ond...@isc.org said:
> perhaps you can try running BIND by hand and see why it crashes on the
> console?
Thank you for nudging me in the right direction. I ran named in the foreground
with -g and could finally see the error:
error short read /dev/urandom: Resource temporarily unavailable
> BTW is there any chance that you and Havard share any common
> bits of configuration?
That would be coincidental; we've not coordinated our configs.
However, it is true that I'm also monitoring my BIND instances,
using collectd, and as far as I know it's using the XML-based
monitoring
Hi Ingeborg,
perhaps you can try running BIND by hand and see why it crashes on the console?
Usually the coredump is controlled by ulimit and it needs the directory
(working directory of the binary, on Linux that would be /prof//cwd) from
where you launch the binary writeable.
If BIND 9
FreeBSD 11.2
Bind 9.12.3-P1
I have a server which runs several nameservers, each in its own chroot with
its own assigned IP-address.
They run like this:
/usr/local/sbin/named -4 -t /resolver -u bind -c /etc/namedb/named.conf
In named.conf:
statistics-channels {
inet port 5 allow
4 matches
Mail list logo
