Last night the name server went down, i.e it stopped having the zones it was supposed to:
Log shows something went terribly wrong:
5-Apr-2017 13:58:21.195 zone domain22.no/IN: loaded serial 2017041136
15-Apr-2017 13:58:21.195 all zones loaded
15-Apr-2017 13:58:21.196 running
16-Apr-2017 03:39:19.651 received control channel command 'reload'
16-Apr-2017 03:39:19.652 loading configuration from '/etc/named.conf'
16-Apr-2017 03:39:19.710 using default UDP/IPv4 port range: [32768, 60999]
16-Apr-2017 03:39:19.710 using default UDP/IPv6 port range: [32768, 60999]
16-Apr-2017 03:39:19.784 sizing zone task pool based on 1 zones
16-Apr-2017 03:39:19.895 the working directory is not writable
16-Apr-2017 03:39:19.897 zone domain1.com/IN: (master) removed
16-Apr-2017 03:39:19.897 zone domain2.com/IN: (master) removed
16-Apr-2017 03:39:19.897 zone domain2.com/IN: (master) removed
..
..
16-Apr-2017 03:39:19.898 reloading configuration succeeded
16-Apr-2017 03:39:19.899 reloading zones succeeded
16-Apr-2017 03:39:19.900 all zones loaded
16-Apr-2017 03:39:19.900 running
I had to run fixdns today morning to get the zones up again.
What could make all the zones be removed 03:39 tonight?