Issue: You're running some Debian distribution, say, Ubuntu Server 18.04 or 20.04. You've specified an IP address parameter in a /etc/bind/named.conf:
query-source address 192.168.30.2; listen-on { 127.0.0.1; 192.168.30.2; };
And, nutz. Now bind9 startup aborts at boot with syslog errors similar to:
Nov 8 11:06:52 cygnus named[1333]: could not get query source dispatcher (192.168.30.2#0) Nov 8 11:06:52 cygnus named[1333]: loading configuration: address not available Nov 8 11:06:52 cygnus named[1333]: exiting (due to fatal error) Nov 8 11:06:52 cygnus systemd[1]: bind9.service: Main process exited, code=exited, status=1/FAILURE Nov 8 11:06:52 cygnus systemd[1]: bind9.service: Failed with result 'exit-code'.
But, you can run the init.d script manually, with:
/etc/init.d/bind9 start
and bind9 starts just fine. You've cranked-up debug for bind9, but logging just keeps reporting [correctly, if maybe lacking in verbosity] the necessary IP address is not available.
You search the InnerTubes and note the issue has been reported as a bug to ISC -- and rejected as not a bug -- multiple times over the past decade. [True; it's not an ISC BIND9 bug.]
Grrr.