing the permissions.

This happened to me too; everytime named started it would change back the owner of the "master" directory from "bind" to "root" according to the mtree file.

In the end I just used the "dynamic" folder to store my dynamic zones in with "bind" as owner, which makes more sense, and also doesn't get its user changed by the mtree.

_______________________________________________
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"

Reply via email to