Flushing the cache

The Time to Live (TTL) factor can also have a bearing on the issues at hand. In instances where a simple dig request will show that the nameserver displays a different record to the local DNS, then (beyond waiting for the automated update to take place) a different course of action is to flush the cache.

In the case of BIND, it is simply a matter of restarting the service like this:

# systemctl restart named

However, without being so drastic, you can also use the following syntax:

# rndc flush

Then run a service status check:

# systemctl status named

In this respect, you should now see the following notices:

received control channel command 'flush'
flushing caches in all views succeeded

Alternatively, you can target a specific domain with the following syntax:

# rndc flushname google.com

And, having run the systemctl status named command, you will see the following reports:

received control channel command 'flushname google.com'
flushing name 'google.com' in all cache views...