Corrupted Slave Data?
tundra at tundraware.com
Thu May 20 13:30:00 UTC 2021
Running bind 9.16.15 on FreeBSD 11.4-STABLE.
Master is out on a cloud server at Digital Ocean. Slave is on-premise.
All on-prem LANs point to the slave instance.
Running split horizon to keep nosey parkers out of our local DNS assignments.
Recently - and for no obvious reason - the on-prem instance stops resolving
properly. The fix is to stop it, clear out the slave files, and restart.
Then it works for a few days and repeats its misbehavior.
The logs show nothing remarkable, at least at first look.
Is there a known slave file corruption problem?
Could someone kindly suggest things we could look into otherwise?
Many Thanks ...
More information about the bind-users